THE VALUE OF OUR THE VALUEDIGITAL OF OUR DIGITAL IDENTITY IDENTITY

Policy Series TABLE OF CONTENTS TATABLEBLE OF CON OFTEN CONTS TENTS

PART 1 KEY MESSAGES PART 1 Digital identity: A driver of growth, a consumer hazard or both? Digital identity: A driver of As the volume and variety of data grows, so does its value ...... 23 growth, a consumer hazard Consumer research reveals a deep lack of trust ...... 26 1 or both? PART 2 A definition of digital identity and how consumers really see it EXECUTIVE SUMMARY Understanding how consumers view their digital identity ...... 37 Few consumers know how their data is used – fewer can control it ...... 39 Many factors impact consumer decision making ...... 42 5 19 Consumer perspectives on the proposed EU privacy regulation ...... 47

PART 2 PART 3 PART 3 A definition of digital The value of digital identity – The value of digital identity – for organisations and consumers Traditional production ...... 56 identity and how consumers for organisations and Retail ...... 62 really see it consumers Financial services ...... 67 Telecommunications and media ...... 73 Public services / health ...... 77 Web 2.0 communities ...... 85 E-commerce ...... 91 33 53 Online info / entertainment ...... 95 PART 4 PART 4 Part 5 The megatrends of digital identity The megatrends of digital A practitioner‘s guide to the A growth driver for a stagnant economy ...... 101 identity future of digital identity The six key trends in digital identity applications ...... 105

PART 5 A practitioner‘s guide to the future of digital identity Innovative approaches to building trust ...... 113 Guiding principles to unlock the value of digital identity ...... 114 99 109 APPENDIX ...... 117 Key Messages

key messages

KM1 | 2 KEY MESSAGES Key Messages KKEYEY MESS MESSAGES AGES

• In an increasingly digital society, personal data • All told, the value created through digital iden- • Most consumers have little idea what happens • Overall, given proper privacy controls and suf- has become a new form of currency. The biggest tity can be massive – at a 22% annual growth to their data. Only 30% have a relatively com- ficient benefits, most consumers are willing to challenge for political and business leaders is to rate, applying personal data can deliver a €330 prehensive understanding of which sectors are share their personal data with public- and pri- establish the trust that enables that currency to billion annual economic benefit for organi- collecting and using their information. Indivi- vate-sector organisations. They want to spend keep flowing. sations in Europe by 2020. duals with higher-than-average awareness of their new currency on deals that they like. data uses require 26% more benefit in return for • “Digital identity“ is the sum of all digitally • Individuals will benefit to an even greater de- sharing their data. • To unlock value, organisations need to make the available information about an individual. It is gree, as the consumer value will be more than benefits of digital identity applications very clear becoming increasingly complete and traceable, twice as large: €670 billion by 2020. The com- • Few individuals are in control of their digital to consumers. Further, they need to embrace the driven by the exponential growth of available bined total digital identity value could amount identity. Just 10% of respondents had ever done new digital identity paradigm of responsibility, data and the big data capabilities to process it. to roughly 8% of the EU-27 GDP. six or more out of eight common privacy-pro- transparency and user control. tecting activities (e.g., private browsing, disab- • How digital identity develops is an important • However, two-thirds of potential value genera- ling cookies, opt-in/out). However, consum- • Privacy is increasingly becoming an area of com- concern for consumers and companies alike. tion – €440 billion in 2020 – is at risk if stake- ers who are able to manage and protect their petition for organisations, which can differen- Individuals are worried about losing both their holders fail to establish a trusted flow of data. privacy are up to 52% more willing to share tiate themselves by providing the right privacy privacy and control over their personal data. information than those who aren‘t – presum- controls and privacy-by-default product design. Companies, on the other hand, fear that mis- • Digital identity is relevant not just to Web 2.0 ably because they can adapt their data sharing Indeed, such a scenario may well play out in the steps – be it through their own, or third-party companies, but to the economy as a whole. The to their individual preferences. desktop and mobile Web browsing market. applications – could compromise their position public sector and health care industry stand to as trusted provider. profit most from personal data applications and • Trust differs per sector: Consumers are on • Policy makers and regulators need to ensure are expected to realise 40% of the total organi- average 30% more willing to share data with adequate privacy safeguards and maintain a • Global trends including the social media boom sational benefit. e-commerce companies, cable operators and flexible approach that will encourage new appli- (a quarter of the world‘s population will be automobile manufacturers than with Web 2.0 cations and allow consumers to make their own members of online social networks by 2015) • While the retail and Internet sectors already communities. informed choices on the extent to which they and the burgeoning “Internet of things“ (some extensively use personal data, other sectors like wish to generate value from their digital identity. 75 million machine-to-machine connections traditional production and the aforementioned • The “right to be forgotten“ has a small but con- will be added in Europe by 2015) result not public services are in the infancy of digital iden- sistently positive impact on the willingness to • Political intervention must account for shifting only in increased information volume, but also tity value generation. share, increasing it by 10% to 18%. levels of acceptance regarding digital identity completely new types of data. applications and the developments in the glo- • Consumer concern has grown along with the • The preferred form of consent strongly depends bal marketplace. Failure to do so may hamper • From a macroeconomic perspective, it becomes increased use of personal data. 88% of people on the type of data: Opt-out is considered innovation as well as the competitiveness of do- clear that digital data is already a growth driver who are online consider at least one industry a adequate for less sensitive data by 69% of re- mestic industries. Finding the right balance can in an otherwise flagging economy. While tradi- threat to their privacy. Yet consumer research spondents, while opt-in is required for highly spur local investment in digital identity appli- tional industries shrank by up to 3.6% from conducted for this study shows that stated con- sensitive data by more than 80%. The issue of cations and also attract industries to European 2008 through 2011 in Europe, data-intensive cerns by individuals do not necessarily result in consent highlights a key dilemma: Control is markets that provide a safe haven for personal sectors – where the use of digital identity is a key behavioural change. important to consumers (82%), but so is conve- data. component of business – thrived with annual nience (63%). growth rates between 15% (e-commerce) and up to 100% (Web 2.0 communities).

3 | 4 Executive Summary

executive summary

ES5 | 6 EXECUTIVE SUMMARY Executive Summary EEXXECUECUTIVE TSUMMIVEAR SUMMY ARY

“Digital identity“: A driver of growth, a con- Views on digital identity tend to take one Digital identity can drive massive growth in an It is important to realise that, sumer hazard or both? of two extremes: Let organisations do what they otherwise stagnant European economy although the public debate on mone- need to in order to realise the economic poten- tisation of personal data is dominated Increasingly, we are living double lives. tial of “Big Data,“ or create powerful safeguards Taking a macroeconomic perspective, it by the perspective on Web 2.0 players, There is our physical, everyday existence – and to keep private information private. But digital becomes clear that digital data is already a growth digital identity is relevant, and impor- there is our digital identity. Most of us are likely identity can‘t be cast in such black-and-white driver in an otherwise flagging economy. These tant, for the whole economy. more familiar with that first life than with the terms. While consumers voice concern about the have not been the golden years for the traditional second, but as the bits of data about us grow use of their data, their behaviours – and their production, retail, financial services and tel- It is important to realise that, although the and combine in the digital world – data on who responses to a survey conducted specifically for ecommunication and media sectors, which saw public debate on monetisation of personal data is we are, our history, our interests – a surprisingly this report – demonstrate that they are willing, compound annual growth rates (CAGR) of 0% to dominated by the perspective on Web 2.0 players complete picture of us emerges. What might also even eager, to share information when they get -3.6% from 2008 through 2011 in Europe. Data- and the digital economy, digital identity is not just be surprising for most consumers is just how ac- an appropriate benefit in return. Indeed, as Eu- intense sectors, on the other hand, where the use the realm of the Facebooks and Googles of the curate and traceable that picture is. ropean Commissioner for Justice Viviane Reding of digital identity is a key component of business, world. It is relevant, and important, for the whole remarked, “Personal data is in today‘s world the performed strongly over the same period: Web economy. In fact, it is the public sector and health In this report, we introduce the concept currency of the digital market. And like any cur- 2.0 communities saw an annual revenue growth care that stand to profit the most from personal of digital identity – the sum of all digitally avail- rency it has to be stable and it has to be trust- of about 100%, while the e-commerce sector had data applications – potentially realising 40% of able information about an individual. How exact- worthy.“1 This is a crucial point. Consumers will a CAGR of 15% and the online information and the total organisational benefit. ly digital identity develops is a massive concern “spend“ their personal data when the deals – and entertainment sector of 22%.2 for consumers and companies alike. Individuals the conditions – are right. The biggest challenge The benefits of digital identity are relevant are worried about losing their privacy and see- for all stakeholders is how to establish a trusted In an overall stagnant European economy, and important for consumers as well. Indeed, BCG ing their personal data used against them – for flow of this data. the applications built on the use of digital identity estimates that the consumer benefit will be more instance by insurers or employers discriminating can drive massive value growth for both public- than double the organisational value – €670 billion against them based on personal data floating This study aims to reshape the discussion. and private-sector organisations: At a 22% annu- a year by 2020 – mainly stemming from reduced around the Internet. Companies, on the other To give business and political leaders a better un- al growth rate, the annual economic benefit can prices (passed on by companies seeing data-driven hand, fear the backlash of approaching digital derstanding of this new phenomenon, it quanti- reach €330 billion by 2020. cost savings), the time savings that self-service identity the wrong way, be it through their own fies, for the first time, the current and potential transactions will bring and the high value individ- applications or third-party applications that run economic value of digital identity across the econ- Those applications can help tackle some of uals place on free online services, supported at on their networks, platforms or channels. omy as a whole and explores how consumers society‘s most pressing problems, too. In the health least in part by the use of personal data. really view their personal data. An innovative care sector, data-driven applications like person- At the same time, the volume of data that survey of 3,000 individuals in Europe looked at alised medicine, decision support systems and Sectors are at different stages of generating is available, collected and analysed is growing their actual decision making, drilling down on electronic records shared among providers can im- value from personal data rapidly, thanks to ubiquitous connectivity, vast the factors that spur, and hinder, the sharing of prove quality and efficiency of care – particularly increases in processing power and the growth data. As a result, we were able to develop a new important goals given Europe‘s ageing population. Eight sectors collectively cover the whole of new sources like social media, online transac- paradigm for unlocking the value of digital iden- In the public sector, personal data-driven initiatives economy, but individually face very different tions, in-product sensors and digital media. The tity in a sustainable, consumer-centred way. These to fight tax evasion and citizen self-service portals, conditions and challenges in applying digital ability to put this data to good use – not the uses tools and insights, we hope, will contribute to the where tasks like driving license renewals can be identity: traditional production, retail, financial consumers fear but ones that can benefit them – development of digital identity applications and performed online, can increase tax revenues and services, telecommunications and media, public is growing, too. policies that strike the right balance – and keep lower spending, giving governments some relief services/health, Web 2.0 communities, e-com- the data flowing. from the budget pressure they are under. merce and online info/entertainment.

1 In January 2012 at the presentation of the proposal for a new General Data Protection Regulation 2 Eurostat; BVH; comScore; BCG analysis

7 | 8 EXECUTIVE SUMMARY Executive Summary EEXXECUECUTIVE TSUMMIVEAR SUMMY ARY

Evolutionary path of amount of available data, which is expected to be a member of a social network and will to some digital identity value grow by 45% per year through 2015, to roughly extent share personal information, including dem- creation Internet Retail sectors 7 zettabytes. To put that in perspective, that is ographical data, likes and preferences, images, Financial more than 1,000 gigabytes of data – twice the video and even location data.3

alue services V capacity of a standard laptop – of data for every Telco & media III person on earth in 2015. Second, there is the The other trend leading to increased data Tapping data ecosystem rapidly improving ability to process and analyse variety is the “Internet of things.“ In-product sen- opportunities II that data. The “data explosion“ is driven by global sors that can “call home“ via the Internet and relay I Leveraging personal Creating basic digital data internally trends resulting not only in increased volume, but usage data are becoming increasingly prevalent. product experience also completely new types of data. Automobiles, home appliances and energy meters External applications are among the traditional product categories that Traditional Sharing data with third parties production The “data explosion“ is driv- have – or soon will have – integrated links to the Internal enhancement in both directions Public en by global trends resulting not Internet. Not including phones or PCs, we expect Usage data for R&D, delivery sector optimisation etc. only in increased volume, but also to see an additional 75 million devices with direct Digitisation completely new types of data. connections to the Internet in Europe by 2015.4 Basic digital identity uses as secure authentication

Time The social media boom is perhaps the For both organisations and consum- most important of these trends, as it has allowed ers, this increased data volume has the po- Digital identity intensity formerly passive data subjects to actively share tential to create significant value, which will and distribute information on a large scale. By manifest itself in a variety of ways. By creat- Each of these sectors is at a very different tomation are currently the major applications of 2015, one quarter of the world‘s population will ing more personalised products – and lever- stage of unlocking value through personal data ap- digital identity, we expect the focus to shift to inno- plications. vative new services and enhanced user experience. Global data explosion 2015: 7.91 × 109 TB Retail and the three Internet sectors are al- The traditional production and pub- Global data transaction volume in terabyte (TB) ready at a relatively advanced level; these organi- lic sectors, meanwhile, are at the very be- 1010 sations routinely leverage personal data internally ginning of the path. Many of these organ-

for R&D to optimise the delivery of their goods and isations are only now starting to create a 109 services and for other purposes. They often tap into basic digital product experience or to embrace × 16,000 Social media boom the wider ecosystem of data-related opportunities digitisation – such as government agencies 108 as well, such as sharing – and selling – information and health care providers migrating to electronic Internet of things to third parties. records and setting up processes that leverage this 107

data along the value chain. 5 1990: 4.97 x 10 TB Online data transactions Companies in the telecommunications and 106 media sector, including cable operators, have access The volume – and variety – of data is expanding Digital service and media to a vast amount of personal data in principle – 105 mainly relating to telephony and Internet usage The economic growth prospects for dig- 1990 1995 2000 2005 2010 2015 or TV viewing – but so far make relatively little use ital identity are fuelled by two important fac-

of it. While customer self-service and process au- tors: First, there is the exponentially expanding Source: comScore; Facebook; Sandvine: “Information White Paper,” 2011; Sandvine: “Digital Contents White Paper,” 2010; METI E-Commerce survey; IDC; consumer research; BCG analysis

3 Source: Morgan Stanley; Strategy Analytics; eMarketer; UN “World Population Forecast”; BCG analysis 4 Source: ABI; Berg Insight; IDC; Yankee Group; Gartner; Analysys Mason; BCG analysis

9 | 10 EXECUTIVE SUMMARY Executive Summary EEXXECUECUTIVE TSUMMIVEAR SUMMY ARY

aging data to enhance existing products – person. That spurs sales for the organisation that selling it can open up a significant new High as these numbers are, they are hard- businesses can create new revenue streams and and satisfaction for the consumer. Traditionally, revenue stream. The online sectors pioneered ly surprising considering the high-profile data expand old ones. Research and development can personalised products came about through be- the secondary monetisation of personal infor- breaches of recent years. When Sony‘s PlayStation be improved, meaning more relevant and more at- spoke solutions at high cost premiums and for mation, but it is a trend taking hold – and rap- Network was breached in April 2011, it didn‘t just tractive products – which benefit all sides. small target audiences. Digital identity enables idly – in other sectors as well. Even the public result in a few user accounts being compromised. personalisation on a wide-scale, cost-effective sector has benefitted from putting its data up Personally identifiable information of 77 million The six key trends in digital identity applications basis. Amazon‘s dynamically generated recom- for bids: In the United States, the state of Flor- subscribers was stolen – a breach so significant mendations – based on a user‘s purchasing and ida took in more than $60 million in 2010 by that the entire service was shut down for 24 days While the actual applications of personal browsing history – are estimated to generate selling information it collects from drivers. and Sony incurred direct costs associated with the information vary from sector to sector, and even some 25% of the e-commerce giant‘s sales. incident of ¥14 billion, or roughly €128 million.7 from organisation to organisation, they invariably Personal data uses concern most individuals, fall into one of six categories. These are the mega- • Enhanced delivery. The availability of both few are fully aware and in control of them Incidents like these also heighten consum- trends of digital identity: more volume and more types of data enables ers‘ more subjective fears that new technologies organisations to increasingly extract operational Along with use of personal data, the con- like RFID, face recognition and user profiling will • Process automation. By integrating personal insight from it. This allows them to improve cerns of people have also grown. The majority of unravel the anonymity and privacy they have data into their processes, organisations can au- the delivery of goods and services. For instance, individuals have concerns about the use of per- long enjoyed – or at least thought they enjoyed. tomate – and simplify and speed up – transac- IBM and the U.S. insurer Wellpoint are devel- sonal data by companies. A major Europe-wide Data collection and use is rarely transparent to tions that traditionally required users to identify oping a system to analyse huge volumes of study, Eurobarometer5, put the figure at 70%, users. Indeed, while the BCG survey found that themselves, express their preferences or perform data from many sources – including research while BCG‘s own representative survey of more most individuals are concerned about their per- some kind of manual action. For example, the papers, medical textbooks and population health than 3,000 European consumers, conducted for sonal information, far fewer realise the extent to Oyster Card carried by London commuters au- data – to detect diseases and suggest treatment this report in August 2012, revealed that 88% of which organisations acquire and leverage it. Only thenticates them when they board buses and options. Such solutions can lead to many more people who are online consider at least one in- 30% of respondents had a relatively comprehen- trains, automating ticketing and access control. patients benefiting from optimal care. dustry or sector a threat to their privacy.6 sive understanding of which sectors were collect- ing and using their information. • User enablement. Digital identity gives in- • Personal data-driven R&D. Using a range of dividuals the ability to perform transactions data sources – such as product reviews, com- Share of concern, autonomously, without any human assistance. ments posted on social media sites and usage awareness and control Customer self-service is a prime example. data from in-product sensors – companies can It continues to grow strongly in sectors where also gain insights that let them better focus Individuals‘ it has been applied for decades and is being their research efforts and shorten development privacy stage Concern Awareness Control introduced in more and more new industries. cycles. This trend has already been adopted by Denmark, for example, has one of the most major consumer goods companies, including advanced e-government programmes in the Procter & Gamble, whose Vocalpoint P&G com- Regard at least one sector Aware that > 50% of sectors Ever done > 3/4 of activities to world and generates €200 million in annual munity gets customers more deeply involved in as privacy threat collect personal data manage their privacy savings with electronic invoicing alone. the development and testing of new products. Share of EU • Personalisation. Customising products and • Secondary monetisation. The data that an or- online population 70 – 90% 15 - 30% 0 - 10% services to a particular individual‘s preferences ganisation holds, or the insight derived from it, and needs makes them more relevant to that is valuable to other parties – valuable enough Source: BCG digital identity survey (n = 3,107, August 2012)

5 Special Eurobarometer 359, field work conducted 2010, published 2011 6 BCG digital identity survey (n = 3,107, August 2012) 7 Sony‘s script of revisions to reporting of fiscal year 2011 published on 23 May 2011 11 | 12 EXECUTIVE SUMMARY Executive Summary EEXXECUECUTIVE TSUMMIVEAR SUMMY ARY

Fewer individuals still are both aware of User controls increase consumers‘ willingness are able to manage and protect their privacy • Data types, sectors, collection methods and data use and able to exert some control over it. to share data (through common actions such as changing pri- uses. Consumers also make clear distinctions When we asked respondents how many of eight vacy settings in a social network or opting out of between different types of data, industry sec- common privacy-protecting activities they had Our research indicates that the level of anx- certain data uses) are up to 52% more willing to tors, collection methods and the uses to which ever performed – actions like changing the pri- iety consumers express regarding their privacy is share information than those who aren‘t. data is put. Their level of trust differs depend- vacy setting in a social network service, or surf- not a strong indicator of their actual data-sharing ing on the sector, for example: Consumers, we ing the Web in “private“ mode – just 10% said behaviour. Different groups of consumers ex- Consumers who are able to saw, are on average nearly 30% more willing to they had ever done six or more of these things. pressed very different levels of concern. Some were manage and protect their priva- share with e-commerce companies, cable oper- Part of the explanation, of course, is that many concerned about the data use of just one, or none, cy are up to 52% more willing to ators and automobile manufacturers than with data-collecting organisations do not offer any of the covered sectors; others were concerned by all share information than those who Web 2.0 communities. options or settings enabling users to manage or nearly all of them. But we found that their actual aren‘t. their privacy. privacy valuation did not differ. Privacy, it seems, is Exhibit Sensitivities of data types shows comparable to climate change: Stated concerns by Ease of use is also critical: For low- and how consumers differentiate between data types. While few consumers are fully aware of individuals do not necessarily result in a change in medium-sensitivity data, the easier consumers While low-sensitivity information is shared by how their data is being used and even fewer are behaviour in their day-to-day lives. perceived those controls to be, the more they more than 40% of people even without receiving capable of actively managing it, most individuals shared. For sharing age and gender information, a direct benefit, most individuals are unwilling expect their digital identity to be treated respon- Different groups of consumers for example, individuals who had a relatively easy to part with highly sensitive data even for large sibly and openly: 79%, for example, said com- expressed very different levels of time managing their privacy were 37% more will- rewards. It should be noted that the research meth- panies should be more transparent about use of concern. But we found that their ac- ing to share than those who had a relatively dif- od we applied provides very meaningful data on personal data. tual privacy valuation did not differ. ficult time. the relative differences, but the absolute monetary

Leveraging digital identity means understand- However, other factors do significantly in- ing how consumers really view it fluence a consumer‘s willingness to share data: Sensitivities of data types As the whole topic is centred around the • Awareness. Consumers with higher awareness Share of respondents willing to provide data to organisations (%) Sensitivity individual, it is vital to understand how con- of data uses require higher benefits in return for sumers really view and are willing to use their sharing their data. Specifically, those with higher- 100% · Age group and gender · Opinion on products 90% digital identity. Past research into the use of than-average awareness require 26% more Low · Name and e-mail personal data has mainly focussed on consumer benefit in return for sharing their data. The cur- 80% · Interests attitudes, asking people about their concerns rently low degree of awareness thus represents 70% Medium · Past purchases · Purchase plans and opinions. But as this report demonstrates, a real risk for organisations seeking to create 60% · Media usage individuals‘ privacy concerns and their behavi- value via digital identity – especially if they con- 50% · Location our don‘t always match up. To home in on actual duct practices that consumers would not know- 40% High · Social network posts · Health record behaviour and provide a quantitative perspective ingly approve of. 30% · Financial data on individuals‘ willingness to share personal 20% · Credit card data

data, we applied a value-based methodology for • Control. Having easy-to-use privacy controls 10% 8 this study‘s survey. and sharing options can substantially increase 0% €0 €5 €10 €15 €20 €25 €30 €35 €40 €45 €50 Compensation (€ per month) a consumer‘s willingness to share personal data. BCG‘s survey found that consumers who Source: BCG digital identity survey (n = 3,107, August 2012)

8 Conjoint methodology is a technique BCG frequently utilises in project work to identify how several dimensions of a product influence consumer decisions (for example, how design, brand, environmental impact and price affect the decision to buy an automobile).

13 | 14 EXECUTIVE SUMMARY Executive Summary EEXXECUECUTIVE TSUMMIVEAR SUMMY ARY

figures cannot be interpreted as the actual “value with the statement “I do not like it when a web- Another lens through which to look at sent greater value. Between these two poles, how- of the data.“ They rather serve as indications of the site asks me for the same information every time applications is the traceability and completeness ever, is space for compromise: solutions that ba- value consumers place on their personal informa- I open it.“ Control is important to consumers – of the personal data that is used. Individuals typ- lance privacy with value generation. For example, if tion. but so is convenience. Balancing these often ically prefer that businesses use digital identities control schemes guaranteed that digital identities conflicting aims will be tricky but critical. that are incomplete and not traceable – data that remained untraceable, more complete identities “The right to be forgotten“ has a small but may provide insight on usage or purchasing pat- could be used. Likewise, if data collection was lim- positive impact on sharing Yet perhaps the most impor- terns, for example, but can‘t be linked back to a ited to less complete information, more traceable tant takeaway from this study‘s re- particular person. Organisations, on the other digital identities could be employed. Just as organ- The study‘s research also shed some light search is this: Consumers want to hand, generally prefer to use complete and trace- isations need to understand the impact of their on how some of the privacy rights and controls share their data – if the benefits and able digital identities, as the granularity in the data collection and use methods, they must carefully in the proposed General Data Protection Regula- the privacy controls are right. and the ability to address specific consumers repre- assess the levels of traceability and complete- tion are viewed by consumers: Yet perhaps the most important takeaway Digital identity framework • The right to be forgotten. This has a small but from this study‘s research is this: Consumers Dark bubble representing po- Prevalent method tential future value creation consistently positive impact on the willingness want to share their data – if the benefits and the of data collection of use case of individuals to share private data. When re- privacy controls are right. Mined spondents were given the right to have their (profiled) data deleted at their request – as opposed to A framework for analysing acceptance of deletion being outside their control – they were digital identity applications Tracked Arrow showing how 10% to 18% more willing to share information. (contextual) use cases evolve Matrix colours indicating how This effect was most pronounced for people While the applications vary, a common sensitive consumers are to share their personal data who had not shared private information with framework can be used to evaluate every use case. Required Personal data application organisations before – or at least, thought they This framework focuses on the two key dimensions (structured) with bubble size representing hadn‘t. in how digital identity is leveraged: the method of current value

data collection and the manner in which it is used. Volunteered • C onsent. The form of consent consumers prefer Consumers consider these same two dimensions (unstructured) strongly depends on the type of data at issue. when viewing the trade-off between the value and Delivery of Enhancement Transfer to Transfer to Major use For some types of less sensitive data, for ex- cost of sharing information. requested of services third party third party of data ample, up to 69% of respondents considered service (anonymised) (traceable) opt-out, or even assumed consent, appropriate. Indeed, their willingness to share declines as For highly sensitive data, in contrast, more than the methods for collection and use move toward ness to be used in their applications and business might do the same thing and face tremendous 80% of respondents said an opt-in mechanism the outer boundaries, where consumers feel they models. In this context, it is interesting to observe consumer backlash. The difference in consumer was a must. have less control over the collection and use of that the level of acceptance of data-intense appli- acceptance can often be explained in terms of their data; for example, when organisations acquire cations can vary greatly between sectors, and even With regard to consent, the survey also data through tracking or data mining, or when they between organisations within the same sector. The level of acceptance of highlights a key dilemma digital identity poses. allow third parties to access it. Thus, consumer data-intense applications can vary Most respondents (82%) expressed the wish to sensitivities can also be plotted on the same frame- So while one company might run into no greatly between sectors, and even decide whether to allow data use in each in- work, and we can see at a glance both the oppor- problems collecting a user‘s location data and between organisations within the stance. However, a majority (63%) also agreed tunity and the risk any single application presents. transferring it to a third party, another company same sector.

15 | 16 EXECUTIVE SUMMARY Executive Summary EEXXECUECUTIVE TSUMMIVEAR SUMMY ARY

brand awareness and loyalty, the “cool“ factor of To help organisations successfully adopt the are the direct costs, as well, to both consumer and While safeguards are needed, a given product or application or the significant new paradigm, we offer the following recommenda- organisation: The level of security – not just tech- the regulations that ultimately re- consumer surplus associated in particular with tions: nical measures, but also policies and access control sult should be flexible enough that “freemium“ services. – needs to reflect the high value of personal data. consumers can make their own in- • E ngage customers for sustainable personal data formed choices about whether and Towards a new paradigm for digital identity usage. Organisations should involve and educate • B uild a data-driven organisation, not just a how they share data. consumers on personal data applications, as well data-driven IT department. Consumers aren‘t As detailed above, the opportunities digi- as on privacy management. And they should do the only ones who benefit from transparency. on certain data uses – which might impact a wide tal identity presents – when managed carefully – so in a carefully thought-out, well-structured way. Organisations do, too. They should know what range of applications beyond the targeted appli- are enormous. But so, too, are the lost opportunities, As we show in this report, three elements are es- data is available both inside and outside their cation. Ultimately, such prohibitions designed to should it not be handled in a thoughtful, balanced sential to ensure a sustainable flow of data: First, own walls, and all parts of the organisation protect individuals can actually hamper or delay way. BCG estimates that two-thirds of the potential the benefit a consumer receives has to exceed the should be encouraged to identify value creation innovations that would benefit them and can put value generation – or €440 billion in 2020 alone – is “cost“ of sharing the data. Second, there needs to opportunities, experiment with new applications domestic industries at a competitive disadvantage at risk if stakeholders fail to establish a trusted flow be transparency on how the data is used. While this and develop analytical talent. relative to their international competition. Bal- of personal data. might initially reduce sharing, it limits the risk of ance means, too, that sometimes a “highest“ level brand damage and helps to attract more informed Striking the right balance in policy and regulation of protection that stalls novel new applications is BCG estimates that two-thirds customers. Third, privacy controls should be avail- less desirable than a “high“ level that spurs them. of the potential value generation – able and easy to use. They will significantly increase Across Europe, policy makers are actively or €440 billion in 2020 alone – is at data sharing by individuals, likely offsetting any discussing proposed rules on privacy. While safe- The most successful policies will put the risk if stakeholders fail to establish a negative impact on sharing resulting from in- guards are needed, the regulations that ultimately consumer front and centre, but also enable exper- trusted flow of personal data. creased transparency. result should be flexible enough that consumers imentation in, and the development of, new uses can make their own informed choices about whether of personal data. They will be firm regarding prin- To unlock the full value, organisations need to • Take accountability for a trusted flow of data. and how they share data. Individuals should be ciples, but allow for flexibility in implementation make the benefits of digital identity applications very Every business and public agency that depends on able to generate value from their digital identity – and enforcement. They will give organisations a clear to consumers. Just as importantly, they need to digital identity applications should define guiding if they wish to do so. At the same time, they should certain amount of leeway, acknowledging that dif- embrace the new digital identity paradigm: It starts principles for how it works with personal data. be able to rest assured that their data isn‘t being ferent markets and regions have different histori- with responsibility. Then it adds transparency to limit These principles should be communicated clearly put to unintended or undesired purposes. cal, cultural and political legacies – and different, the risks presented by hidden practices that, once un- both within and outside the organisation to ensure constantly shifting attitudes towards privacy and covered, become the focal point of a public firestorm. unmistakable accountability and commitment. At Once again, balance is key. Privacy protec- new uses of personal data. A balanced approach, the same time, an organisational structure and tions shouldn‘t be so stringent that they discour- both in policy and in the applications themselves, Organisations need to communicate and be processes that hardwire compliance with those age innovation, nor should they be so weak that is essential if we are to realise the enormous bene- open about how exactly they acquire and use data. principles should be established. they enable misuse. It also requires sometimes fits digital identity has to offer. Finally, there must be user control. As evidenced by taking a step back from some of the emotion that the consumer survey data we present, individuals‘ • Increase data security in order to safeguard can imbue discussions on privacy. Politicised and preferences regarding privacy and data sharing digital identity. As the examples in this report hyped concerns about the privacy implications of differ – and differ widely. Giving them options and demonstrate, data breaches cause damage on new applications – concerns voiced recently in controls might reduce sharing for specific individ- multiple fronts: Consumers see their personal in- the case of Google‘s “Street View“ feature – can uals in specific situations, but overall it will likely formation compromised; organisations see trust morph into widespread outcry as it did in Germa- increase the sharing of personal data. erode and their brand and reputation suffer. There ny.9 That, in turn, could trigger blanket prohibitions

9 Spiegel Online, “Die Foto-Hysterie verdeckt echte Datenschutz-Probleme,“ 3 November 2010

17 | 18 Part 1

Part 1

part 1 Digital identity: A driver of growth, a consumer hazard or both?

P119 | 20 Part 1 a consumer hazard Part 1 orA dr both?iver of growth,

Digital identity – a driver of growth, a consumer hazard or both? Consumers benefit in a host of ways, too. third party‘s hands. If consumers hold back their They save time through automated transactions. personal information, the value potential of They save money when the efficiencies businesses digital identity will be held back as well. “Like any reap are passed down in the form of lower prices. currency, [personal data] has to be stable and it By now, we have all heard the analogies Organisations understand that there is And they place an especially large value on the has to be trustworthy,“ Reding went on to say. highlighting the importance – and value – of particular value to be found in digital identity. free online services that are supported, at least in For business leaders – and policy makers now data. It is the new gold. The “new raw material Online businesses, for example, regularly use a part, by the use of personal data – enough that discussing possible data privacy protections – of the 21st century,“ as Francis Maude, the UK consumer‘s purchasing history and preferences to consumers are willing to divulge pieces of that the challenge is to establish the trust that keeps Minister for the Cabinet Office, put it.1 Whatever recommend new products or to target their mar- data in return for using Facebook, Google and the currency flowing. the label, the basic message is the same: Data is keting efforts. But digital identity – and its bene- other services. As European Commissioner for an untapped resource of vast potential that only fits – is relevant to the entire economy. Indeed, its Justice Viviane Reding said in January 2012, For business leaders and policy needs to be unearthed and refined before the applications can help solve some of society‘s most when presenting the proposal for the new makers the challenge is to establish riches roll in. pressing problems. In health care, personalised General Data Protection Regulation, “Personal the trust that keeps the currency medicine, decision support systems and electronic data is in today‘s world the currency of the digi- flowing. If only. records that can easily be exchanged between tal market.“ doctors, hospitals and labs can improve quality of The discussion on digital identity has tend- For while the analogies make for catchy care – and even prolong lives. In the public sector, Currency, however, can be volatile, and ed to pit two ideas against each other: What is wordplay, they all have the same flaw: Data is data-driven initiatives to fight tax evasion and personal data is no exception. Data breaches and the potential economic value of personal data? still a big unknown. It is valuable, of course – but create citizen self-service portals (where trans- fears about data misuse are giving consumers How can consumers be protected from its mis- how will that value be realised? What types of actions can be performed online instead of in pause. They are concerned about their privacy use? What it needs to do, instead, is balance these applications and innovation can data enable – offices) can increase tax revenues and lower and what happens to their data once it is in a two goals, creating applications and policies that and what will they bring? What challenges lie in costs – easing the budget pressure almost all the path, and how can those obstacles be over- European governments now face. Global data explosion come? These are questions that are only starting 9 to be answered. And they are questions that be- 2015: 7.91 × 10 TB The value that can be derived Global data transaction come even more important – and more difficult from digital identity applications is volume in terabyte (TB) to answer – when the data in question is personal 1010 potentially enormous, and can be a information. significant force in spurring an other- 109 wise stagnant European economy. x 16,000 Individuals today enjoy a kind of dual Social media boom citizenship: We have our normal, everyday lives 108 in the physical world, but we also have our All told, the value that can be derived from Internet of things digital identity – all the bits and pieces of infor- digital identity applications is potentially enor- 107 5 mation about us that are readily, and increasing- mous, and can be a significant force in spurring 1990: 4.97 x 10 TB Online data transactions ly, available in digital form; data that is collected an otherwise stagnant European economy. BCG 106 and analysed to create a surprisingly accurate, estimates that at a 22% annual growth rate, the Digital service and media and ever-improving, picture of who we are, annual economic benefit for private- and public- 105 what we do, what we like (and what we dislike, sector organisations can reach €330 billion by 1990 1995 2000 2005 2010 2015

too). 2020. Source: comScore; Facebook; Sandvine: “Information White Paper,” 2011; Sandvine: “Digital Contents White Paper,” 2010; METI E-Commerce survey; IDC; consumer research; BCG analysis

1 The Guardian, 18 April 2012

21 | 22 Part 1 a consumer hazard Part 1 orA dr both?iver of growth,

spur innovation – and value – while safeguarding The “data explosion“ is being driven, made available for later use. Even trans- From a macroeconomic per- privacy. That means taking a close, critical look at specifically, by four global trends: actions that are ultimately abandoned – like spective, it is clear that the digital both the potential of digital identity and consum- the £1.02 billion worth of goods UK consum- data explosion can fuel economic er sensitivities. The insight that results will drive • Social media boom. These online services – ers put in their online shopping carts in growth. flexible solutions that will shape our future. including Facebook, Twitter, VK, LinkedIn 2011 but never purchased – are captured.4 and XING – let users not only interact with economic growth. It is no secret that growth has As the volume and variety of data each other, but also share news and infor- • Digital services and media. Faster, more been elusive for traditional business sectors in grows, so does its value mation, including demographical data, prevalent Internet access, low-priced Europe. Between 2008 and 2011, the financial likes and preferences, images, video clips storage media and the growth of mobile services sector saw revenues remain flat, while Gold may be a scarce natural resource, and even location data (where they are and devices have resulted in a vast increase in the the retail industry had a compound annual but digital data is the opposite: It is readily what they are doing there). While these amount of digital multimedia consumed growth rate (CAGR) of -1.6% and the traditional available and rapidly increasing at an exponen- services often come under scrutiny for their by users through services like YouTube or production sector a CAGR of -3.6%. Yet over tial rate. BCG expects that the global data trans- privacy practices, the best known continue Spotify. The shift to digital is exemplified the same period, highly data-intense industries action volume will grow at 45% annually through to see their ranks swell: Already, Facebook by Netflix, the popular US video service that thrived. The e-commerce industry enjoyed a 2015, to 7.91 × 109 terabytes, or approximately has 1 billion users worldwide, and by 2015, delivers titles by DVD and online streaming, CAGR of nearly 15%, and for Web 2.0 communi- 7 zettabytes. To put that in perspective, that is one quarter of the world‘s population will which saw a 72% increase in the number ties – the Facebooks and LinkedIns of the world – enough data to fill the hard discs of two standard be a member of a social network.2 of customers who utilise streaming only in the rate was just over 100%. Within the Internet laptops for every person on earth in 2015. 2011.5 sector, revenue has been closely tied to the use of • The Internet of things. In-product sensors personal data: On social networking sites, adver- The explosive growth of digital data is a that can “call home“ via the Internet and tising is targeted to users‘ profiles and posts; on phenomenon that both consumers and organi- relay usage data are becoming increasingly From a macroeconomic perspective, it e-commerce and digital media sites, personalised sations are very familiar with. A decade ago, a prevalent. Automobiles, home appliances is clear that the digital data explosion can fuel recommendations drive usage and sales. consumer in Europe could buy a 64-megabyte and energy meters are among the tradition- memory card – capable of storing three minutes of al product categories that have – or soon Sector growth high-definition video – for about €20. Today for will have – integrated links to the Internet. Revenue development 2008 - 2011 that price, she could buy a 32-gigabyte card Not including phones or PCs, we expect to in Europe (CAGR in %) capable of storing some 25 hours of HD content. see an additional 75 million devices with Traditional production -3.6% direct connections to the Internet in Europe But where the data is flowing from is the by 2015.3 Potentially hundreds of millions Retail -1.6% more interesting part of the story. The flood of more will be equipped with near-field Financial services 0.0% digital data is coming to a large extent from new communication chipsets that allow them and even unexpected places – sources that are to connect to the Internet through other Telco & media -1.5% particularly rich in the personal information that, devices. Public services / health collected, aggregated and analysed, extends our Web 2.0 communities 100% digital identities. • Online data transactions. Each time a transaction occurs online – a song purchase, E-commerce 15% an airline seat upgrade, a payment of an Online info / entertainment 22% electricity bill – data is collected, stored and

Source: Eurostat; BVH; comScore; BCG analysis

2 Source: Morgan Stanley; Strategy Analytics; eMarketer; UN “World Population Forecast“; BCG analysis 4 Research commissioned by Experian, published 19 March 2012 3 Source: ABI; Berg Insight; IDC; Yankee Group; Gartner; Analysys Mason; BCG analysis 5 Interpret, 7 March 2012

23 | 24 Part 1 a consumer hazard Part 1 orA dr both?iver of growth,

None of this has escaped the attention of • Self-service apps and portals. Telecommuni- organisations in the value chain can achieve present a considerable risk; one that has to be traditional businesses, who are starting to ramp cations providers and financial services com- overall cost savings of 10% to 20% by better understood, addressed and mitigated. up their own efforts to leverage digital identity. panies are moving more customer transactions leveraging personal data – no small benefit for Among the more prominent examples: from the offline labour-intense world to online an industry that has been seeing ever-growing Consumer research reveals a deep lack self-service portals (e.g., My Vodafone). And for cost increases, fuelled in large part by Europe‘s of trust • Personalised automobile systems. While many good reason: Online and mobile transactions are ageing population. car manufacturers now routinely incorporate far less expensive to support. Indeed, for banks, The concerns and anxieties of consumers re- digital sensors in their automobiles – typically they run just 5% of the cost of branch trans- Digital identity enables businesses to bet- garding the use of their personal data are mount- to monitor vehicle systems to enhance perfor- actions.7 ter understand what their customers like, and ing. Eurobarometer, a major Europe-wide study mance and troubleshoot problems – some are want, to serve them more efficiently and to conducted in 2010, revealed that 70% of consum- going a step further, integrating sensors with • Actual behaviour-based insurance policies. develop new and enhanced revenue streams. ers were concerned about how companies were personalised services. The result: new subscrip- The emergence of mobile devices capable of The benefits they achieve are passed down to using their personal information.10 BCG‘s own tion-based revenue streams, or rejuvenated transmitting location and usage data allows in- consumers, as well, in the form of lower prices, research study of more than 3,000 European con- offerings that are more compelling to buyers. surance companies to base their premiums on time savings or new and valuable services. This sumers, conducted for this report in August 2012, One initiative, BMW ConnectedDrive, inte- actual instead of predicted behaviour. Case in isn‘t happening on a small scale. We estimate not only confirmed the high level of concern, but grates in-car satellite navigation – an option point: GMAC Insurance has pioneered a pay-as- that digital identity applications can bring a demonstrated that, if anything, trust was eroding that, industry-wide, has lost ground to less ex- you-drive (PAYD) car insurance policy in coope- quantifiable annual benefit of approximately further. Nearly 90% of respondents considered at pensive after-market GPS units – with safety, ration with GM. Customers can let their OnStar €1 trillion in Europe by 2020. least one industry sector a threat to their privacy, entertainment and Internet services. Users can telematics system automatically transmit the 82% thought an independent entity should ensure search Google directly from their navigation unit mileage of their car and pay insurance premi- Achieving this potential is by no means a the data security compliance of organisations and or send a destination address to the system from ums only for the miles actually driven.8 given, however. Consumer sensitivities – heightened, 79% thought companies should be more trans- Google Maps. This initiative enabled BMW to as we will show, by an organisational approach parent about the use of personal data. take a product that was facing commoditisation As the euro crisis puts pressures to digital identity that largely ignores them – and differentiate it, keeping its sales healthy. on government budgets, early adop- ters are showing how data-driven Consumers‘ concerns • Big brands are Social marketing campaigns. value creation can help ease the strain. Share of respondents leveraging online social networks to improve their engagement – and boost their visibility – In the health care and public sectors, the Overwhelmingly, consumers are wary of how their data is being used with consumers. Farmers Insurance, the third potential benefits of digital identity applications largest general insurer in the United States, may be especially compelling. As the euro crisis Consider at least one industry sector as 88% teamed with Zygna, maker of the popular puts pressures on government budgets, early privacy threat Facebook-based game FarmVille, to promote adopters are showing how data-driven value its brand. An in-game Farmers airship provided creation can help ease the strain. In Italy – which Want an independent entity to control organi- 82% sations‘ data security players with insurance to protect their FarmVille loses an estimated €120 billion each year to tax compliance crops; clicking on the airship took players to the evasion – a system called “Redditometro“ analyses Think companies should company‘s Facebook fan page. Before this cam- some 100 indicators of a citizen‘s lifestyle to be more transparent about 79% personal data use paign, Farmers had 6,500 Facebook fans. Ulti- optimise tax audits. In 2011 alone, €11.5 billion 9 mately, it had 2.4 million – all potential advo- was recovered. In the health care sector, BCG 0% 100% cates for spreading the word about the company.6 estimates that insurers, providers and other

10 6 Neal Schaffer, October 2010; Farmers Insurance Special Eurobarometer 359, field work conducted 2010, published 2011 7 BCG analysis 8 GMAC Insurance and General Motors websites 9 The International Herald Tribune, 7 February 2012 25 | 26 Part 1 a consumer hazard Part 1 orA dr both?iver of growth,

Consumer concerns fall into two broad breaches by hackers or disgruntled em- Many of these data breaches could have didates and universities increasingly use them categories: ployees have compromised personal infor- been avoided had organisations taken the proper to screen applicants, individuals want to know mation stored on corporate servers (names, precautions to safeguard information, such as that their information won‘t be used against • Actual damage: These are the worries that mailing addresses, national insurance using strong encryption to encode user account them. They worry that new technologies, like stem from the real-world events – and real- numbers, credit card information, e-mail data, sufficiently restricting employee access to face recognition, will give them less anonymity. world damage – that have increasingly addresses and so on) and led to credit data and securing data backups. Their failure to They wonder whether companies and govern- filled the headlines in recent years. Data card fraud, identity theft, spam and other take readily available steps only heightens con- ment agencies are creating detailed profiles illegal or undesirable consequences. sumer mistrust. It also causes brand and reputation on them, and how they might be using them. damage when the breach occurs. These, perhaps, are the hardest worries to allay, High-profile data since they can‘t be solved simply by beefing up breaches since 2011 That last point is important, for the damages security. incurred when data is breached tend to be mutu- Up to 40 million employee records could have been stolen al: both consumers and organisations are harmed. New and emerging technologies can com- RSARSA March 2011 through spear-phishing attacks Sony‘s PlayStation Network breach, for example, plicate the picture further, as many can be seen as

E-mail addresses and names of millions of customers, several resulted in costs the company put at ¥14 bil- both a privacy threat and a privacy enhancement. epsilon March 2011 financial institutions (e.g., CitiGroup) and the non-profit organi- sation College Board were exposed lion by the end of its 2011 fiscal year – roughly Biometrics can boost security by requiring a non- €128 million.11 Outside experts estimate total costs alterable, non-transferable identifying character- A total of 77 million PlayStation Network accounts were hacked, 12 Sony April 2011 exposing names, e-mail addresses and credit card data. Sony lost including brand damage at up to $4.6 billion. istic (a fingerprint, a retina scan, a facial pattern) millions during the long downtime of the network for access. But it can also potentially be used to • “What If“ fears and worries: These are the erode anonymity; for example, face recognition The UK online travel provider‘s customer database was hacked, with Travelodge June 2011 its clients‘ e-mail addresses and names exposed subjective concerns; the worries about “Big software that automatically identifies and tags Brother“ and other unscrupulous misuse. As social network users in photographs uploaded to Malware attack on ESTsoft‘s servers led to exposure of the names, ESTsoft July 2011 user IDs, hashed passwords, birthdates, genders, telephone numbers, employers take to social media sites to vet can- the site. and street and e-mail addresses of 35 million Koreans

Tricare / A SAIC backup tape was stolen from a Tricare employee, containing Share of concern, September 2011 medical and financial information of 5.1 million individuals; a $4.9 SAIC billion lawsuit followed awareness and control

Attackers hacked a cloud-based NASDAQ system, which enabled NASDAQ October 2011 them to monitor boardroom-level communications of more than Individuals‘ Concern Awareness Control 10,000 executives privacy stage

During routine maintenance procedures, customers‘ phone numbers Telefonica O January 2012 were logged to websites when they were accessing the service via 2 3G or WAP, exposing the data to website publishers Regard at least one sector Aware that > 50% of sectors Ever done > 3/4 of activities to Files containing 6.4 million passwords were found on underground as privacy threat collect personal data manage their privacy LinkedIn June 2012 hacker websites by security specialists and are expected to belong to LinkedIn members Share of EU Hackers targeted Philips‘ servers and obtained more than 200,000 online population 70 – 90% 15 - 30% 0 - 10% Philips August 2012 e-mail addresses, along with customer information such as names, postal addresses, birthdays, passwords and phone numbers

Source: Press; company websites Source: BCG digital identity survey (n = 3,107, August 2012)

11 Sony‘s script of revisions to reporting of fiscal year 2011 published on 23 May 2011 12 Ponemon Institute 2011

27 | 28 Part 1 a consumer hazard Part 1 orA dr both?iver of growth,

their browser – just 10% had ever done six or when their customers do find out what is going more of these things. on, they might do something about it.

Consumers are largely in the dark about Customers aren‘t likely to stay how, when and where their personal data is used. in the dark about data use forever, Transparency is often missing or insufficient – and companies run a real risk that as evidenced by the 79% of survey respondents when their customers do find out who said it should be improved. what is going on, they might do something about it. When customers of the American retail chain Target Corporation signed up for the company‘s This report will show that the risks compa- REDcard shopping card, they expected to gain nies face still exist and are significant. But it will benefits like special discounts. What they likely also show that they can be reduced – to the benefit didn‘t expect – and what Target did not tell them – of consumers and organisations alike. This insight IPSUM was that their purchase history was also being comes from the survey commissioned specifically used to calculate a “pregnancy prediction score“ for this report (and detailed in part 2), in which for them (and a very accurate one, too, as expect- respondents from Germany, the Netherlands ant mothers tend to suddenly start buying cer- and Poland were presented with scenarios and tain products, like supplements and additive-free questions that narrowed in on their attitudes – foods). For Target and other retailers, pregnant and their actual behaviour – regarding the sharing women are a valuable subset of buyers, since of data. What made this survey unique was that it they tend to purchase more and with lower price didn‘t focus on the potential perils of sharing in- sensitivity. But after Target started using those formation so much as how the decision to share scores to develop lists of possibly pregnant cus- is made – the factors that influence it, and how tomers and send them coupons for baby-related consumers weigh them. This goes to the heart products, some of the recipients reacted badly – of what data-dependent organisations need to in one incident, recounted in The New York Times know as they try to strike a balance between Magazine in 2012, an angry father confronted value and privacy. a Target manager after the chain guessed his While most consumers are concerned BCG‘s survey revealed that only 30% of teenage daughter‘s pregnancy before he did.13 about personal data applications, far fewer are respondents had a relatively comprehensive Balanced solutions protect privacy, generate aware of just how their information is being understanding of which sectors were collecting For the retailer, the backlash was relatively value and create competitive advantage acquired and leveraged. and using their data. Fewer still are able to exert minor. It simply tweaked the mailings so coupons some control over that collection and use. When for random products were included, too – making Privacy and the way organisations treat While most consumers are survey participants were asked to note which its targeted marketing appear not quite so tar- consumers‘ data is becoming an increasingly concerned about personal data ap- of eight common privacy protection activities geted. But the incident demonstrates two crucial important competitive factor. Anyone who plications, far fewer are aware of they had ever performed – actions like changing facts: Customers aren‘t likely to stay in the dark doubts this need only look at the desktop and just how their information is being privacy settings in a social network, surfing the about data use forever (especially with the growth mobile Web browsing market. When Microsoft acquired and leveraged. Web in “private mode“ or disabling cookies in of social media), and companies run a real risk that announced that its next version of Internet

13 New York Times Magazine, 16 February 2012

29 | 30 Part 1 a consumer hazard Part 1 orA dr both?iver of growth,

Analysed sectors This study aims to reshape the discussion In part 4, the report outlines the six most impor- by providing political and business leaders with a tant trends of digital identity, the “megatrends“ Manufacturing industries 1 Traditional production quantitative angle, empirical evidence on that are relevant for every organisation, in both consumers‘ actual data-sharing decision making the public and the private sector, looking to iden- 2 Retail and a holistic perspective. tify opportunities for value generation. Services industries 3 Financial services This study aims to reshape Finally, part 5 presents a new paradigm for re- the discussion by providing political sponsibly unlocking the value of digital identity – 4 Telco & media and business leaders with a quan- in a sustainable, consumer-centred way – and Public sector 5 Public services / health titative angle, empirical evi- offers guiding principles for key stakeholders. dence on consumers‘ actual data- 6 Web 2.0 communities sharing decision making and a holistic perspective. Internet industry 7 E-commerce In part 2, we present the results of a 8 Online info / entertainment consumer research study that went beyond traditional survey questions on privacy concerns to look at actual behaviour. Drilling down on the factors that spur, and hinder, the sharing of Explorer would enable a “do not track“ feature alike are worried about data misuse by apps, personal data, it quantifies consumers‘ decision by default – so Websites will know that a user particularly as the number of available apps making, as well as their sensitivities. does not tolerate cookies or any other tracking continues to grow rapidly. Currently, neither features – the move was clearly intended to lure Apple nor Google provide much in the way Part 3 quantifies, for the first time, the customers away from competing browsers (like of user control. How they address the issue – current and potential economic value of digital Firefox, Opera and Chrome) that are not expect- whether they give users tools to manage their digi- identity for organisations and consumers. As ed to follow suit. tal identities within the app ecosystem, or take the opportunities and challenges vary – often no action – will have important implications in greatly –by industry and area, it takes a sector- Another area where privacy and personal terms of both privacy and competitiveness. by-sector approach, laying out the economic data management may likely be an important value, most important current applications and competitive dimension is the mobile operat- Digital identity can change balance sheets – cutting-edge case studies for each. The result is ing system space. Worldwide, more than and change our future. But solutions need to put a comprehensive, detailed overview of the value 850 million smartphones will run either the customer front and centre. Only when there generation potential for the European economy Apple‘s iOS or Google‘s Android by the is trust that organisations are handling infor- as a whole. end of 2012.14 Both platforms enable mation responsibly and providing sufficient third-party apps to track highly sensitive individual benefit, will data be shared in a sustain- personal data, like location information, once a able way. user agrees to terms of use upon installation with just one tap. This model has already raised privacy concerns; consumers and policy makers

14 Source: Gartner (July 2012); BCG analysis

31 | 32 Part 2

part 2 A definition of digital identity and how consumers really see it

P233 | 34 aA ndefinid howtion of dconsumeigital identity rs really see it Part 2

A definition of digital identity and how consumers really see it A definition of digital identity the data record – such as a customer ID num- ber or name and date of birth – could be used Data about us – what we like, what we do, to trace it back to a specific individual. who we are – is scattered all across the digital land- Digital identity: Putting the pieces of our lives advances in the ability to aggregate, process and scape. It is distributed on different systems, on diffe- Yet, as we will show, it is not always easy to together gain insight from it, is not only making our digital rent storage devices, on social networks, on corporate determine what information can, and cannot, identities more detailed, but also more accurate, servers, in government databases. No matter where be traced back to someone. First, let‘s look at By now, we are used to divulging little more useful and ever more valuable. It is a phe- it is, this information shares one important characte- the different types of data from the least to bits about ourselves as we navigate our physical nomenon that is sparking innovation, anticipation ristic: It can be traced back to a specific person. the most traceable: and digital worlds: credit card information for a and not just a little anxiety. purchase; an e-mail address to receive discounts; Digital identity is the sum of all digitally • Anonymous data. Data that is collected a social security number on a medical form. By It is also a phenomenon that is often mis- available data about an individual, irrespective of without any identifier and was never linked themselves, these pieces of personal data tell only understood, or not really understood at all. To its degree of validity, its form or its accessibility. It to an individual. An example of anonymous part of our story; together, they form a richer, full- that end, we define the key concepts involved can include any – and often all – of the following: data would be a questionnaire returned by er picture. in digital identity, share findings from a unique mail with no name or return address. consumer research study and introduce an inno- • Inherent characteristics. Where does an individ- That picture is becoming increasingly com- vative framework for identifying the opportunities – ual come from, and who is he or she? Date of • Anonymised data. Previously identifiable plete. The tremendous growth in the volume of and risks – presented by today‘s most important birth, gender and nationality are examples of this information that has been de-identified: personal data now shared or collected, along with digital identity applications. type of information. Anything that could link the information back to a specific individual – such as an • Acquired characteristics. What is an individual‘s ID code, a credit card number, even a serial Digital identity definition story; their history? Here, information such as number for a mobile phone – has been re- Individual address, medical record and purchase history are moved and a third party possessing the in- preferences relevant. formation would be unable to reconstruct · Favorite brands What does · Taste in music she/he like? it (if done properly, see below). · Interests · ... • Individual preferences. What does an individu- al like? Data types here would include interests, • Pseudonymous data. Data records that hobbies and favourite bands and television shows. contain no clear ID, but an explicit identifier Acquired Digital identity that can be used to link the data directly to attributes Sum of all digitally a specific individual. Examples of such an · Address What did available data about identifier are a taxpayer ID or a customer · Medical record she/he do? an individual, irre- · Purchase history spective of its degree The blurring lines of personal and anony- account number. Pseudonymous data thus · ... of validity, its form mous data provides protection of personal data to the or its accessibility extent that the translation of the identifier The definition of personal data is fairly into a clear ID is protected. Inherent characteristics straightforward: information relating to an · Date of birth Where does identified or identifiable individual. In the • Clear personal data. Information that is · Gender she/he come · Nationality latter instance, the person would not be ex- most easily traceable to an individual, as from? · ... plicitly identified, but one or more aspects of the data record contains a clear ID.

35 | 36 aA ndefinid howtion of dconsumeigital identity rs really see it Part 2

Organisations seeking to expand the use – ice Netflix (with random numbers replacing pants about their opinions regarding how their The survey was split into two parts. The first and value – of personal information while user names). By comparing the rankings and personal information might be used. While these followed a conventional path, with ques- mitigating privacy concerns often turn to a timestamps with ratings that had been post- are important, words don‘t always equal action, tions that explored attitudes and concerns time-honoured solution: “anonymising“ the ed on another site, the Internet Movie Data- and what is essential to know is consumers‘ ac- regarding privacy (other questions obtained data. The idea is simple: By deleting or ob- base (IMDb), they were able to link some tual behaviour. When presented with a choice to demographic data). The second part was de- scuring the codes and identifiers that link back of the Netflix ratings with rankings that share – or not share – personal data, how do they signed around the conjoint techniques. Every to a specific individual, any subsequent party had been left, under a user‘s own name, on actually make their decision? participant was presented with 12 sets of using the data would have no way to know IMDb. This enabled them to identify the users “deals,“ each consisting of three scenarios in who it related to. Anonymised data is there- behind those supposedly anonymous Netflix To better understand that decision making, which payment was offered in exchange for fore key for many scientific and commercial rankings.1 the study made use of conjoint methodology – data. The scenarios involved different types of applications of digital identity by third parties. a technique BCG frequently utilises in project data, to be put to different uses, collected in After all, why should an individual mind if an What this means is that organisations have work to identify how various dimensions of a different ways, by organisations from differ- organisation is sharing data that in no way to take special care and regularly revisit their product (such as the design, brand, cost and ent sectors. For each set, respondents picked identifies them? processes when they want to use anonymised environmental impact of an automobile) influence the scenario, or exchange, they viewed most data – and determine whether the informa- purchasing decisions. In the same way, we used favourably. But what is anonymous one day isn‘t always tion is indeed truly untraceable. With each conjoint methodology to explore how different so anonymous the next. As the volume and passing day, the odds are better that it‘s not. dimensions of data sharing (such as the type of Before carrying out the survey, we conducted variety of data grows – along with the ability data being collected, the sector doing the collect- numerous qualitative interviews to ensure to analyse it – links can sometimes be made ing and the method of collection) impact the that key names and descriptions were under- even in the absence of explicit identifiers. Understanding how consumers view decision to share personal information. This is stood by respondents from different regions With enough effort – and occasionally not their digital identity the first time that such techniques were used in and backgrounds. Accordingly, we avoided much effort at all – the data can be traced such a comprehensive way to analyse the privacy technical terminology in the survey. For in- back to a single individual. Sometimes this is Organisations knowthat when they design valuations of consumers. stance, instead of asking consumers whether done by comparing or piecing together differ- a product or create a service, understanding con- they thought “opt-in“ was necessary, we ent bits of data, a sort of high-tech deductive sumer attitudes is a critical success factor. The asked if they thought an organisation seeking reasoning. Other times, a recently discovered same applies for digital identity applications. To The survey to use their data ought to get their agreement technology makes it possible to unlock pre- leverage them successfully and realise the opti- beforehand. Using a standard test to gauge viously hidden or meaningless identifying mal benefits and value they can bring, one has A total of 3,107 Europeans were surveyed in the accuracy of conjoint surveys – the root- details. to understand how consumers view their digital the summer of 2012. The group was comprised likelihood analysis (RLH) – we found the re- identity. What are their sensitivities? What factors of 1,026 individuals from the Netherlands, sulting significance of this survey to be good.2 Consumers would likely be surprised – and do they weigh when making the decision whether 1,041 from Germany and 1,040 from Poland; concerned – by the speed and accuracy with or not to share personal data? And more to the these regions were chosen because they offer which seemingly unidentifiable information point, how do they weigh those factors? a cross-section of Europe including countries In the end, the study enabled us to quan- can become identifiable. In one well-known with a reputation for high privacy valuation tify and visualise the willingness of individuals example, two researchers at the University To help answer these questions, we commis- and a former Warsaw Pact country. Partici- to share their personal data with organisations. of Texas at Austin de-anonymised subscrib- sioned a consumer research study – one very differ- pants were selected as a representative sample The results – a key component of this report – er movie ratings that had been anonymised ent from past research on the use of personal data. according to the demographic factors of age, will hopefully contribute valuable empirical and published by the U.S. online video serv- Traditionally, surveys on this topic have focused gender and region within their country. evidence of how consumers view their digital mainly on consumer attitudes. They ask partici- identity.

1 Narayanan, Shmatikov: “Robust De-anonymization of Large Sparse Datasets,“ 2008 2 Root-likelihood (RLH) is 0.64. This implies that predictions based on the conjoint model are twice as good as random predictions.

37 | 38 aA ndefinid howtion of dconsumeigital identity rs really see it Part 2

Few consumers know how their data is not a strong indicator of their actual behav- showing that the disconnect between individuals‘ and viral communications, this is a less promis- used – fewer can control it iour when it comes to sharing data. When we concern and behaviour applies to the broader ar- ing proposition than ever. looked at respondents at opposite ends of the ray of data-sharing scenarios. The survey revealed that while most con- anxiety spectrum – one group that expressed Significantly, this correlation between sumers are concerned about their digital identi- concern about the use of personal data by one • Awareness. The survey also revealed that con- awareness and willingness to share varies by sector, ty, fewer are specifically aware of how it is being or no industry sector, and another group that sumers are largely unaware of what organisa- as shown in Exhibit Impact of awareness on data used, and fewer still are able to manage it through was concerned by all or nearly all of the sectors, tions are doing with their personal data. Only sharing. In retail, for example, the link is particu- privacy controls and other means. Our research we found that their actual privacy valuations 30% of respondents had a relatively compre- larly strong: Consumers with high awareness are also shed light on how these three dimensions – (determined from the deal scenarios they chose, hensive understanding of which sectors4 were 53% less likely to share. For banks, however, high- concern, awareness and control – impact the wil- and reflecting their willingness to share) did not collecting and using their data. This awareness awareness customers are just 13% less likely to lingness to share data. Taking a close look at each, differ. In this respect, privacy is similar to cli- gap, it turns out, poses a real risk to organisa- share, and for public agencies, the figure is only then, and understanding the role it plays in digital mate change: Stated concerns do not necessari- tions, as the study found that consumers with 9%. It‘s worth noting that banks and public agen- identity, is essential for developing applications ly translate into changed behaviour. higher awareness of data collection and use are cies have already a relatively high level of con- that aren‘t just innovative, but also successful. less willing to share data. Specifically, those with sumer awareness of collection and use practices. The relationship between expressed concern higher-than-average awareness were, over- More than 80% of survey participants, for exam- • C oncern. To no surprise, given previous research and privacy valuation is shown below, in Exhibit all, 26% less willing to divulge personal infor- ple, knew that banks were collecting personal data and the public discourse, we found a high level Expressed level of concern vs. valuation of personal mation. A cynic might be tempted to think that (ranking first among the sectors). Retail, on the of concern regarding the way organisations may data. This “weak link“ is not unprecedented: Pre- ignorance is bliss – provided that ignorance can other hand, rates the second lowest level of aware- be using personal data: 88% of respondents be- vious research had found that an individual‘s stat- be maintained – or, to put it another way, that ness, at just 32%. One possible interpretation: The lieved that at least one industry is a threat to ed concern about sharing private information was it is all right if an organisation keeps its use of more unexpected the use of personal data by an their privacy. But the research also indicated a poor indicator of whether or not they would join personal data secret as long as it is very good at organisation is, the more pronounced is the effect that the level of concern consumers express is Facebook.3 Our findings take this a step further – keeping it secret. But in this age of social media of awareness on willingness to share.

Expressed level of concern vs. Impact of awareness valuation of personal data Impact of data collection awareness per sector on willingness to share per sector on data sharing Value of personal data High -30% -53% -34% = Expressed H igh concern -32%-20% -25% -9% -23% Value of -20% -25% -32% personal -13% data -24% L ow Social Online Online Public Banks Mobile Health Loyalty Retailers Cable Car manu- net- search shops agencies phone insurances cards network facturers works engines operators providers Low Lowest Expressed concern Highest Low awareness of sector data collection High awareness of sector data collection

Source: BCG digital identity survey (n = 3,107, August 2012) Source: BCG digital identity survey (n = 3,107, August 2012)

3 Acquisti, Gross: “Imagined Communities: Awareness, Information Sharing, Privacy on the Facebook,“ 2006 4 While our research study focused on eight sectors, the survey included eleven types of organisations. We did this in order to capture the granular differences between different types of organisations within some sectors. For example, the public service and health sector is comprised of both public agencies and health insurers, who don’t always use data, or impact an individual’s views and behaviour, in the same way. 39 | 40 aA ndefinid howtion of dconsumeigital identity rs really see it Part 2

For a telling example of the low overall • Control. Privacy controls also have an impact on Exhibit Impact of control on data sharing selves are different. We see these differing per- awareness of data collection and use practices, a consumer’s willingness to share, the research shows how willingness to share varied with spectives play out in real life all the time. For exam- one need only look at Internet search engines. While found – but in a positive way, if they are imple- participants’ proficiency in managing their pri- ple, some frail and elderly individuals find it 75% of respondents said they had shared personal mented well. The survey found that consumers vacy. Respondents were split into nine groups, perfectly appropriate to have a video camera in- information with organisations, only a quarter of who are able to manage and protect their pri- corresponding to how many of the most com- stalled in their bathrooms, as a monitoring system these participants said they had shared such data vacy (via common actions such as changing pri- mon privacy protection activities they had ever to sound the alarm if they fall and get hurt. Very with search engines. Yet at the same time, the vacy settings in a social network, disabling cook- performed. Participants with the lowest assumed few young people, however, would want that overwhelming majority of respondents told us that ies in their browser, surfing the Web in “private” ability to manage their privacy (those who had same surveillance in their own bathroom. they used search engines. As these sites almost al- mode, or opting out of certain data uses) are up never performed any privacy protection activi- ways store search history – which often includes to 52% more willing to share information than ties) were also the least willing to share data. As While accessible, intuitive privacy features highly sensitive data – this was vivid evidence of those who aren’t. Ease of use is critical, espe- user proficiency increased, willingness to share have a positive impact on consumers’ willing- the frequent disconnect between assumed and cially for low- and medium-sensitivity data. For increased – though among the most proficient ness to share personal data, the effect varies with actual knowledge when it comes to what organi- sharing age and gender information, for exam- users, the correlation was not as strong (though the type of information at issue. As Exhibit Im- sations do with personal information. ple, individuals who had a relatively easy time they still had well above-average willingness to pact of privacy protection ease of use shows, managing their privacy were 37% more willing share). the impact is great-est for data of low sensitiv- The takeaway for organisations seeking to share than those who had a relatively difficult ity (such as age, gender and interests) and me- to create value through digital identity is clear: time. (The correlation between ease of use and It is significant, too, that few respondents dium sensitivity (past purchases and location, Low awareness can ultimately work against sharing does not hold for highly sensitive infor- were able to exert much control over the col- among other data types). But, as might be them – especially, if they conduct practices mation, like medical and financial data.) lection and use of their personal data. Just expected, the effect is neutral for highly that consumers would not knowingly approve of. 10% had ever done six or more of the common sensitive types of data, like health and financial privacy-protecting activities. And only around information. These are data types people would half had done more than three. rather not share – no matter how easy it is to manage. Impact of control These findings lead to one crucial over- on data sharing Consumers' willingness to share data arching conclusion: With easier-to-use priva- Many factors impact consumer decision cy protection features, and privacy education, making High +52% organisations might be able to significantly boost consumers‘ willingness to share data. The survey’s conjoint methodology let us 63% 62% 61% 60% explore how different factors influence the 56% 56% Why are consumers more willing to share decision to share, or not share, data. What we 55% when they have better controls? One idea is that found was that consumers not only give different 49% individuals are more willing to take risks when weight to different factors – the data type, for 41% they feel more in charge. Yet while this might be example, is more than twice as important as the Low a contributing factor in their decision process, collection method – but that even within a single we would argue that choices and control simp- dimension they make clear distinctions: For Lowest Highest ly help individuals adapt their sharing to their instance, consumers are on average nearly 30% specific preferences. As we will show, preferences more willing to share with cable operators and Consumers' privacy protection proficiency and sensitivities for personal data sharing vary e-commerce companies than they are with Web Note: The privacy protection proficiency axis shows how many of the eight wide-ly among people – just as people them- 2.0 communities. privacy-protecting activities respondents had ever applied, from lowest (none) to highest (all eight). Source: BCG digital identity survey (n = 3,107, August 2012)

41 | 42 aA ndefinid howtion of dconsumeigital identity rs really see it Part 2

Impact of privacy In our research we explored how six major di- purchases less so; while interests, e-mail address protection ease of use Willingess to share per data type mensions influence the consumer decisions to share and opinions on products are considered most Age group, gender +37% personal data with organisations. In descending or- readily shared with organisations. Opinion on products Low +36% der of importance (to their decision making) they are: sensitivity Name and e-mail +31% Exhibit Sensitivities of data types shows Interests +45% • Data type. Not all data is considered equal how consumers differentiate between data types.

Past purchases +38% when it comes to privacy. We found that con- While low-sensitivity information (like age and

Purchase plans +28% sumers consider some types to be much more gender) was shared by more than 40% of sur- Medium sensitive than others – and, accordingly, they are vey respondents even without receiving a direct sensitivity Media usage +31% less willing to share it. There is more hesitancy benefit, most participants were unwilling to part Location +39% to share credit card information, for example, with highly sensitive data (such as a health re- Social network posts +4% than one’s age group and gender. Financial data, cord) even for large rewards. Notably, also social Health info -2% health records, and social network posts are also network posts belong to the category of highly High sensitivity Financial data +3% considered very private; location data and past sensitive personal data. Respondents‘ experience of ease of use:

Credit card data -7% Rather difficult Rather easy Sensitivities

Low of data types High Share of respondents willing to provide data to organisations (%) Sensitivity Source: BCG digital identity survey (n = 3,107, August 2012) 100% · Age group and gender · Opinion on products 90% Low · Name and e-mail Dimensions of consumer 80% · Interests decision to share Dimension Weight Margin of impact for decision to share data Lower willingness to share Higher 70% Medium · Past purchases personal data Ø · Purchase plans 60% · Media usage Right to 10% Not defined At your request 50% · Location be forgotten 40% High · Social network posts 30% · Health record · Financial data Collection 13% Tracked Volunteered 20% · Credit card data method 10%

0% €0 €5 €10 €15 €20 €25 €30 € 35 €40 €45 €50 Compensation (€ per month) Use of 13% Transfer to third party Delivery of the data (traceable) requested service Source: BCG digital identity survey (n = 3,107, August 2012)

Sector 16% Social network Online shop • Compensation. By increasing compensation5, which, as seen above, is often withheld even organisations can often – but not always – when the compensation is high. While we test- Age Credit offset other factors that may be hindering the ed cash compensation only in our experiment Data type 31% group card and data gender sharing of private information. The offset works to achieve comparable results, for actual digital particularly well for low- and medium-sensitivi- identity applications “compensation“ can come Characteristics of dimension with Note: Based on primary consumer research with representative sample of ty data; less so for highly sensitive information, in the form of many non-monetary benefits. n = 3,107 (NED: 1,026; GER: 1,041; PL: 1,040) Lowest willingness to share Others Source: BCG digital identity survey (n = 3,107, August 2012) Highest willingness to share 5 It should be noted that the research method we applied provides very meaningful data on the relative differences, but the absolute monetary figures cannot be interpreted as the actual “value of the data.“ Rather they serve as indications of the value consumers place on their personal information.

43 | 44 aA ndefinid howtion of dconsumeigital identity rs really see it Part 2

• Sector. Consumer willingness to share person- sumers, while only 31% disapproved of organi- tion can be purged from organisations‘ reposi- nificant for users with low overall willingness al data differs from sector to sector. The study sations collecting data in order to deliver a prod- tories (whether this occurs manually on request to share data. found, for example, consumers are most willing uct or service. or automatically after a specified time frame). The to share personal data – about 13% more than survey found that the impact of this right va- As shown in Exhibit Competitive position the average of sectors – with e-commerce com- • Collection method. Consumers, the survey ries depending on the type of data at issue, of industries, the type of organisation seek- panies. On the other end of the spectrum, they found, are more willing to share when the in- but its presence has a positive impact on the ing to collect and use personal data influences demand far higher compensation from social formation is actively provided by them either decision to share for each one. The survey also people‘s willingness to share. Social networking networking sites than from any other sector. voluntarily (e.g., date of birth provided in a form, found that the right to be forgotten is most sig- sites like Facebook have to provide consumers despite it is not required) or because it is re- • Use. How the data will be used – whether it quired for delivering a service (like address for Consumers' willingness will be used to provide the requested service, shipping a package) than when it is acquired to share data required to enhance the delivery of another service or be with them being passive via tracking (such as Share of respondents willing to provide data for digital wallet per sector Online shops transferred to a third party – also impacts the location data continuously transmitted by a 70% consumer’s decision to share or not. The valu- consumer’s smartphone) or data mining (infor- Search engines Banks ation is congruent in this instance with voiced mation derived from analysing data sets). 60% Mobile phone operators opinions: Fully 70% of survey respondents dis- Social networks approved of organisations allowing third parties • The right to be forgotten. Another factor con- 50% to use data that could be traced back to con- sumers weigh is whether their personal informa-

+ 9% 40% + 28% Competitive position

of industries 30%

Perception of industries as privacy threat Recquired compensation of industries

€ Lower Compensation Higher 20% Compensation ( per month) € € € € € € € € € € € Share of mentions as top concern for privacy in % Ø 0 10 20 30 40 50 60 70 80 90 100

Social networking sites 48 +13.4%

Online search engines 32 +5.3% Source: BCG digital identity survey (n = 3,107, August 2012)

Online shops 19 -12.6% with 13% more benefit than other sectors, on ave- most willing to share the data with them (as Public agencies 19 -0.6% rage, need to in order to get the same data. depicted in the exhibit above). Ranking just Banks 18 +5.0% below the e-commerce sector were search Mobile phone operators 16 +3.9% That consumers are more apt to share with engines, banks, and mobile phone operators, all

Health insurances 14 +3.2% some sectors than others can have significant at a very similar level. Social networks, on the ramifications for the development of some high- other hand, brought up the rear. One could argue Loyalty cards 12 +4.8% profile – and potentially high-value – digital that the e-commerce sector’s high ranking re- Retailers 10 -3.9% identity applications. Case in point: digital wallets. flects the experience consumers already have 8 -11.3% Cable network operators Our research study revealed that e-commerce with it sharing data relevant for payment and off- Car manufacturers 4 -7.1% companies have a significant competitive advan- ers based on users‘ purchase history. But what- tage in acquiring the personal data needed for ever the reason, this is a strong indication that Source: BCG digital identity survey (n = 3,107, August 2012) digital wallet applications; consumers were e-commerce companies are in a prime position

45 | 46 aA ndefinid howtion of dconsumeigital identity rs really see it Part 2

to get consumers‘ approval for the personal data The right to be forgotten. As lined out Consumers‘ view on consent usage required to provide digital wallet appli- above, this has a small but consistently posi- Share of respondents considering method of consent as appropriate per data type

cations. tive impact on the willingness of individuals to 39% 31% 61% 40% 52% 53% 49% 56% 62% 80% 83% 83% share private data. When respondents were giv- Consumer perspectives on the proposed en the right to have their data deleted at their EU privacy regulation request – as opposed to deletion being outside

their control – they were 10 to 18% more willing 37% 43% 42% The study’s research also shed light on to share information (Exhibit Impact of privacy 39% 36% 36% how consumers view some of the privacy rights protection ease of use shows how this broke down 33% 31% 31% and controls in the EU’s proposed General Data per data type). This effect was most pronounced 26% Protection Regulation: for people who had not shared private informa- 24% 18% 16% 14% 14% tion with organisations before – or at least, didn‘t 12% 12% 12% 10% 7% 6% 4% 3% 3% think they had. Age Opinion Name Interests Past Purchase Media Location Social Health Financial Credit group on prod- and purchases plans usage network info data card and ucts e-mail posts data gender Low sensitivity Medium sensitivity High sensitivity Impact of the right to be forgotten Note: Numbers may not sum to total due to rounding. No agreement Opt-out Opt-in Source: BCG digital identity survey (n = 3,107, August 2012) per data type Willingess to share per data type Consent. The form of consent consumers For some kinds of less sensitive data, for example, Age group, gender -16% prefer strongly depends on the type of data at up to 69% of respondents considered opt-out, or Low Opinion on products -18% issue (see Exhibit Consumers‘ view on consent). even assumed consent, appropriate. Yet when it sensitivity Name and e-mail -11% Dilemma of consumers‘ Interests -13% Convenience desire for control and Past purchases -11% Control "I do not like if a website asks me for the same information every time convenience "I would like to have the opportunity I open it." Medium Purchase plans -11% to always decide whether my data can sensitivity be collected." Media usage -14% 44% 23% Location -13%

Social network posts -11% 40% Health info -15% 63% High 82% sensitivity Financial data -13% 38%

Credit card data -10% 23%

Low High 12% 8% Data deletion 3% 3% 6% At users‘ request Automatically after one year Not defined Strongly agree Agree Undecided Disagree Strongly disagree Note: Based on primary consumer research with representative sample of n = 3,107 (NED: 1,026; GER: 1,041; PL: 1,040) Source: BCG digital identity survey (n = 3,107, August 2012) Source: BCG digital identity survey (n = 3,107, August 2012)

47 | 48 aA ndefinid howtion of dconsumeigital identity rs really see it Part 2

came to credit card data or financial information – like if a website asks me for the same information • The fundamentalists. Comprising approximately and less than 0.1% would not share any data at all. among the most highly sensitive types of data – every time I open it.” Control and convenience are 25% of the population, this segment is generally While nearly all respondents fell into the ”prag- 83% of respondents said an opt-in mechanism important aims. They are often conflicting aims, distrustful of organisations that ask for personal matics” segment, we did find that privacy values was a must. With regard to consent, the survey too. Balancing them will not be easy – but it will information. They are not willing to sell any private differed by country. Dutch respondents had the also highlights a key dilemma digital identity poses: be critical. information, regardless of the incentives they highest (meaning they had the lowest average Consumers want more control of privacy, are offered, and when given a choice, will always willingness to share), with Germans slightly yet at the same time, want more convenience choose privacy controls over benefits. behind and Polish respondents with the lowest in the way they perform transactions and interact Most consumers want to share their data – if average privacy value. with organisations. As Exhibit Dilemma of con- the benefits and controls are right • The pragmatics. The group comprises the ma- sumers‘ desire for control and convenience jority of the population, about 65%. These individ- A framework for homing in on value and sen- illustrates, most respondents – 82% – expressed Back in the 1990s, the renowned privacy uals weigh the benefits of sharing their private sitivities the wish to decide for themselves whether researcher Alan Westin identified a specific seg- information against the degree of intrusiveness. to allow data use in each instance. But a majority – mentation of the population, three groups defined They believe that organisations, whether in the In part 3 of this report, we look at the key 63% – also agreed with the statement, “I do not by their privacy views and behaviour.6 private or public sectors, should “earn“ their trust applications of personal data for each sector, as and want the opportunity to decide for them- well as the less developed “cutting-edge” appli- Segmentation along selves whether to opt out of data uses. They are cations whose ultimate impact is hard to predict – privacy value willing to share data if their privacy concerns are but could be dramatic. Share of respondents willing to provide data1 properly addressed. 100% While the applications vary, a common • The unconcerned. Making up just 10% of the framework can be used to evaluate every use case. 80% population, this segment is generally trustful of This framework focuses on the two key dimensions organisations that want to collect their personal in how digital identity is leveraged: the method of 60% information. They will readily give up privacy con- data collection and the manner in which it is used. trol to secure benefits and they are not in favour Consumers consider these same two dimensions 40% of enacting new privacy laws or regulations. when viewing the trade-off between the value and cost of sharing information. Indeed, their wil- 20% In the Internet era, 20 years is a long time, lingness to share declines as the methods for col- and our research shows that a significant shift in pri- lection and use move toward the outer boundaries;

0% vacy valuation and behaviour has taken place since for example, when organisations acquire data via 5 10 15 20 25 30 35 40 45 50€ C ompensation ( € per month) Westin formulated this segmentation. Indeed, on tracking or data mining, and when they sell it to the Internet today there are essentially no “fun- third parties. Thus, consumer sensitivities can also Very low Low value Medium value High value Very high value damentalists” and no “unconcerned.” Nearly the be plotted on the same framework, and at a quick value entire population are ”pragmatics“ – willing to glance we can see both the opportunity and the Ø Value €10 €10 €20 €31 €44 share their data for the right benefit (as shown risk any single digital identity application presents Share of n 5% 23% 42% 27% 3% in Exhibit Segmentation along privacy value). We (as shown in Exhibit Digital identity framework). found that only 5% of respondents had a very low

1 Scenario: Average for sectors and data types; data collection: “Actively, by providing data“; data use: “Allow other companies to use data anoymised“; privacy value, and less than 0.1% were giving all Another lens through which to look at appli- the right to be forgotten: “Not defined“ of their data away for free. At the same time, just cations is the traceability and completeness of the Note: Segmentation for Germany Source: BCG digital identity survey (n = 1,041 for Germany, August 2012) 3% of respondents had a very high privacy value, personal data that is used. Individuals typically

6 Alan F. Westin, 1991; Kumaraguru: “Privacy indexes: a survey of Westin’s studies,“ 2005

49 | 50 aA ndefinid howtion of dconsumeigital identity rs really see it Part 2

Digital identity framework Dark bubble representing po- Prevalent method tential future value creation of data collection of use case

Mined (profiled)

Tracked Arrow showing how (contextual) use cases evolve Matrix colours indicating how sensitive consumers are to share their personal data Required Personal data application (structured) with bubble size representing current value

Volunteered (unstructured)

Delivery of Enhancement Transfer to Transfer to Major use requested of services third party third party of data service (anonymised) (traceable)

prefer that businesses use digital identities that We hope that those frameworks, combined are incomplete and not traceable – data that may with the insight derived from the survey, will help provide insight on usage or purchasing patterns, organisations from all sectors to home in on ap- for example, but can’t be linked back to a par- plications that don’t just offer value potential, but ticular person. Organisations, on the other hand, the balance necessary to realise it. generally prefer to use complete and traceable dig- ital identities. Between these two poles, however, is space for compromise: solutions that balance privacy with value generation. For example, if control schemes guaranteed that digital identities remained untraceable, more complete identities could be used. Likewise, if data collection was limit- ed to less complete information, more traceable digital identities could be employed. Just as organ- isations need to understand the impact of their collection and use methods, they must carefully assess the levels of traceability and completeness to be used in their applications and business models.

51 | 52 Part 3

part 3 The value of digital identity – for organisations and consumers

P353 | 54 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

The value of digital identity – for organisations and consumers For the purposes of this calculation, we • governments: The value for governments defined value as follows: and the public sector includes increased tax revenues and spending reductions (nearly • consumers: The value for individuals in- all of the latter, however, are expected to be While retailers and online businesses may it include second-order effects, for example, if cludes the consumer surplus of Internet handed back to citizens in the form of tax have been the first to appreciate – and seek out – personal data applications improve medical care services2; lower prices (or taxes) resulting reduction or other relief). the value potential in digital identity, they now and prolong lives, that means people spending from organisations passing back efficiency have plenty of company. Even organisations from more money. gains they derived from using personal data; Our quantifications do not include the most traditional and least data-intensive and time savings achieved via self-service second-order effects. As those effects and sectors have received the message: There is op- Yet opportunities – and the strategies to and other digital identity applications. as-of-yet unknown applications come to portunity in digital identity – opportunity to im- realise them – will vary from sector to sector. Key fruition, they will add to the value generated prove customer service, marketing efficiency, new applications, as well as the drivers that make per- • businesses: For private-sector companies, by personal data applications. product development and, in a nutshell, growth. sonal data available, also vary among the sectors. organisational value consists of additional So, too, do the less developed applications, the revenues along with the cost savings that All told, the quantifiable benefit of per- cutting-edge uses of digital identity whose pros- remain after any “hand-backs” to consumers. sonal data applications can reach €1 trillion pects are harder to assess. annually by 2020 – with private and public or- ganisations reaping about a third of the total, Accordingly, in analysing the uses and ben- and consumers the rest (the latter explained in efits of digital identity, this report looks at the 1. Traditional production State of the sector part by the high value consumers place on free eight major sectors separately. For each, it ex- online services they receive in exchange for their plores key applications and where they fit Traditional production companies like auto- data). This may well turn out to be a conservative according to the framework introduced in part 2. • While digital identity is only beginning to be mobile and consumer goods manufacturers are estimate, since we based it on today’s major use It looks at the potential benefits for both organi- embraced, it is expected to be a key growth quite familiar with the concept of “just-in-time“ cases, which represent only a subset of the appli- sations and consumers, outlines case studies and driver in an otherwise stagnant sector. production. For them, digital identity can be con- cations we will see ten years from now. Nor does notes applications on the periphery. sidered a “just-in-time“ revenue enhancer. • Social media and in-product sensors are driving data availability, which in turn will In a sector where growth has stagnated Quantifying the potential value of digital sector-specific business case for each appli- drive new revenue-enhancing uses of digital overall, several factors combine to give digital identity applications cation, based on data from research houses identity. identity significant potential to create value. First, and public sources (e.g., development of popu- there is the explosive growth of social media, which To estimate the potential value of lation, development of market sizes), as well • Targeted marketing, currently the main use is facilitating new channels of interaction with digital identity applications, we performed as BCG’s own project work experience (e.g., case, will continue to grow in sophistication customers and generating an enormous amount of a bottom-up calculation based on the key benchmarks for efficiency improvements). and importance. unstructured data: customer feedback, usage data, use cases per sector. First, we identified ap- Finally, for each use case, we derived a “digital reviews and opinions. This type of information, and proximately 20 core applications of personal identity value share”1 to reflect the particular • Personalised products and better customer the insight it can reveal, has traditionally been hard data (such as targeted marketing, citizen self- importance of personal data to that appli- insight promise significant benefits to con- to come by in the sector. Second, companies are service and loyalty programmes). Then, tak- cation. sumers – both quantifiable (time and money starting to incorporate sensors in their products – ing the sectors one by one, we developed a saved) and unquantifiable (increased satis- devices that collect and transmit “back home“ data faction). such as usage and location information.

1 See appendix: “Application of digital identity value shares in quantification model.“ 2 The “consumer surplus” is the value Internet services (such as social networking sites) have for consumers beyond what they have to pay for those services (we include Internet access fees and device costs in their total payment); see BCG: “The Internet Economy in the G-20,“ March 2012.

55 | 56 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

The dramatic increase in data availability is out. Big data expenditures – expected to reach additional revenue. A case in point: SNCF, one Digital identity value Traditional production will create an array of opportunities. Chief among $17 billion in 2015, with a 39% CAGR worldwide – of the European railway leaders, developed a EU-27 in 2011 and 2020, billion € them: enhanced and personalised products, new are indicative of the sector‘s growing focus on digital reservation and ticketing solution for subscription-based services and targeted market- converting data into value.3 mobile phones. It smartly leverages location ing campaigns. And make no mistake: The word data; for instance, a customer can purchase a 70 CAGR ticket for the next train simply by passing the 95 % Key personal data applications phone in front of a terminal on the platform. Traditional production For SNCF, it is not just about transaction cost Prevalent method of data collection 3 - Targeted marketing savings, but about giving consumers greater convenience and more choices. 63 +22% Mined (profiled data) • Targeted marketing: As pioneered by the e-commerce sector, targeted advertising lets Organisations 12 Tracked 2 – Product enhancements (contextual data) companies deliver more personalised ads to Consumers 11 7 +25% specific audiences defined by demographics 1 2011 2020 or behaviour. The traditional production sector Required runs mainly branding campaigns and has been (structured data) Source: Company reports; EITO; EIU; Gartner; Eurostat; OECD; IDC; Forrester; shifting these to the online space, increasingly IAB; BCG economic value modelling

Volunteered leveraging targeting techniques to reduce (unstructured data) 1 – Consumer insight waste. Certain forms of targeting can increase • Machine-to-machine (M2M) connections efficiency by an order of magnitude and, based that trigger preventive actions – and reduce Delivery of Enhancement Transfer to Transfer to Major use requested of services third party third party of data on our project work in the digital marketing product failures. Sensors in automobiles al- service (anonymised) (traceable) space, BCG considers an overall efficiency in- ready track system performance. Taking this

Year Value generation Consumer sensitivity crease of 30% for branding campaigns to be a step further, the data could be analysed for

Very low (< 10) High (> 30 – < 35) realistic. symptoms of potential trouble, enabling correc- 2011 2020 Low (> 10 – < 20) Very high (> 35) tive action or early warnings to the driver. Small Large Medium (> 20 – < 30) On the cutting edge – where applications may be more sophisticated and intriguing, but Computers and mobile phones aren‘t the Source: BCG digital identity survey (n = 3,107, August 2012); BCG analysis their prospects are more uncertain – we have only devices that connect to the Internet. Advanc-

observed the following uses: es in communication technologies and miniatu- Key applications of digital identity Gamble‘s Vocalpoint provides an online forum risation – combined with cost reductions – have to engage target customers in the development • Connected devices that don‘t just monitor but made it possible to give practically any product • consumer insight: By collecting and analysing of new products (PUR water systems being one learn. These devices analyse the daily habits or the ability to “phone home“ or communicate the increasing amount of consumer data that product shaped by this process). behaviour of users and adjust functionality and with other machines. Indeed, by 2015, more than is available – from online comments on social preferences accordingly. The new Nest thermo- 75 million additional devices other than com- media sites to real-time usage data transmitted • Product enhancements: By leveraging usage or stat, for example, learns when users tend to be puters and phones will be directly linked to the by sensors – companies can develop products location data transmitted by connected devices, out of the house, turning the heat down when Internet in Europe via mobile telephony net- and delivery processes that are better suited companies can develop new subscription-based they are at work and turning it up just before works. Potentially hundreds of millions more to consumer needs. For example, Procter & services or personalise existing products for they come home.4 will be equipped with near-field communication

3 Source: IDC (March 2012) 4 www.nest.com

57 | 58 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

Connected devices The benefits for organisations and consumers Next-generation growth beyond mobile Selected software and device platforms installed base develop- tablets and smart- ment in five years after product launch phones and computers phones reach mass Mobile connections growth 2010-2015 (CAGR %) • Improved consumer insight helps companies Global installed base (million) iPhone (Q3 2007) adoption at an 100 improve delivery processes and better focus 200 unprecedented pace ... Consumer devices R&D efforts. 180 80 160 • Personalisation features make products more 140 60 120 Health care attractive to consumers, allowing companies to increase sales and/or maintain price points. 100 iPad Other M2M (launched Q3 2010) 40 80 Security / safety Telematics iPod (Q1 2002) • Data-driven subscription services provide addi- 60 Remote monitoring 20 tional revenue streams and create a continuous 40 Netscape (Q4 1994) Retail Energy / utilities relationship with customers. 20 PC1 (1982) Fleet management 0 0 Connections 2015 (million) 0 10 20 30 40 • Targeted advertising maximises returns on a Q1 Q3 Q5 Q7 Q9 Q11 Q13 Q15 Q17 Q19 marketing investment, homing in on those Quarters since launch 1 Based on estimates of PC shipments, and assuming a launch date of 1982 (even Bubble size = connections 2015 consumers most likely to respond favourably. though PCs were available prior to that date, the widely recognised launch of the PC was in 1981, by IBM) Note: Netscape users limited to U.S. only Source: ABI Research; Berg Insight; IDC; Yankee Group; Gartner; Analysys Mason; BCG analysis • Personalised products bear tremendous un- Source: Company reports; Morgan Stanley research; companies' websites; BCG

tapped potential to save time and energy for analysis consumers. Real-time traffic guidance and chipsets like radio-frequency identification (RFID) shooting and preventive maintenance can be en- or Bluetooth, which allow them to link with the hanced because devices can now literally sound … and will become the Internet through other devices. the alarm when problems may be lurking. “remote control of life“

But it is what they will be communicating A second development sparking opportu- that will make things interesting. Onboard sen- nities for personal data applications in traditional sors – also more advanced and economical than production is the increasing ubiquity of mobile ever – will measure all manner of information; devices. In Europe‘s largest markets, smartphone first and foremost, usage data that can be linked penetration is at 40% to 60% – and growing.5 to the device‘s owner. Collected, transmitted and Meanwhile, Apple has sold more iPad tablets analysed, this data can lead to insight that can be after two years than it sold iPods after five. The leveraged across the value chain. popularity of these connected devices, combined Smart mobile devices with their intuitive interfaces and ability to run Research and development efforts can third-party software, make them ideally suited be better focussed, thanks to usage data tell- to serve as ”remote controls of life.” Users will ing manufacturers what products and features be managing all manner of products with just a customers like. Personalised products can be few taps and in the process, establishing a con- more easily, and creatively, developed. Trouble- nection between those gadgets and the Internet.

5 Source: Gartner (July 2012)

59 | 60 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

intelligent management of home energy con- down all of their lights, appliances, heating 2. Retail accordingly. They have also been able to reduce sumption are still in the early stages of their units and so on. waste by optimising demand projections and respective life cycles. • Retailers are already making extensive use stock levels. Further, the sector has actively em- But the system goes even further: Users can of digital identity, aided by the availability braced the larger data ecosystem, partnering with RWE SmartHome – giving consum- create profiles to personalise – and opti- of rich data sets derived from loyalty pro- consumer goods producers as well as fellow re- ers control of their home energy mise – their energy use, specifying, for grammes and their ability to track trans- tailers to share data and generate value from the consumption example, when lights and heating should actions. added insight. be turned on and off, or noting preferences In Germany, RWE launched SmartHome in like their preferred room temperature • The uses of personal data are relatively ad- Looking forward, smartphones are revital- 2010 to help its customers manage their or how bright they like the lights in the vanced – rivalling, and in some cases sur- ising loyalty programmes and targeted adver- energy use and costs. A family of devices – living room. As convenience and comfort passing, the Internet sector. Partnering and tising, extending both their reach and their including thermostats, actuators and plug are enhanced, so, too, are savings. RWE sharing data with external parties, such as breadth. New mobile loyalty apps not only enable adapters – connect to lights, heating con- expects that SmartHome can help to cut consumer goods producers and other retail- smaller retailers to set up programmes, but let trols and appliances in the home. These household energy costs by an average of ers, is not uncommon. companies push offers more directly to users. devices are then linked to a central con- 20%.6 As energy use is reduced, pollution, Real-time location data lets them deliver personal- trol unit through a wireless network. From carbon emissions and the drain on natural • Current applications leverage digital identity ised offers at just the right moment, such as when a single interface on their PCs or smart- resources are decreased as well. to target personalised offers, improve inter- a customer is near, or actually in, the store. phones, users can turn on, adjust or power nal processes and stock levels and gain cus- tomer insight. With retailers already leveraging digital identity for customer insight, targeted marketing, • The mobile communications area presents a internal efficiency and partnerships with third major opportunity. Real-time location data parties, the potential for additional value growth will allow retailers to more effectively deliv- appears considerably smaller than in other sec- er targeted, timely offers, and mobile appli- tors. But this is an industry that has been on the cations will enable smaller retailers to launch forefront of putting data to work, and its embrace loyalty programmes (solving the “no room in of cutting-edge applications, combined especially the wallet“ problem). with the burgeoning growth of mobile devices, may lead to some unforeseen growth.

State of the sector Key applications of digital identity Continuous pressure on profit margins Etiam rhoncus. Maecenas has given the retail industry plenty of incentive • c onsumer insight: By analysing feedback and tempus, tellus eget condimentum to leverage digital identity – and it has not let the data in the social Web, retailers can improve rhoncus, sem quam semper libero, opportunity pass. The sector is already making their demand planning, product offering and sit amet adipiscing sem. extensive use of personal information; by some process design. Fashion retailers are particularly measures more so, even, than the Internet sector. proactive in this regard, as intimate customer By tracking transactions and leveraging loyalty knowledge can decide a season‘s success. Aber- programmes, retailers have been able to gain insight crombie & Fitch, for example, has more than 7 on customer preferences and target their offers million Facebook fans, while H&M and Zara each

6 Source: RWE

61 | 62 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

Key personal data applications cate targeted offers. When executed well, they Loyalty programmes have been a power- Retail Prevalent method can increase revenues and customer retention. ful tool for retailers, enabling them to incenti- of data collection 4 - Targeted marketing A particularly successful example can be found vise repeat purchases (through rewards and

Mined at Tesco. Its “Tesco Card“ programme enables discounts) while gaining insight about custom- (profiled data) the retailer to thoroughly analyse customer ers’ purchases and preferences. However, tra- data and create and send personalised vouch- ditional loyalty programmes have limitations. ers. Tesco – whose loyalty programme members For one thing, there is only so much room in a Tracked 3 – Loyalty programmes (contextual data) account for 80% of its sales – has a voucher consumer‘s wallet, apparently limiting the num- redemption rate of about 15%, far above the ber of concurrent programmes consumers will 9 Required industry standard of 1 to 2%. use to two. Sign-up procedures can be cumber- (structured data) 2 –Service enhancements some, and the programmes are very expensive to • Targeted marketing: Retailers have been at the set up, putting them out of reach for most small Volunteered forefront of leveraging digital data to deliver and medium-size retailers. As a result, the retail (unstructured data) 1 – Consumer insight offers and advertisements to narrowly focussed sector has reached a saturation point for loyalty groups of customers. Now they are extending programmes. Delivery of Enhancement Transfer to Transfer to Major use requested of services third party third party of data this into the mobile space, sending offers direct- service (anonymised) (traceable) ly to customers‘ smartphones and leveraging Smartphones promise to shake things up, Year Value generation Consumer sensitivity real-time location data to create and deliver expanding both the reach and depth of these in- Very low (< 10) High (> 30 – < 35) offers that are both relevant and timely. Fast- itiatives. Mobile loyalty apps and wireless inte- 2011 2020 Low (> 10 – < 20) Very high (> 35) food chains like McDonald‘s are leading the way. gration with point-of-sale terminals mean that Small Large Medium (> 20 – < 30) the device itself can be used in place of cards, Source: BCG analysis On the cutting edge – where applications adding convenience and creating a “bottomless may be more sophisticated and intriguing, but wallet“ where dozens of programmes can be as Digital identity value have more than 12 million, and H&M has even their prospects are more uncertain – we have easily managed as two. Retail 0.5 million people talking about its brand on observed the following uses: Facebook.7 The mobile element also enables retailers Current to leverage new types of data and create innova- loyalty Number of monthly used loyalty programmes, EU-27 in 2011 and 2020, billion € • s ervice enhancements: Companies can enhance % of loyalty members, global1, 2011 tive features and incentives for their customers. the shopping experience through more active in- programmes' Mobile apps can transmit location data, enabling limitation 40 CAGR volvement of customers and integration of social timely offers (for example, a discount delivered 39 Less than 30% of 95 % 31 % Internet services. For instance, Woolworths Aus- 30 % loyalty card users while the customer is in the store – or even the 30 with > 2 active tralia extended the grocery shopping experience programmes competitor‘s across the street). They could award +22% 31 to major commuter hubs with its virtual shopping points for scanning products that interest the 14 20 Organisations 16 % 7 wall and an accompanying smartphone app has customer or provide access to reviews. Consumers Consumers been downloaded more than 1.5 million times.8 11 % benefit from more relevant offers and informa- 7 8 +25% 10 6 % 5 % tion. Retailers benefit from greater customer 2011 2020 3 % • loyalty programmes: These programmes give 0 insight – and an even deeper relationship than retailers an effective way to collect and analyse 0 1 2 3 4 5 >5 before. Nor would loyalty programmes remain Source: Company reports; EITO; EIU; Gartner; Eurostat; OECD; IDC; Forrester; 1 Average for U.S., United Kingdom, France, Germany, Brazil, China, Japan, IAB; BCG economic value modelling a customer‘s purchase data and to communi- India, Russia, and Canada | Source: Gartner; BCG analysis the domain of the largest retailers; mobile

7 Source: Socialbakers (October 2012) 9 Source: Humby, Hunt and Phillips: “Scoring Points: How Tesco Continues to Win Customer Loyalty,“ 2008; press 8 Source: Woolworths; press

63 | 64 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

devices and apps open up the possibilities – and The benefits for organisations and consumers ucts (using a smartphone camera). While $110 million in retail revenue could be linked the payoff – to even the smallest shops. there is a direct benefit to consumers – back to Shopkick. • Loyalty programmes – traditional and mobile – “kicks“ can be redeemed for rewards or donat- • Mobile loyalty apps. Bringing loyalty pro- provide retailers with deep consumer insight, ed to charity – the payoff for retailers can be With more than 3 million users and grammes to smartphones offers several key as all purchases can be tracked. They also boost even greater. Shopkick drives foot traffic into 7,000 participating stores – including major advantages for retailers: It increases the reach customer retention and repeat purchase rates. stores, while the data it collects – including chains like Best Buy, Old Navy and Target – of the programmes, enabling businesses to use location and interests – drives sales. Retailers Shopkick demonstrates that well-crafted, real-time location data to send offers at oppor- • Targeted marketing delivers relevant offers to can offer highly relevant deals and discounts non-intrusive mobile apps can be a boon to tune moments. It also enables smaller retailers – those consumers most likely to respond, maxi- while the user is still in the store. In 2011, both retailers and shoppers. who otherwise could neither afford the setup mising the effectiveness of a marketing invest- costs of a conventional programme nor con- ment and reducing customer acquisition costs. Innovative approaches vince customers to carry one more card in an in retail overstuffed wallet – to launch their own loy- • Greater consumer insight lets companies reduce wynsh Wynsh puts customers in driver‘s seat of marketing … alty programmes. waste by better managing inventories, product offerings and internal processes.

• Greater customer participation. Mobile apps You decide where your discount should be offered and social media enable customers to more • Real-time location data enables consumers to actively communicate their likes and interests receive timely, relevant offers – which are partic- to retailers. Such insight can lead to highly ularly convenient when received near or in the effective personalised discounts. For example, store. the Wynsh app lets users photograph a product and ask for special offers. Since an individual • Innovative mobile applications give consumers

discount has more value to a customer than a new and useful functionalities – such as com- App identifies nearby User photographs Photos are sent to Retailer grants individual general offer, retailers can achieve the same re- municating (via a photo or barcode scan trans- participating stores products of interest retailer discount to user

sult – a sale – with a lower discount. mitted by smartphone) what products they are * Discounts granted are variable and are set by the brand in question. shopping for, and receiving on-the-spot dis- • Virtual shopping ”walls.” Blurring the line counts. between retail and e-commerce, this inno- Shopkick … and Shopkick increases loyalty reach vative application allows retailers to post visual representations of their products in public Shopkick – an innovative mobile loyalty spaces such as subway stations. When custom- and retail marketing solution ers see a product they want, they scan its bar- code with their smartphone camera. After the Shopkick leverages an object nearly purchase is authenticated, a home delivery is every shopper brings into a store: their mo- scheduled. Tesco Home Plus has built virtual bile phone. The free app, which combines shopping walls at public transport facilities both loyalty and targeting elements, awards

across Korea. With a low capital expenditure, points – called “kicks“ – to users each time User installs app on App identifies nearby Device in stores automa- User can redeem 10 smartphone participating shops and tically registers walk-in collected “kicks“ for it was able to boost its online sales by 130%. they enter a participating store; addition- restaurants and user gets “kicks“ discounts and free al points are earned for scanning prod- points services

10 Source: Tesco; BCG analysis Source: Wynsh; Shopkick

65 | 66 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

3. Financial services major use case is process automation; primari- Key personal data applications Prevalent method 4 - Targeted marketing Financial services ly, self-service transactions through the Web of data collection • Banks and insurance companies possess a and, increasingly, mobile. These efforts are pay- Mined 3 – Scoring & wealth of data due to their everyday busi- ing off handsomely: Online and mobile trans- (profiled data) rating ness. action costs run about 5% of their in-branch counterparts.11 Tracked • The current major use case is process auto- (contextual data) mation – especially customer self-service – There is much more potential to tap. The but the sector can significantly increase sector can boost the value it creates with dig- Required 2 – Personalised digital identity value by moving beyond these ital identity fivefold by moving beyond core (structured data) products / services core applications. applications. Some of these we are already

starting to see, in more sophisticated uses of Volunteered • Increased use of mobile devices will not only data to personalise products, more accurately (unstructured data) 1 – Process automation make more data available in real time (such score and rate potential customers and target Delivery of Enhancement Transfer to Transfer to Major use as location and transaction information), but advertising. requested of services third party third party of data will allow sector companies to develop per- service (anonymised) (traceable)

sonalised services and enhance their targeted The mobile space offers even more pos- Year Value generation Consumer sensitivity

marketing efforts. sibilities, including digital wallets and mobile Very low (< 10) High (> 30 – < 35) payment, as well as the ability to target ad- 2011 2020 Low (> 10 – < 20) Very high (> 35) • Mobile wallet applications are potentially vertising and even product offerings based on Small Large Medium (> 20 – < 30) lucrative – but only if the purchasing data real-time location data. Source: BCG digital identity survey (n = 3,107, August 2012); BCG analysis collected can be used or monetised by way of targeted advertising. Digital identity value Key applications of digital identity • Personalised products / services: New data Financial services types (such as location and usage informa- EU-27 in 2011 and 2020, billion € 96 State of the sector • Process automation: Customer self-service – tion) available via mobile devices and sensors CAGR through the Web or mobile applications – allow for an array of innovative products. GM‘s The financial sector possesses an enor- provides an automated, anywhere-anytime OnStar system, for example, is a subscription- 95 % mous cache of personal data as a result of its alternative to more expensive and more time- based service in which an in-vehicle device 41 +21% everyday business. That volume is only going to consuming phone or in-branch transactions. captures diagnostic and usage data (such as grow with increased use of electronic payment But automation can mean much more than mileage) and provides navigation and roadside methods. Meanwhile, the burgeoning use of simply paying bills or checking an account assistance. Through a partnership with GMAC 29 mobile devices will provide new data types – in balance. For example, AXA Insurance has de- Insurance (and with driver permission), usage Organisations 7 particular, location information – and make them veloped a smartphone app that lets customers data is sent to the insurer, which can result in a 55 +8% available in real time. This wealth of data will file claims right at the scene of an automobile low-mileage discount. Consumers 22 mean a wealth of possibilities. accident. They can use their mobile device‘s GPS capability to find the closest approved • Scoring and rating: By accessing and inter- 2011 2020 So far, the sector is just scratching the repair shop, and even tap a button to place a preting data from additional available sources, Source: Company reports; EITO; EIU; Gartner; Eurostat; OECD; IDC; Forrester; surface of what it can do with digital identity. The call for a tow. such as the social Web, companies can improve IAB; BCG economic value modelling

11 BCG analysis

67 | 68 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

their scorings of customers, thereby improving health risks as blood and urine tests. Although 70% of all bank transactions, and walking into a Basing insurance premiums on actual, risk management and fraud prevention. Case in this practice is highly controversial, with criti- branch will be an increasingly infrequent event. rather than predicted, behaviour. As tele- point: Aviva ran a trial with 60,000 individu- cal questions regarding individual consent yet Good news for those who‘d rather spend their matics-capable devices emerge – recording every- als in the US in 2010, showing that online data to be answered, this example illustrates how lunch hours elsewhere. thing from how far an individual walks each day detailing food purchases and social activities much insight about a person can be derived to how fast they drive their car – they will enable could be as effective in identifying potential from such data. On the cutting edge – where applications insurers to base premiums on actual activities may be more sophisticated and intriguing, but and behaviour instead of statistics and predic- Development of self-service their prospects are more uncertain – we have ob- tions. This will give consumers additional choices, in banking served the following uses: like paying car insurance only for the miles actu- Share (in %) of interaction type with banks from 1990 to 2020 in the U.S. ally driven. 100 Mobile banking Intensified use of location data to inno- vate services. The location data provided by mo- Digital wallets. While not presently a 80 bile devices can help companies match the right mainstream payment option, an increasing num- Self- service Online banking customer with the right offer – at the right time. ber of mobile handsets are coming equipped 60 68% For example, insurer Tokio Marine, in coopera- with near field communication (NFC) technology, tion with mobile network operator NTT DOCOMO, which will allow consumers to complete a trans- 40 ATM offers specialised one-time insurance policies – action simply by holding their phone to a point- via mobile devices – based on a user‘s location and of-sale unit. In itself, these payment transactions Call centre 20 Agent inferred activity (e. g., golf insurance for someone will not lead to significant revenue gains. But 32% Branches currently approaching a golf course). the data that is collected can give valuable insight 0 1990 1995 2000 2005 2010 2015 2020 Digital wallet / Digital wallet revenue pools in billion USD p.a. and competitive intensity1, 2015 Actual Forecast m-payment revenue pools Source: NRC; Deutsche Bank research; Forrester; BCG analysis Potential revenue pool, 2015 Targeted marketing • Targeted marketing: Companies are using transaction costs while improving customer con- High ($10B – $20B Deals/offers the growing amount, and detail, of personal venience by increasing availability of the service. per year) and advertising data to more efficiently target their marketing activities – homing in on specific audiences With the rise of online and mobile bank- instead of a more general placement. ing, customers are able to not only serve them- Medium Transaction fees ($5B – $10B selves, but to do so from any place, at any time. per year) International Banking transactions tend to be high- Little wonder, then, that self-service now ac- transfers Micromerchant volume, standardised processes – a balance in- counts for nearly 50% of all bank transactions, acquiring Unbanked 2 quiry, a transfer between accounts, a credit card more than double the figure from the early Low (prepaid/mobile money) (< $1B – $5B Domestic payment. That makes them particularly well suit- 1990s. And the upward trajectory will continue, per year) transfers Virtual currency ed for customer self-service. Automated teller assisted by biometric technologies that will en-

machines were an early example of “banker- hance authentication and convenience (no pass Low Medium High less“ banking and demonstrated that by taking codes to remember). By 2020, ATM-based, mobile (0% – 10%) (10% – 25%) (25% – > 45%) humans out of the equation, banks could save and online self-service will account for nearly 1 Percentage of business pursuing business opportunity (based on 100 companies observed) Competitive intensity 2 Relevant only in emerging markets with undeveloped banking system Source: BCG: “How Banks Can Take the Lead in Mobile Payments,“ June 2012

69 | 70 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

about a consumer‘s interests and preferences – in- Before the full potential of the digital wallet Location-based one-time insurances sight that can make financial companies an impor- can be realised, however, there has to be widespread tant player in the targeted advertising space and acceptance of it by merchants. The increasing avail- Tokio Marine Nichido open up new revenue pools. ability of near-field communication in smartphones, as well as point-of-sale technologies that use it, is a For the financial services sector, digital hopeful sign that we may not need to wait for long. wallets, or mobile payments, will bring transaction fees from three major sources in developed mar- The benefits for organisations and consumers kets: Insures damages against Insures golf equipment • Process automation enables a company‘s IT systems others‘ property or health against damage and theft • Domestic and international transfers diverted to do the work of more expensive manual processes. from traditional payment channels (digital wal- lets are expected to capture just a small share of • Personalised products and services enhance the

the overall transaction volume). sales potential of existing revenue streams and Insures the cost of a hole Insures the hospital costs open up all-new ones. in one celebration (drinks after a golf accident & presents) • Payments in the virtual world, where consumers look for alternatives to credit cards. • The insight that can be gleaned from purchasing Smartphones are used to localise opted-in users Data used to push location-specific insurance offers to users data can be monetised via sale to, or partner-

• Payments to small merchants who lack credit ships with, advertisers and other organisations. Source: Tokio Marine; NTT Docomo card terminals, and to larger merchants instead of cash payments. • Targeted marketing allows companies to more Pay-as-you-drive cost-efficiently reach those consumers most insurance Yet transaction fees are just a part of the likely to respond, maximising the impact of a GMAC

potential revenue digital wallets can bring finan- marketing investment. GMAC Insurance cial services companies. Advertising that is targe- Vehicle Adjusted ted based on mobile payment data is, in our esti- • Consumers save time and gain convenience diagnostics insurance fee reports mate, up to 10 times more valuable to marketers through anywhere-anytime self-service trans- than advertising targeted according to context or actions. behaviour. This makes the data recorded by digital PAYD1 insurance wallet applications highly valuable as well. • More accurate scorings and ratings lead to fairer pricing for insurance and other services. GM Nor are the finance companies the only ones to benefit. Brick-and-mortar merchants will Tokio Marine Nichido and GMAC: Innovative be able to measure how online advertisements im- one-time and “pay-as-you-drive“ insurance pact offline purchases, so they can gauge – and policies Driver‘s improve – the efficiency of their campaigns accord- telematic data Usage data ingly. That’s a capability e-commerce companies Smartphones and sensors make it pos- (mileage) have long enjoyed, but traditional retailers had sible to capture real-time location and usage OnStar previously lacked. data – and already, that is transforming the way

1 Pay-as-you-drive Source: GM; GMAC; OnStar; BCG analysis

71 | 72 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

some financial products are sold. Case in point: 4. Telecommunications ices, putting companies under pressure to dev- driven value creation, and current applications insurance. and media elop higher-value services while streamlining costs. fall far behind the Internet and retail sectors in their sophistication. Currently, the chief use case In Japan, telecommunications giant • Companies have access to a vast amount Sector companies have, in principle, access is process automation. This is no small thing, NTT DOCOMO has teamed up with the insur- of personal information mainly relating to to riches of data. But there are serious caveats. given that many companies still incur significant er Tokio Marine Nichido to offer innovative telephony usage or TV-viewing data, but so Telecommunications and cable companies oper- operational expenditures for their provisioning “one-time“ policies based on a mobile phone far make relatively little use of it. ate the fixed and mobile networks that data passes processes, and the shift to customer self-service user‘s location. Through the GPS capabili- through – putting them in a prime position to and the resulting efficiency gains can be a large ties on a smartphone, a user‘s location data is • Digital identity is expected to be a huge capture usage and location data – but, under EU value driver. But by moving beyond core appli- acquired, and from it, behaviour can be in- growth driver in a sector where growth and electronic privacy rules they cannot collect and use cations, telecommunications and media compa- ferred. For example, someone located in an competition is closely linked to the ability to data, such as traffic data or location data, unless nies can increase digital identity value fivefold. airport is likely taking a trip; someone on a innovate and enrich existing services. users have given their prior consent. Users can also golf course may be playing a round or two. withdraw prior given consent at any time. Developing and benefiting from those The service then sends a message to the • Currently, the chief application is process new applications may nevertheless require a care- phone, offering a relevant policy. NTT DOCOMO automation (in particular, customer self-serv- Like financial institutions, this sector is ful approach, as subscribers need to be able to gains a new revenue stream from location ice), which is helping to reduce the tradi- only starting to explore the possibilities of data- trust their network operators to guarantee the data it already collects, and Tokio Marine tionally high level of operational expenditures Nichido is able to offer just the right policy at just in this sector‘s service-provisioning processes. Key personal data applications the right moment. Services like these also open Telco & media Prevalent method the door to new, specialised insurances for all • Like the financial sector, telecommunications, 4 - Targeted marketing of data collection manner of activities and situations. cable and media companies can see a fivefold increase in digital identity value by moving Mined (profiled data) Meanwhile, in the United States, fairer beyond core applications. automobile insurance premiums – based on ac- tual, rather than predicted, driving behaviour – • The trust that customers put in their network Tracked (contextual data) are now possible, thanks to in-car devices that providers critically depends on their commit- collect and transmit usage data like mileage. ment not to track behaviour over their networks OnStar, from General Motors, is a subscription- other than for purposes related to service deliv- Required 2 – Service 3 - Monetisation of (structured data) enhancements consumer insight based service that has tradi-tionally used such ery, unless customers give their consent. data to provide navigation and diagnostics. Now Volunteered 1 – Process automation it is being used to get low-mileage drivers lower State of the sector (unstructured data) insurance premiums. For a fee, GM sends OnStar data to GMAC Insurance, which then uses it to For telecommunications and media com- Delivery of Enhancement Transfer to Transfer to Major use requested of services third party third party of data adjust insurance premiums according to actual panies, digital identity holds the promise to service (anonymised) (traceable) miles driven. The result: a win across the board. cushion – considerably – the impact of an increas- Year Value generation Consumer sensitivity Consumers save on insurance, GM monetises ingly competitive marketplace. The relief won‘t Very low (< 10) High (> 30 – < 35) data while incentivising more drivers to subscribe come a moment too soon. Disruptive IP services – 2011 2020 Low (> 10 – < 20) Very high (> 35) to OnStar and GMAC improves its insurance Skype, YouTube and Spotify, to name a few – Small Large Medium (> 20 – < 30) pricing – and gains a competitive advantage threaten the revenues generated by their core serv- among a crowded field of rivals. Source: BCG digital identity survey (n = 3,107, August 2012); BCG analysis

73 | 74 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

Digital identity value • s ervice enhancements: By leveraging data Secondary monetisation Telco & media of location data provided by or inferred about users (such as lo- € EU-27 in 2011 and 2020, billion cation data captured by mobile devices on the network), companies can develop – on their CAGR 23 own or in partnership with other sectors – TomTom

personalised services that increase usage and Vodafone provides revenue. Case in point: Netflix, the video stream- TomTom with anonymised 19 +16% real-time location data ing and rental service, claims that its personal Vodafone 6 Organisations recommendations algorithm is responsible for TomTom analyses data to 5 improve traffic forecasts Consumers 4 +17% 12 Vodafone tracks 1 75% of what subscribers watch. movement of mobile 2011 2020 phone users

Source: Company reports; EITO; EIU; Gartner; Eurostat; OECD; IDC; Forrester; • Monetisation of consumer insight: The data Source: Vodafone; TomTom IAB; BCG economic value modelling that companies capture, or the insight derived from it, can be sold to third parties for use in Utilising insight generated from per- The benefits for organisations and consumers confidentiality of communications over their net- their own services. For example, Telefónica has sonal data to increase customer retention and works. But if they tread carefully and transparently set up a business unit – Dynamic Insights – ded- reduce churn. By focusing on key “influencer” • Process automation reduces transaction costs with regard to intended data collection and usage icated solely to seizing the opportunities pre- customers – those whom other customers tend for companies and improves customer satis- processes, and consumers are given an informed sented by the company’s data wealth. Its first to follow – telecommunications companies can faction – which in turn boosts retention rates. choice to consent, network operators would be commercial product, ”Smart Steps,” will leverage more efficiently allocate their retention resourc-es. • Partnering with other sectors to create new in a prime position to capture usage and loca- location data to help retailers better understand For example, Telekomunikacja Polska – the larg- data-driven services opens up new revenue tion data (which it-self can be monetised). They what drives foot traffic to their stores.13 est fixed network operator in Poland – analyses streams – as does simply marketing the data, could also facilitate partnerships to attract third- call records to understand social networks and or insight derived from it, to third parties. party applications to their digital platforms – • Targeted marketing: Leveraging personal data identify prized “multipliers“ within its custom- relationships where another sector‘s expertise and (including real-time location data provided by er base. Doing so has enabled it to improve the • Targeted advertising – especially when en- this sector‘s data combine to create value for both. mobile devices on the network), companies can accuracy of its churn model by 47%15. (Whether hanced by real-time location data – maximis- deliver digital ads to those individuals most an operator can take this approach will depend on es the company‘s own marketing investment. Key applications of digital identity likely to respond. For companies in the telecom- country-specific telecommunication services laws.) It can also mean additional revenue when the munications and media space, this is relevant data is used by third parties to target their own • Process automation: Through the use of se- both for their own marketing and as a facili- Enabling customers to opt in to lo- ads and offers. cure digital authentication, companies can tator of targeted marketing solutions. For examp- cation-based special offers from third par- enable customers to perform transactions in le, Sky is testing ”AdSmart“ in the UK as a solution ties. Leveraging the location data captured and • Self-service anytime-anywhere transactions self-service or completely automate them, both that target ads within Sky programming playing transmitted by mobile devices on their networks, save consumers time, are more convenient online and with mobile devices. This can reduce through set-top boxes and on mobile screens.14 mobile telecommunications providers can offer than manual processes and improve the overall manual effort and costs – sometimes consider- third parties the ability to send highly relevant customer experience. ably. But it often requires rethinking processes On the cutting edge – where applications offers and deals to participating subscribers in

end-to-end; for instance if a product portfolio may be more sophisticated and intriguing, but real time. Case in point: Telefónica o2 estab- • Operational benefits achieved through effi- is too complex, people will have to go to shops their prospects are more uncertain – we have ob- lished its “Priority“ programme in the UK to mar- ciency improvements are partially passed on so it can be explained by a sales clerk, no matter served the following uses: ket event tickets and coupons to subscribers who to consumers in the form of lower prices and how well the online ordering process works. opted in. enhanced services.

12 Source: Netflix‘s own tech blog, 6 April 2012 15 Source: Sybase: “The Importance of Social Network Analysis,“ 2011 13 Launch announced by Telefónica in October 2012 14 Source: BSkyB; press 75 | 76 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

5. Public services / health Key personal data applications Public services /health SFR Régie – leveraging subscriber relation- Prevalent method ship and localisation data • The public services and health sector posses- of data collection

ses large amounts of personal data, but rela- Mined 4 – Personalised (profiled data) 5 – Tax collection Heightened competition and shrinking tively little of it has been digitised. medicine margins may be reining in growth in tradition- al telecommunications services, but for some • It has more growth potential for organi- Tracked 3 – Process automation providers, new opportunities lie, literally, in the sational value than any other sector – large- (contextual data) health care palm of their hands. The increasing populari- ly due to the low-hanging fruit that awaits

ty and capabilities of smartphones, combined once digitisation efforts gain traction. Required 2 – Process automation with the real-time location data that mobile (structured data) public services carriers can so easily obtain, is spurring a host • Digital identity applications currently few

of new data applications. And that‘s spurring and far between, with process automation Volunteered 1 - Citizen self-service (unstructured data) revenues. being the initial focus.

Delivery of Enhancement Transfer to Transfer to Major use The French mobile phone company • The euro crisis has given European govern- requested of services third party third party of data service (anonymised) (traceable) SFR leverages its large customer base and its ments an incentive to increase efficiency, access to location data to offer companies a jump-starting eGov initiatives. Year Value generation Consumer sensitivity powerful way to target their marketing: send a Very low (< 10) High (> 30 – < 35) deal or discount to a mobile subscriber based • Some advanced, if still early-on applications – 2011 2020 Low (> 10 – < 20) Very high (> 35) Small Large on where they are (walking near a restaurant, such as tax fraud detection and personalised Medium (> 20 – < 30)

standing in a shopping mall, entering a de- medicine – demonstrate the high-value po- Source: BCG digital identity survey (n = 3,107, August 2012); BCG analysis partment store and so on). The service – run tential that can be realised. through SFR‘s new mobile advertising arm, The picture is starting to change, though. Digital identity value SFR Régie – is non-intrusive: Subscribers have State of the sector The euro crisis has spurred governments to boost Public services / health to opt in to receive deals; and because the efficiency, jump-starting eGov and citizen self- EU-27 in 2011 and 2020, billion €

marketing is targeted, only relevant offers are Without question, the public services and service initiatives. Meanwhile, the increasing use, 387 CAGR made, so users aren‘t bombarded with unwant- health sector came late to the digital identity par- and advancement, of IT in health care, combined 95 % ed advertisements. ty – and by some measures, it still has yet to ar- with competition and pressure to cut costs, has 120 +35% rive. While organisations possess vast amounts of led to more digitisation and the development of So far, some 3.6 million SFR subscribers personal data, the majority of it remains in paper several very promising digital identity appli- have opted in, receiving timely offers from SFR form. Moreover, there have traditionally been few cations. To be sure, the public sector still lags be- clients like Air France, McDonald‘s and Re- incentives to digitise data and boost efficiency. hind all others, but it is, gradually, starting to gain 62 267 nault. This is just the beginning, too: As more Public entities face little competition, and in the ground. Organisations 8 +19% telecommunications data – not just localisa- health care arena, the complex relationship be- tion info, but usage and status – is incorpo- tween providers and payors, along with the The slow start means that there is plenty Consumers 54 rated into efforts like SFR‘s, targeting should sector’s compensation schemes, means that the of low-hanging fruit. This and the sector’s sheer 2011 2020 become even more specific, and effective. party that collects the information isn’t always the size16 explains why there is more growth poten- Source: Company reports; EITO; EIU; Gartner; Eurostat; OECD; IDC; Forrester; one to benefit from its use. tial for organisational value here than in any IAB; BCG economic value modelling

77 | 78 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

Citizen self-service manual transactions – coming in at roughly 40% tions, and full medical reports – to be accessed the cost today, and dropping to less than 30% by by a variety of stakeholders. A unique personal Transactions Citizen‘s benefits Government‘s benefits the next decade. The cost efficiency is driven by identity number enables reliable identification 1 Share of transactions offered Time per transaction Cost per transaction replacing labour costs that tend to increase at of each patient while ensuring data protection. as self-service 100% 100% 109 % the rate of inflation with IT software and hard- 5 % 80 % 100 % Manual ware costs that are scaling. Self-service trans- • Personalised medicine: By collecting and 95 % actions are also far more convenient for many analysing a comprehensive, up-to-date array 80% 80% -61% -72% people, with no more waits on long lines. In the of patient data, customised health care is fa- illustrative example outline above, what used to cilitated, with treatment focusing not just a 60% ~ 45 min 60% take 45 minutes of legwork now takes 15 minutes presenting condition but the patient’s overall at a PC. situation. A leading example of this is Kaiser -200% Permanente’s personalised medicine initiative. 40% 40% 39 % So far, however, public agencies in Europe – Among the industry’s most advanced data unlike their private-sector counterparts – have users, the U.S.-based health care payor/provid- 31 % 20% Self-service 20% been slow to embrace these solutions. We esti- er captures a broad collection of medical re- 20 % mate that in 2011 just five percent of all trans- cords, from hospitals, medical offices, labs and actions were completed via self-service. That’s pharmacies, analysing them as a set to focus 0 costing governments more money and citizens treatment on the patient’s total health.18 2011 2020 0 Manual Self-service Manual Self-service more time than either should be spending. Manual Self-service ~15 min 2011 2020 Applications using personal data promise • Process automation – public services: Public to revolutionise medical treatment and deliver Source: BCG analysis agencies can use digital data to automate deci- significant cost savings. But realising the full sion making and processes. For instance, algo- potential of personalised medicine, decision rithms based on the personal data of job centre support systems and other tools that leverage other sector. Process automation is currently the global leader in eGov, expects that nearly 80% clients can be used to identify the best-suited pro- digital identity will require a health care setting major value contributor, but emerging uses, like of municipal transactions will be digital by fessional training programme and thereby sup- where IT is utilised broadly – and effectively. Is personalised medicine, hold vast potential – not 2015 and already generates €200 million in port the decision making of job centre employees Europe there yet? BCG’s Health Care IT Score was just to improve efficiency, but lives. annual savings with electronic invoicing alone.17 to enable more efficient use of the public funds. designed to indicate how well a country‘s health care system is prepared to reap digital identity Key applications of digital identity Like banking and telecommunications, • Process automation health care: By making value. the public sector is rich in standardised, high- patient data more easily available to health • citizen self-service: Automated transactions – volume transactions – the kind ideally suited care professionals, process flows, patient care The results show two things clearly. facilitated by secure digital authentication of for customer, or in this case citizen, self-service. and research can all be improved. Collection First, the starting position is very different from citizens – provide a more convenient, more Examples abound: applying for a birth certi- of data via mobile devices further increases country to country and the focus areas to ad- efficient alternative to traditional manual ficate, renewing a driver’s license, registering the volume, data types and timeliness of the vance IT in health care differ as well. Secondly, all processes. A vast array of transactions – every- a new address. Indeed, BCG estimates that by available information. An example of how this European nations have room for improvement. thing from reporting stolen bicycles to obtain- 2020, 80% of all public administration trans- works can be found within the government- ing marriage licenses – can be, and increasing- actions could be offered as self-service in the funded Swedish health care system. It allows ly are, performed via self-service. Denmark, a EU-27 countries. They cost less to perform than for health data – including diagnoses, prescrip-

16 More than 50% of all eight sectors combined. See appendix for a detailed breakdown of sector sizes. 18 Among the prominent examples of how Kaiser Permanente‘s rich data sets were applied is when Merck had to pull blockbuster medica- 17 Source: Case study “eInvoicing in Denmark,“ 2007; press tion Vioxx from the market in 2004, after a study based on Kaiser Permanente data provided proof of Vioxx‘ comparatively high risks.

79 | 80 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

Maturity of IT utilisation in eral practitioners, can keep up with the latest The benefits for organisations and consumers European health care research and advancements in diagnosis and Scores represent absolute maturity, whereas 1 is the lowest maturity score and 5 is the highest maturity score, 2012 treatment, such solutions can lead to many • Citizen self-service reduces costs per trans- more patients benefitting from optimal care. action significantly. Meanwhile, individuals 1 Identification Cost traceability BCG HC IT Score IT adoption Interoperability save time – they no longer need to travel to • Further personalising medical treatment by and wait at public agency offices. DEN 4 3 4 5 3 taking behavioural and genetic data into account. By collecting data on an individual’s • Analysing data to crack down on tax evasion and

SWE 4 3 4 5 3 own DNA, companies like 23andMe aim to gen- fraud boosts government revenues. erate new insights to help guide treatment. Other companies, for instance the online por- • Personalised medicine and other data-driven AUT 3 2 3 4 3 tal PatientsLikeMe, contribute to adding richer health care improvements can result in more data by involving patients more directly. effective patient treatment, better outcomes and HUN 3 2 2 3 3 longer life expectancy. They can also help to cut health care costs. 6 % 6 % 6 % NED 3 3 4 2 3 Five major personal Personal data Description data applications 5 % 5 % 5 % 5 % 5 % application GB 3 3 3 4 2 in health care

Proactive care Utilise personal data to actively monitor and prepare health care visit, not only during visit but also before visit to prepare patient and after visit to give patient further instructions GER 2 2 3 1 2 Example: Sending an e-mail to a patient after a visit, including a summary, relevant educational health material and follow-up appointments

1 Rounded average of maturity scores from rating elements Source: BCG analysis Decision support systems Utilise big data capabilities to analyse digitalised medical records and medical knowledge databases to support doctors in diagnosis and treatment process

Example: Wellpoint and IBM developing a system based on the powerful AI unit “Watson“ • Tax collection: To help detect tax fraud, person- On the cutting edge – where applications with the goal to provide up-to-date, evidence-based and individualised health care al data – such as bank records and credit card may be more sophisticated and intriguing, but

statements – is analysed and compared with their prospects are more uncertain – we have Personalised medicine Base treatment of patients not only on diagnostics, but also on more comprehensive disclosed information. Italy recovered €11.5 observed the following uses: personal data, e.g., genetic characteristics, behavioural data

billion in 2011 by thwarting tax evasion. It uses Example: Analysing genetic data and medical record of patient to predict chances of future a system called “Redditometro“ that analyses • Automated medical decision support. These illness and recommend preventive actions based on patient‘s behaviour some 100 life-style indicators, including gym systems, like the one under development by memberships and private school tuition, to op- IBM and the US insurance company Wellpoint, Process automation Utilise technological capacities to automate services, i.e., telemedicine for video sessions and diagnostics, where only basic information such as heart rate, blood pressure etc. is required timise tax audits. The system assigns taxpayers analyse huge volumes of data from many scores for their risk of evasion, so that audits sources – including patient records, research Example: Reducing hospitalisation time through telemonitoring older persons and those in 19 high-risk groups after a heart attack through digital blood pressure and heart rate checks can focus on highest-risk individuals. papers, medical textbooks, and population health via face-to-face video conferences data – to detect disease and suggest treatment

options. As no medical doctor, especially gen- Enhanced R&D Extract insight from personal data analysis to align R&D goals with regard to a more personalised approach for each patient rather than “one-size-fits-all“ medicine

19 Sources: Republic of Italy; The International Herald Tribune, 7 February 2012 Example: Develop more comprehensive ways to analyse the genetic data to support the development of more effective medicine, with a more preventive rather than reactive focus

81 | 82 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

• The sale of personal information that is unique to • The major part of savings achieved through per- rate is 9% for the United States and 11% for Italy - fighting tax evasion with the public sector – such as motor vehicle data use- sonal data use in health care will also be handed Japan). This translates into many billions of “Redditometro“ ful to insurance companies20 – can provide a minor back to consumers in the form of lower prices and Euros in lost government revenues. source of additional income, but needs to be care- health insurance premiums. Italy has the dubious distinction of fully balanced against the potential loss of trust Digital identity applications offer a means probably losing more revenue through tax citizens have in their states regarding their data. The shadow economy – in which goods to stem those losses and lessen the crushing evasion than any other European nation. Little and services are paid for in cash without any pressure governments are under – more so since wonder, then, that it is now taking the most • All operational benefits from public-sector digi- tax collected – is a problem for all countries, but the euro crisis – to realise savings. BCG considers aggressive, and innovative, measures to com- tal identity use will be handed back to consumers particularly within the EU-27, where it averages a reduction of the shadow economy to 10% of bat the problem. Italy’s automated system to through lower taxes or improved public services. 22% of each nation’s GDP (for comparison, the GDP as fully attainable by 2020. detect potential tax fraud – “Redditometro“ – collects data from financial institutions, pub- Retrievable tax revenues of Doing so will take a combination of in- lic agencies and other organisations to infer shadow economy in EU-27 Shadow economy in % of GDP Effective tax rate in % Full potential retrievable tax Share of GDP (%) revenues 2020 in billion € novative measures, such as automated tax col- an individual’s standard of living based on Bulgaria 8.5% lection and the analysis of tax and bank data more than 100 indicators (such as home mort- Romania 6.5% Lithuania 6.8% to identify, and discourage, potential tax eva- gages, school tuition payments, car owner- Estonia 6.9% sion. Already we are seeing some important ship and gym memberships). That standard Latvia 5.7% first steps leveraging personal data. Germany is compared to the citizen’s declared tax Cyprus 7.2% has introduced a standardised tax ID to trace amount (in Italy, filing tax declarations on- Greece 6.2% tax collections more effectively and create more line is mandatory), and instances of potential Malta 6.3% Poland 6.1% transparency, with the goal of making tax fraud fraud are flagged for further investigation. Italy 7.5% more difficult. The EU is currently discussing the Slovenia 6.2% implementation of a similar approach across all Redditometro results in a much more Hungary 6.5% its member states. effective auditing process – focusing not on a 5.0% Portugal random sampling of tax filings, but on those Spain 4.3% Belgium 5.7% Much more will be needed, but the com- most likely to actually involve some degree Sweden 4.7% bination of technological advances and incen- of tax evasion. The system’s success has been Czech Republic 6.9% tives – both increasingly rapidly – will drive dramatic: Italy recovered €11.5 billion in Slovakia 2.5% progress, drawing more tax revenue out of the 2011. This figure is expected to further im- Denmark 3.9% shadow economies. prove over the coming years as the system Finland 3.5% Germany 2.6% is optimised and more data types are added. Ireland 1.9% France 2.4% IBM and Wellpoint – automated medical Netherlands 1.4% decision support system United Kingdom 1.1% Austria 0% Luxembourg 0% Medical care can also benefit from 30% 40% 10% 20% 30% 40% 50% 30 60 90 120 aggregating and interpreting large collec- 10% Ø 22% tions of data. Doctors, after all, make diag- Full Current potential average < 2% 2 - 4% 4 - 6% > 6% noses and develop treatment options not just

Source: Tax Justice Network “The Cost of Tax Abuse,“ 2011; Dr. Friedrich Schneider: “The Shadow Economy in Europe,“ 2009; BCG analysis according to their examination of a patient,

20 In 2011, French government passed a law, enabling it to sell license plate data to commercial companies. The data is currently offered for €0.087 and €0.200 per record according to purchased volume.

83 | 84 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

but on that individual’s medical history, along 6. Web 2.0 communities Global growth of social networks with the latest research on conditions and care. A new system from IBM and Wellpoint • Personal data – created and shared by users – Number of social network users worldwide in millions, 2007-2015 can guide them to the right conclusion by is the primary asset. CAGR 2,000 +10 % finding and analysing the relevant data with 1,800 lightning speed. • The types of content generated by Web 2.0 1,600 communities are wide-ranging and detailed: ~ 25 % of global population 1,400 Based on IBM’s powerful analyti- demographics, employment and education CAGR cal engine, “Watson,” the system uses data data, interests, family and social connec- 1,200 +30 % from various sources – including patient tions, photos, videos, product feedback and 1,000 medical records, treatment and claim his- location data – to name a few. 800 tory, research papers and medical textbooks 600

(all increasingly available in digital form) – to • Data volume and types will increase as so- 400 ~10 % of global population detect likely diseases and injuries and pro- cial media continues its strong growth and 200

vide treatment options. It considers possible mobile applications enable users to spend 0 drug interactions when suggesting medicine more time using these services. 2007 2008 2009 2010 2011 2012 2013 2014 2015 use, and even looks at national databases on Actual Forecast disease outbreaks to inform its judgement. • Targeted marketing and other ways to mon- Precise responses are provided in less than etise the user-generated content are cur- Source: Morgan Stanley research; Strategy Analytics research; eMarketer; UN “World Population Forecast”; BCG analysis three seconds. The result is up-to-date, evi- rently the main applications. dence-based, individualised health care – and All of this can be – and is – used to gen- State of the sector a way to cut costs while improving treatment. • Integration of a social component on other erate consumer insight and target marketing. websites, blogs and services is increasing More data types are coming (imagine paying Personal data is the main asset of social Systems like these do face challenges, and could spur a host of innovative – and with a phone-based digital wallet and having and professional networks like Facebook, LinkedIn however. Providers may be hesitant to invest potentially lucrative – applications. the purchase posted directly on Facebook). Com- and XING. These sites exist, after all, to give users in the technology if only insurance compa- panies already use social media data to focus a platform on which to create and share content – nies benefit from the cost savings. There may Social networks are enjoying rapid – and research efforts, understand customer interests not just details like employment, education and be legal restrictions on using certain types of widespread – growth. In the United Kingdom, and know which discounts to send which users. interests, but photos, video clips, social connec- medical data. And there are medical doctors approximately 60% of the population is on tions and location information. who need to adopt their practices. Facebook; in Iceland the figure is around 70%. There’s also the participatory aspect of The strong growth will continue throughout the the social Web. Dialogues with customers, and Web 2.0 communities are starting to purse world. By 2015, a quarter of the globe’s citizens surveys to gauge interests and gain feedback a range of digital identity applications. Among will have joined a social network. (and even new product ideas) happen in real- them: partnering with other sectors to add a time. That gives customers a potent voice in how social component to existing services; and mone- Social media is already a fertile source of products are developed and supported, it lets tising the insight gleaned from user data. But personal data, including interests, educational companies better serve them and it builds strong- for now, at least, the main focus is on targeted and professional histories, preferences, places er, more lasting, relationships. marketing and other ways to monetise the user- visited, activities performed and milestones in generated content. The profiling of users is not one’s life. only used for targeting advertisements, but is a

85 | 86 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

Facebook penetration Key personal data applications in Europe Web 2.0 communities Prevalent method Facebook users as share of country population, 2012 of data collection

0-10 % Mined (profiled data) 3 – Targeted marketing 10-25 % 25-35 %

35-45 % Tracked 1 – Service enhancements > 45 % (contextual data)

Required 4 – Default identity (structured data) provider

Volunteered 2 – Monetisation of user- (unstructured data) generated content

Delivery of Enhancement Transfer to Transfer to Major use requested of services third party third party of data service (anonymised) (traceable)

Year Value generation Consumer sensitivity

Very low (< 10) High (> 30 – < 35) 2011 2020 Low (> 10 – < 20) Very high (> 35) Small Large Medium (> 20 – < 30)

Source: BCG digital identity survey (n = 3,107, August 2012); BCG analysis

creases the dependency of Web 2.0 communi- Digital identity value Source: Socialbakers, July 2012; BCG analysis ties on advertising revenue. Web 2.0 communities EU-27 in 2011 and 2020, billion € substantial element in improving the user expe- mendations on whom to connect or what to • Targeted marketing: Targeting is especially CAGR rience and increasing usage; for instance, see, listen and read. Thus, the social and pro- important for social networks as otherwise the 39 through recommending new connections and fessional networks strengthen their customers‘ ad inventory would be of very low value for ad- 7 +35% providing more relevant news and updates. loyalty and increase the time spent on the site. vertisers. Facebook, for example, offers a broad range of criteria by which advertisers can target • Monetisation of user-generated content: The consumers, including geography, demographics, Key applications of digital identity insight generated from highly detailed user education level and user-defined “likes” (a po- 32 +19% 9 content – or simply the data itself – can be pular Facebook feature which enables users Organisations 1 • service enhancements: User data is applied to marketed to third parties. A social network for to tag products and other things that interest Consumers 8 develop a personalised product experience for professionals, for example, can generate valu- them). 2011 2020 the network members. Based on profiles and able competitive intelligence for employers and Source: Company reports; EITO; EIU; Gartner; Eurostat; OECD; IDC; Forrester; the derived preferences, users receive recom- recruiters. Monetising data and insight de- IAB; BCG economic value modelling

87 | 88 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

• Default identity provider: Increasingly, social effective as the user base and available data Media usage and ad spend development media accounts are being used as a single grow. Share of media usage and ad spend per channel in %, 2011 U.S. % sign-on to access a wide-range of external 50

sites and services. Using their Facebook login, • Secondary monetisation of data or the insight Time spend 2. Growing usage share for example, users can sign on to a wide range derived from it can provide additional revenue 40 Ad spend of websites and even see who in their social streams.

network has also visited the site and what 30 they may have done there (such as reading or • The use of a social media account as a default 1. Catch-up of ad spend sharing an article). This can greatly extend the identity on the Web and for mobile applica- 20 43 42 reach of the social network. tions increases customer loyalty and retention.

25 28 On the cutting edge – where applications • Consumers gain – at no financial cost – a plat- 10 22 15 11 10 may be more sophisticated and intriguing, but form on which to represent themselves, share 7 their prospects are more uncertain – we have content, communicate with friends and family 0 1 observed the following uses: and network with professional contacts. Print Radio TV PC Mobile Offline Online • Development of more comprehensive mining • The consumer surplus far exceeds the revenue Trend: Growing Steady Declining techniques to generate insight from social that social and professional networks gene-

networks and predict future trends. Fashion rate. According to a BCG report published in Source: Kleiner Perkins Caufield & Byers: “Internet Trends 2012“; BCG analysis designers, for example, have been able to detect 2012, the consumer surplus of Internet serv- nascent apparel trends more accurately based ices – including social networking – averages tration and usage of online media will contribute community widgets, such as “like“ and “share“ on insight inferred from photos, comments and €2,900 annually per capita in Germany.21 to a continued growth in online ad spend. buttons, enabling social networks to expand their location data generated by social networks. reach on the Web and more extensively track As different forms of media gain, and As they pursue new forms of revenue, social their users’ online behaviour. The insight they • Expanding into ad management platforms lose, popularity, advertising spending adjusts, networks might want to take a cue from Google. gain, combined with the targeting capabilities on third-party websites. Increasingly, Web 2.0 following the media usage of consumers. The Leveraging its superior know-how in capturing the they employ on their own sites, could be em- communities are leveraging their user profiles process doesn’t happen overnight, though. Ad- content, and context, of websites, Google has de- ployed to match ads with third-party inventory. to enable targeted advertising on third-party vertisers and publishers need time to discover veloped an advertising business that now accounts sites as well as their own. This is facilitated by the optimal marketing applications and tactics for 96% of its revenue.22 In the process, it has be- Going forward, advertisers and publish- the use of social media accounts as a single for each new media channel. What we are seeing come a key platform for managing ads on third- ers will very likely find themselves with three login and widgets integrated on third-party now is their steady shift from traditional adver- party sites, using its technology to match online main options in targeting online ads: Google’s websites. Users are recognised on the external tising hotspots like print and radio to new ones advertisements with the best-suited inventory. context-based targeting; social networks’ target- site and ads can be targeted accordingly. like online (via desktops and laptops) and mobile ing based on detailed user profiles; and third- (via smartphones and tablets). Social media sites could follow a similar party targeting platforms (like AudienceScience route. While they are already using the user pro- and nugg.ad). For them, a sound strategy is to try The benefits for organisations and consumers While advertising spending has nearly files and personal data they possess to target per- all three and closely follow and track campaigns, caught up with media usage for PC-based online sonalised ads on their own sites, significant po- determining which route delivers the most effi- • Targeted advertising – on its own site and on channels, the “catch-up“ process is just beginning tential lies also on third-party websites. Already, cient results in different settings. third-party sites – will become increasingly for mobile devices. The steadily increasing pene- many of these sites are implementing Web 2.0

21 BCG: “The Internet Economy in the G-20,“ March 2012; BCG‘s “Connected World“ reports 22 Source: Google 2011 annual report

89 | 90 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

7. E-commerce of data. Not only does it influence purchasing Key personal data applications Prevalent method E-commerce decisions, but it is starting to be leveraged in of data collection • Sector already makes extensive use of per- other ways: analysed for insight that can im- sonal data. prove internal processes, or monetised in sale to Mined 5 – Targeted marketing third parties. (profiled data) 3 - Recommendations • Recommendations, the major data-driven application, are responsible for up to 25% The social media boom is creating more Tracked of revenue at best-of-breed sites. customer touch points, and the intensified usage 2 - Consumer insight 4 – Monetisation (contextual data) of consumer insight of mobile apps means more traffic to e-commerce • The most advanced sector in its use of tar- sites. Mobile devices, in particular, hold vast po- geted advertisements. tential for the sector. Location information can Required (structured data) be used to send offers at opportune times (such • The growth of mobile devices and social as when a consumer is in a brick-and-mortar

media sites bodes well for e-commerce store); apps can enable consumers to scan or Volunteered companies, creating more touch points with photograph an in-store item and immediately (unstructured data) 1 – Reviews consumers and more possibilities for inno- receive a better price online. Among the differ- Delivery of Enhancement Transfer to Transfer to Major use vative applications. ent sectors, e-commerce has been, arguably, the requested of services third party third party of data most proactive in leveraging digital data. Advanc- service (anonymised) (traceable) es in technology and the social Web will hand it Year Value generation Consumer sensitivity State of the sector new opportunities to do even more. Very low (< 10) High (> 30 – < 35) 2011 2020 Low (> 10 – < 20) Very high (> 35) The e-commerce sector – which depends Small Large Medium (> 20 – < 30) on personal data to operate – makes pervasive, Key applications of digital identity Source: BCG digital identity survey (n = 3,107, August 2012); BCG analysis and sophisticated, use of digital personal infor- Digital identity value E-commerce mation. It is the most advanced sector regarding • Reviews: By encouraging and publishing user- € the use of targeted ads, and its primary digital created product reviews, e-commerce sites lev- • Recommendations: Analysing a compre- EU-27 in 2011 and 2020, billion identity application – leveraging data to gener- erage the trust customers have in each other’s hensive set of consumer data – such as past 226 CAGR ate recommendations – is already a proven value experiences to drive sales. The more sophisti- purchases, products viewed and other users’ 95 % 45 +20% driver. Indeed, one of the sector’s leaders, Amazon, cated systems allow customers to comment on purchases – e-commerce merchants generate is expected to see up to 25% of its revenue trig- individual reviews, creating a dialogue among relevant, user-specific recommendations that gered by its recommendations system, which steers users that fosters even more trust. increase sales. consumers to items they might like based on pro- files developed from past purchases and other • consumer insight: Usage and purchasing data • Monetisation of consumer insight: The data 93 181 +9% Organisations 9 data. Improving recommendations engines will is continuously analysed to derive insight for generated by or about customers – and the in- continue to be a sector focus, and average perfor- demand planning and offer design. Being able sight derived from it – is often valuable to third Consumers 84 mance will get closer to today’s best in class. to align inventory closely with demand and parties (such as consumer goods manufactur- recognising trends early are critical success ers), and its sale can provide an additional rev- 2011 2020 User-generated content – largely in the factors in the competition within the e-com- enue stream. Source: Company reports; EITO; EIU; Gartner; Eurostat; OECD; IDC; Forrester; form of reviews – is also an important source merce sector and with offline retailers. IAB; BCG economic value modelling

91 | 92 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

• Targeted marketing: E-commerce sites are with some forms of targeting increasing effi- the topics that are being addressed there, • Technical. Here ads are delivering based advanced users of targeted ads, employing ciency by a factor of six.23 and ads matched to those topics are dis- on technical specifications that the user’s them to drive more traffic to their sites and to played. For example, on a blog page discus- browser automatically transmits. These improve conversion rates. • Selling customer insight to third parties in sing sites seen during a trip to Barcelona, can include the used device, Internet con- other sectors opens up a new revenue stream contextual targeting would most probably nectivity speed, clock settings and browser On the cutting edge – where applications for e-commerce companies. result in travel- and tourism-related ads version. For example, video ads would only may be more sophisticated and intriguing, but appearing. be delivered to devices with a fast Internet their prospects are more uncertain – we have • For consumers, e-commerce provides greater connection, or special discounts for dinner observed the following uses: convenience than in-store transactions and • geographical. In this form of targeting, would only be displayed during the evening, enables the identification of the best available an audience is selected by location, which when people are most likely to be thinking • Integrating social features with e-com- price. is derived either from an IP address (typi- about dinner. merce. The rise of Web 2.0 communities has cally indicating country, region and city) or made it possible to incorporate a social aspect • Recommendations generated by vendors guide through data on a GPS-capable device. Ads • f requency capping. This technique requires into purchasing decisions. Used imaginatively, consumers to products that interest them, are then targeted based on that location; measuring how often a user is exposed to these features can help drive sales. Levi’s while user reviews give them confidence in for example, a national website would show a particular ad. The number of times that “Friends Store,” for example, displays products their purchasing decision. ads from a local retailer only to users locat- user is shown the ad can then be restrict- that a visitor’s friends liked – thereby leveraging ed in that retailer’s city. ed – for instance, to four times a day – in the trust consumers have in people they know. order to optimise efficiency and minimise Overview of targeting techniques – used • Registration-based. Leveraging the infor- annoyance. • Extending reach into the offline world. stand-alone or in combination mation a user provided during the regis- Using mobile apps, some e-commerce com- tration process, this form of targeting can panies are finding innovative ways to create • behavioural. In this technique, a user‘s be based on a wide range of data, includ- more touch points with customers and extend browsing behaviour is monitored via cook- ing age, interests and home address. These their reach into the offline world. Amazon, for ies, tracking the websites he visits and the offers and ads can be highly personalised, instance, has developed an app that lets users content he clicks on. From this informa- since they have a lot of personal data to scan products to compare prices. The idea is tion a profile is created. For instance, a user work with. For example, registered users of that a customer standing in a store will see a reading regularly golf articles might be an online retailer may receive a special dis- lower price on their phone and complete their marked as having an interest in sports, an count offer on their birthday. purchase online. interest in golf, being male and earning an above-average income. This example • Retargeted. Retargeting is used to display also demonstrates a key weakness of this ads of products a user has recently viewed The benefits for organisations and consumers technique: If several individuals – family in an online shop but hasn’t bought. For ex- members, for example – are using the same ample, a user might look at sports shoes on • Recommendations influence buying deci- computer, the profile becomes blurred. a fashion retailer‘s website, but then leave sions and drive sales – up to 25% of revenue the site without purchasing them. Retarget- in Amazon’s case. • contextual. This technique does not re- ing techniques will then display ads for that quire personal profiling. Instead, the con- specific pair of shoes at the news or weather • Targeted advertisements remain a key driver tent of a website is analysed to determine site the user visits the next morning. of highly cost-efficient customer acquisition,

23 Criteo claims such improvement over standard banner ads backed by joint study conducted with Nielsen in 2010.

93 | 94 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

8. Online info / entertainment Meanwhile, personalised services – often Key personal data applications Prevalent method Online info / entertainment offered on a subscription basis – and secondary of data collection • Targeted advertising and recommendations monetisation (selling data or insight derived from are the major applications of digital data. it to third parties) both offer growing revenue Mined 4 – Targeted marketing streams that the sector is starting to pursue. The (profiled data) 2 - Recommendations • The increasing use of mobile and Internet- integration of a social component in services – ready devices is expanding the sector’s reach via partnerships with Web 2.0 communities – pre- and the data it generates. sents opportunities for further personalisation, Tracked 1 – Personalised 3 – Monetisation (contextual data) products / services of consumer insight enhanced customer experiences and new data- • As users shift to online media for news and driven applications. entertainment, advertisers are shifting as Required well, enhancing revenue opportunities. Key applications of digital identity (structured data)

• Incorporation of social media into services • Personalised products/services: By using per- Volunteered will enable increased personalisation of serv- sonal data to infer the interests of users, content (unstructured data) ices. can be customised for them. These personalised services can be monetised through subscrip- Delivery of Enhancement Transfer to Transfer to Major use requested of services third party third party of data State of the sector tion fees or advertisements. Spotify, for example, service (anonymised) (traceable)

will individually stream a user’s preferred music Year Value generation Consumer sensitivity

Smartphones, tablets, Internet-ready tele- whenever it is connected to the Internet. Very low (< 10) High (> 30 – < 35) visions and other connected devices have given 2011 2020 Low (> 10 – < 20) Very high (> 35) online information and entertainment serv- • Recommendations: By analysing a variety of Small Large Medium (> 20 – < 30)

ices more reach than ever – and as more users data sources, entertainment choices that accu- Source: BCG digital ientity survey (n = 3,107, August 2012) shift to online media for their news, video and rately reflect a user’s interests can be suggest- Digital identity value music, more advertisers are shifting with them. ed. These recommendations, in turn, can im- tary data to several resellers. Those companies, Online info / entertainment The result for the sector is a confluence of good prove content navigation, increase usage time mainly start-ups, use the data to, among other EU-27 in 2011 and 2020, billion € tidings: more users, generating more data, en- and boost up-selling. A well-known example is things, capture the stock market sentiment. 146 CAGR hancing targeting capabilities sought out by YouTube’s video recommendation system, which more advertisers. Indeed, the major current value generates recommendations based on user pro- • Targeted marketing: The personal data that 95 % 20 +16% derived from personal data is targeted advertising. file, viewing history, ratings of content and other is collected can be leveraged to deliver digital data sources. ads to narrowly defined target groups. Google, The sector’s other key application – espe- for example, collects a variety of data – demo- cially important for entertainment services – is • Monetisation of consumer insight: The data graphics, preferences and browsing history, to 90 126 +4% recommendations. Using systems that analyse that information and entertainment services name a few – in order to profile users. These Organisations 5 different types of data, such as previous selec- generate or collect from their customers can profiles are then used to target advertising tions by a user and selections made by others, be highly valuable to third parties. Monetising from third parties. Consumers 85 highly relevant suggestions can be made. That this information – or the insight derived from

not only improves content navigation but also it – creates an additional revenue source. Case On the cutting edge – where applications 2011 2020 increases usage time and up-selling. in point: Twitter started syndicating its proprie- may be more sophisticated and intriguing, but Source: Company reports; EITO; EIU; Gartner; Eurostat; OECD; IDC; Forrester; IAB; BCG economic value modelling

95 | 96 afoThen vrdalue oconsume ofr gdigiatanisl idenattitry ionss Part 3

Innovative into physical applications, too. Recently, it added Spotify - Integration of social media with personalisation IFTTT WeMo – a family of sensor and switch devices for wide range of privacy options for users home automation – to its list of channels. That

Choose ”Trigger“ and ”Action“ from wide Create ”Recipe“ using ”Ingredients“ “Recipe“ automates everyday life means that now when a new Facebook friend is For entertainment sites, social media of- range of online and offline applications added, a light in the living room can turn on. fers both potential and peril. Features that facil- itate sharing – letting a user’s friends see what But it also means a lot more: a level of he or she is interested in – are a powerful way personalisation that is potentially boundless. to spread the word about a service, reach new Building bridges between the physical and digital audiences and build the user base. But they have worlds won’t just bring convenience to users – to be implemented in a way that builds trust, but revenue to those who do it well. too. Users may not want too much information shared, or shared with everyone. Privacy con- trols have to be available, intuitive and granular. The benefits for organisations and consumers Balancing sharing with privacy is no • Personalisation enhances the value of infor- simple task, but an interesting model for it is mation and entertainment services – realised Spotify, the popular streaming music service. It through either subscription fees or advertising. is tightly integrated with social media; specifi- cally, with Facebook. Users can share, automat- Source: IFTTT • Selling collected and user-generated data to ically, information on songs they are currently their prospects are more uncertain – we have transmit and receive data – creating a vast, and third parties provides an additional source of listening to, or what they have listened to pre- observed the following uses: rapidly expanding, Internet of things. This opens revenue. viously. Their favourite tunes and playlists can up a wealth of possibilities, only some of which be accessed, too, by Facebook friends. This pro- • Incorporating social media features in or- we are starting to realise now. • Customers benefit from access to vast content vides value to users, who get recommendations der to further personalise media consump- offerings at no, or for little, charge. Online info from people they trust – and to Spotify, which tion. By now, many websites incorporate basic One such application is personalised au- and entertainment services represent a major expands its reach. social features such as “like“ and “share“ wid- tomation that incorporates, and links, both Web share of the large consumer surplus of Internet gets. Future differentiation depends on deeper services and “real world“ components. IFTTT – “If services.24 But what makes that value sustainable and more natural ways of making info and en- This Then That” – is one interesting example of is the way privacy is addressed. Spotify offers a tertainment services more social. For instance, this kind of “blended“ automation that serves as • Sophisticated, accurate recommendations range of options that give users very person- Spotify incorporates features that allow users an outlook of things to come. Its premise is sim- systems allow users to more easily navigate alised, and very detailed, control over sharing. to access the playlists and favourite songs of ple: Users set a trigger (the “this“) that causes an services and find content they like. They can choose what playlists and favourites their friends, and let them share their own mu- action (the “that“) to be taken. But the twist is to share; whom to share with; and how much sic selections automatically on social media that the trigger-action combinations – or recipes, information gets posted. They can choose to sites. as IFTTT calls them – can mix different services. enjoy private listening sessions, too, when So if a new Facebook friend is added (the “this“), they don’t want their 600 best friends knowing The world around us isn’t just growing but a tweet about it can automatically be made (the just how much they enjoy Broadway musicals. becoming more interconnected. Today’s commu- “that“). IFTTT’s first step was to integrate differ- Spotify’s savvy use of social media features has nications technologies enable physical objects to ent Internet channels, but now it is expanding helped spur a rapid growth in its subscribers.

24 BCG: “The Internet Economy in the G-20,“ March 2012; BCG‘s “Connected World“ reports

97 | 98 part 4 The megatrends of digital identity P499 TheofThe meg d megigiatrentadatrsl idenendtsi ty Part 4

The megatrends of digital identity

A growth driver for a stagnant economy applications will be unveiled over the next dec- identity intensity – the estimated share of digital ticated and vast use of personal data. Growth in ade. But one fact is clear: The opportunities digital identity value of the overall sector size – as the those sectors will be spurred mainly by an ongoing As detailed earlier in this report, many of identity offers are enormous. It can drive massive application of personal data will play a more and increase in Internet penetration of and usage by the most common applications of personal data growth in an otherwise stagnant European econ- more important role in those sectors. This ratio is the European population. Although the extent to are currently only starting to tap their potential omy, and the value it provides can amount to expected to remain relatively stable for Internet which digital identity is applied in the retail sec- value. It is still too early to tell which of the roughly 8% of EU-27 GDP by 2020 (see Exhibit companies because they already make sophisti- tor is at a similar level to that of Internet compa- emerging uses will succeed and what additional below). cated and vast use of personal data. Growth in nies, this sector is expected to have only a modest those sectors will be spurred mainly by an on- growth rate because the retail sector will not grow Digital identity value EU-27 in 2011 and 2020, billion € EU-27 in 2011, billion € going increase in Internet penetration of and as dynamically. key driver of overall usage by the European population. Although the economy extent to which digital identity is applied in the The benefits of digital identity are relevant retail sector is at a similar level to that of Internet and important for consumers as well. Indeed, ~ 8% of 997 GDP companies because they already make sophis- we estimate that the consumer benefit will be

+14% ~ 2x Digital identity value 328 budget defícit Consumer value of digital identity 2020, in billion € for consumers 6 1 3 1 7 1

unquantifiable 2nd 315 order effects 12,629 41 Organisations 53 669 520 23 24 97 Consumers 262 Further 213 61% potential

2011 2020 GDP Budget deficit Digital identity value

85 Source: OECD; BCG economic value modelling 22 8 84 Although the public debate on mone- sectors with strong growth prospects. Companies Current 39% value 2011 54 tisation of personal data is dominated by the from the manufacturing, financial services, and

perspectives of Web 2.0 players and the digital telecommunications and media industries are % of value 20% 40% 60% 80% 100% economy, digital identity is not just the realm expected to achieve digital identity value growth of the Facebooks and Googles of the world. It is rates comparable to those of the Internet sec-

relevant, and important, for the economy as a tor. This is due to a significant increase in digital 1

whole. In fact, it is the public sector and health Traditional Retail Financial Telco & Public / Web 2.0 E-com. Online info / care industry that stand to profit the most from Public sector and health care production services media health comm. entertain. personal data applications – potentially realis- industry stand to profit the most 25% +1% +8% +17% +19% +17% +9% +4%

ing 40% of the total organisational benefit (as from personal data applications – 1 Rounded values Source: Company reports; EITO; EIU; Gartner; Eurostat; OECD; IDC; Forrester; IAB; BCG economic value modelling detailed in Exhibit Digital identity value for organ- potentially realising 40% of the total G rowth C A R 2011–2020 isations). But these are not the only traditional organisational benefit.

1 See appendix “Development of digital identity intensity per sector“ for details

101 | 102 TheofThe meg d megigiatrentadatrsl idenendtsi ty Part 4

Digital identity value Overview of quantifi- for organisations able and unquantifi- Private- and public-sector value of digital identity 2020, billion € able benefits of digital Stakeholders identity applications

24 6 Consumers Private sector Public sector Further 14 112 84% 15 potential 52 34 36 Cost reduction • Lower prices because of • More efficient ads • Less manual transactions Quantifiable handover of private / public sector savings Revenue increase 7 • Time savings • Personalised products • Reduced tax evasion

16% Better life

Current B enefits 5 1 5 • Longer, healthier life Second-order effects value 2011 11 7 9 Unquantifiable 8 • More convenient use • Longer living, healthier people spend more % of value 20% 40% 60% 80% 100% Applications not known today

they will see thanks to applications using person- goods and services and for other purposes. They 1

Traditional Retail Financial Telco & Public / Web 2.0 E-com. Online info / al data. Second-order effects will add to the to- often tap into the wider ecosystem of data-relat- production services media health comm. entertain. tal, too: People who live longer lives, for example, ed opportunities as well, such as sharing – and 22% +7% +21% +16% +35% +24% +20% +16% will spend more money. Finally, there are all the selling – information to third parties. as yet unknown applications that will bear fruit 1 Rounded values Source: Company reports; EITO; EIU; Gartner; Eurostat; OECD; IDC; Forrester; IAB; BCG economic value modelling as they inevitably enter the scene over the course Companies in the telecommunications G rowth C A R 2011–2020 of the next decade. and media sector, including cable operators, have access to a vast amount of personal data in prin- more than double the organisational value – €670 Germany.2 The equally large consumer value in the Sectors are at different stages of generating ciple – mainly relating to telephony and Internet billion a year by 2020 – mainly stemming from public and health care sectors results from the value from personal data usage or TV viewing – but so far make relatively reduced prices (passed on by companies seeing fact that most of the benefits achieved by those little use of it. While customer self-service and data-driven cost savings), the time savings that organisations will be passed on to citizens in the Part 3 shows in great detail how the var- process automation are currently the major ap- self-service and automated transactions will bring form of tax reductions or other savings.3 ious sectors make very different use of personal plications of digital identity, we expect the focus and the high value individuals place on free online data. We summarise the stage a sector or single to shift to innovative new services and enhanced services, supported at least in part by the use of The total benefit of digital identity will organisation is at in terms of its use of digital user experience. personal data (as shown in Exhibit Digital identity likely be even higher than the €1 trillion in identity with the path depicted in Exhibit Evolu- value for consumers). The latter in particular is a quantifiable benefits we anticipate by 2020. tionary path of digital identity value creation. The traditional production and public sec- source of tremendous consumer value created by As shown in Exhibit Overview of quantifiable tors, meanwhile, are only at the very beginning of Internet companies. The consumer surplus of In- and unquantifiable benefits of digital identity Retail and the three Internet sectors are the path. Many of these organisations are only ternet services – including social networking, on- applications, various unquantifiable benefits will already at a relatively advanced level; these or- now starting to create a basic digital product ex- line shopping, e-mail and online news – averages be created as well, such as the longer lives indi- ganisations routinely leverage personal data in- perience or embrace digitisation – such as govern- for instance about €2,900 annually per capita in viduals may enjoy and the greater conveniences ternally for R&D to optimise the delivery of their ment agencies and health care providers migrating

2 BCG: “The Internet Economy in the G-20,“ March 2012; BCG‘s “Connected World“ reports 3 One notable exception is reduced tax evasion, which increases the fairness of taxation, but does not result in monetary benefits for citizens overall. 103 | 104 TheofThe meg d megigiatrentadatrsl idenendtsi ty Part 4

Evolutionary path of Internet This is just the beginning. Emerging technol- • Personalisation. Customising products and Value Retail digital identity value sectors ogies like near-field communication – where services to a particular individual‘s preferenc- creation Financial services users authorise payment simply by placing es and needs makes them more relevant to

Telco & their smartphone near a reader – will spur all that person. That spurs sales for the organi- III media manner of automated processes. sation and satisfaction for the consumer. Tra- Tapping data ecosystem opportunities ditionally, personalised products came about II I Leveraging personal • User enablement. Digital identity gives in- through bespoke solutions at high cost pre- Creating basic digital data internally product experience dividuals the ability to perform transactions miums and for small target audiences. Digital External applications autonomously without any human assistance. identity enables personalisation on a widesca- Traditional Sharing data with third parties Customer self-service is a prime example. It le, cost-effective basis (due to the availability production Internal enhancement in both directions continues to grow strongly in sectors where it of more data and improved processing power). Public Usage data for R&D, delivery sector has been applied for decades – particularly in optimisation etc. Digitisation banking and the telecommunications industry – Online retailers leverage a user‘s purchase and Basic digital identity uses as and is being introduced in more and more browsing history to generate personalised rec- secure authentication new industries. Its popularity is easy to under- ommendations and custom storefronts – fea- Time stand: Consumers save time and effort, and turing products relevant to each individual con-

Digital identity intensity organisations save money by replacing man- sumer. When executed well, these applications ual labour with IT solutions – and as labour can create great value: Amazon‘s recommen- The traditional production and costs increase and IT costs drop, the savings dations, for example, are estimated to generate public sectors are at the very begin- automate – and simplify and speed up – trans- become even more significant. By passing on some 25% of the e-commerce giant‘s sales. ning of the path, only now starting to actions that traditionally required users to some of these savings in the form of redu- create a basic digital product experi- identify themselves, express their preferenc- ced prices, companies can also leverage user More traditional sectors, too, are discovering ence or embrace digitisation. es or perform some kind of manual action. enablement to gain a competitive advantage. that data-driven personalisation can drive new offerings and enhance the appeal of those that to electronic records and setting up processes that One of the most common applications of auto- These types of applications are becoming a already exist. Retailers can deliver coupons and leverage this data along the value chain. mation is the online ordering process used by e- focal point of the public sector, as citizen self- offers by e-mail or mobile based on a customer‘s commerce sites. Billing and shipping information service portals can simultaneously reduce past purchases, browsing history or even their The six key trends in digital identity is collected once and stored; it is then used to pre- budget pressure and increase efficiency. current location (the latter is made possible applications fill forms for subsequent purchases. This not only Denmark, for example, has one of the most using the GPS capabilities of smartphones). saves consumers time, but also prevents errors. advanced e-government programmes in the While the actual applications of personal world and generates €200 million in annu- • Enhanced delivery. The availability of both information vary from sector to sector, and even Increasingly, process automation can be found in al savings with electronic invoicing alone.4 more volume and more types of data enables from organisation to organisation, they invariably the offline world as well. For example, the Oyster organisations to increasingly extract operation- fall into one of six categories. These are the me- Card carried by London commuters authentica- The most successful applications all share one al insight from it. This allows them to improve gatrends of digital identity that are relevant for tes them when they board buses and trains, auto- vital trait: They provide a simple and intuitive the delivery of goods and services. For instance, organisations at each stage of the path: mating ticketing and access control. Meanwhile, user experience. As Jeff Bezos, founder and CEO IBM and the U.S. insurer Wellpoint are devel- many products can store and apply user prefer- of Amazon, puts it: “Every time a customer con- oping a system to analyse huge volumes of • Process automation. By integrating personal ences – like an automobile that automatically tacts us, we see it as a defect.“5 data from many sources – including research data into their processes, organisations can adjusts the seat and mirrors to each driver‘s liking. papers, medical textbooks and population health

4 Source: Case study “eInvoicing in Denmark,“ 2007; press 5 Wired Magazine online, 13 November 2011

105 | 106 TheofThe meg d megigiatrentadatrsl idenendtsi ty Part 4

data – to detect diseases and suggest treatment relevant for any organisation that has access Key trends of digital identity value creation options. Such solutions can lead to many to a proprietary data set and user consent for Key trends Description more patients benefiting from optimal care. third-party application of the personal in- Applying digital identity of users to automate processes as users do not have to identify formation or sufficiently anonymized data. Process automation themselves and preferences are known as part of the profile Different industries are at different stages in

leveraging data to enhance delivery, and not For instance, in the online sector, Twitter recently Empowering users to perform transactions autonomously without any need for a human User enablement surprisingly, the online sector is currently ahead. sold access to the billions of tweets its users openly counterpart But even in the public sector, the opportunities publish through the service. Third-party compa- are rich: Applications are starting to emerge nies syndicating this data use it, among other Providing more relevant products, services and advertisements based on knowledge Personalisation about consumers that is available and processed in real-time that can fight tax evasion and fraud in govern- things, to capture the stock market sentiment. ment programmes. Health care, too, stands to Adding of new data sources to profiles and applying of big data capabilities to extract new reap substantial benefits from enhanced deliv- Even the public sector has tried to monetise its Enhanced delivery insights to improve delivery of products and services ery, with the insight gleaned from vast data sets personal data. In the United States, the state of leading to more effective patient treatment. Oklahoma realises about $13 million annually by Personal data- Using new and combined data sources and big data analytics to increase R&D efficiency selling information it collects from drivers. This driven R&D and effectiveness • Personal data-driven R&D. Using a range of and similar European examples of secondary

data sources – such as product reviews, com- monetisation raise questions about the balance Secondary Exploring ways to sell personal data in anonymised or fully traceable form to third parties monetisation ments posted on social media sites and usage of the expected return and the potential loss of data from in-product sensors – companies can citizens‘ trust. Is it worth it to market such data also gain insights that let them better focus without the explicit consent of the individuals their research efforts and shorten development for a few million euro when there are applica- Relative importance of cycles. This trend has already been adopted by tions of personal data worth billions of euro that digital identity trends major consumer goods companies, including require citizens‘ trust and collaboration? per sector Procter & Gamble, whose Vocalpoint P&G com-

munity gets customers more deeply involved in So how can organisations and policy mak- Traditional Retail Financial Telco & Public / Web 2.0 E-com. Online info / the development and testing of new products. ers unlock the value potential of digital identity? In production services media health comm. entertain. the following part 5, we present a new paradigm Using personal data to drive R&D is a particu- for responsible application of personal data – Process automation larly compelling proposition for traditional in- in a sustainable, consumer-centred way – and

dustries where, historically, little usage data has offer guiding principles for key stakeholders. User enablement been available and consumer research can be expensive – much more so than in other sectors Is it worth it to market such data Personalisation such as the Internet and software industries. without the explicit consent of the in- dividuals for a few million euro when Enhanced delivery • Secondary monetisation. The data that an there are applications of personal data organisation holds and the insight derived worth billions of euro that require citi- from it is valuable to other parties – valu- Personal data- zens‘ trust and collaboration? driven R&D able enough that marketing it can open up a significant new revenue stream. This is highly Secondary monetisation

6 Source: NIC Inc. website; Oklahoma State Budget; Oklahoma Tax Commission website Relative value generation potential Low Medium High 7 In 2011, the French government passed a law enabling it to sell license plate data to commercial companies. The data is currently offered for a price of €0.087 to €0.200 per record, depending on the purchased volume. 107 | 108 Part 5

part 5 A practitioner‘s guide to the future of digital identity

P5109 | 110 to the future ofA pra dcigititionetalr‘s i dguiende tity

Part 5 A practitioner‘s guide to the future of digital identity

As detailed in the previous parts of the re- BCG estimates that two-thirds responsibly, ensuring an adequately high level of petitive advantage. Indeed, such a scenario may port, the opportunities presented by digital iden- of the potential digital identity security. Next, they should act transparently. By well play out in the desktop and mobile Web tity – when managed carefully – are enormous. value – or about €440 billion in 2020 being open and clear about what they are doing browsing market. When Microsoft announced But so, too, are the lost opportunities, should it alone – is at risk if stakeholders fail with private information, organisations limit the that its next version of Internet Explorer would not be handled in a thoughtful, balanced way. to establish a trusted flow of person- risks presented by hidden practices that, once un- enable a “do not track“ feature by default – so Our research shows that the latter scenario is all covered, can become the focal point of a public websites will know that a user does not toler- al data. too possible. Today, consumers have a generally firestorm. Just as importantly, transparency is crit- ate cookies or any other tracking features – the low awareness of how their personal information ical to communicate to consumers the benefits of move was clearly intended to lure customers is being used; if that awareness should rise, it will BCG estimates that two-thirds of the po- digital identity applications. Finally, there must be away from competing browsers (like Firefox, likely result in a decreased willingness to share tential digital identity value – or about €440 bil- user control. As evidenced by the consumer survey Opera and Chrome) that are not expected to fol- data – all else being equal. On the other hand, lion in 2020 alone – is at risk if stakeholders fail commissioned for this report, individuals‘ prefer- low suit. Apple‘s Safari had already been config- easy-to-use privacy options and controls can to establish a trusted flow of personal data. All ences regarding privacy and data sharing differ – ured to block third-party cookies by default for significantly increase that willingness to share. sectors face this risk, although Web 2.0 commu- and differ widely. Options and controls allow in- several years. Now, with Microsoft‘s initiative, Anyone seeking to create value through digital nities have the highest exposure, being strongly dividuals to adapt their sharing to their specific two camps, each with about 50% market share, identity needs to understand the implications of dependent on very sensitive digital identity ap- needs. While this might reduce sharing by some are forming in both the desktop and the mobile that dynamic. If consumers‘ awareness of data plications. Nor is it digital identity value alone – individuals in certain situations, overall it will like- Web browsing market, with their approaches to collection and use increases without organisa- the additional revenues or efficiency gains de- ly increase the sharing of personal data. privacy being key differentiators. tions establishing practices they approve of, a rived from personal data applications – that is major part of that value will be lost. at stake: Missteps in handling consumers‘ data Another area where privacy and person- can go much further, causing damage to an Privacy is increasingly becoming an area of al data management may likely be an important organisation‘s brand, its client relationships and competitive differentiation competitive dimension is the mobile operating Digital identity value at risk its reputation. system space. Worldwide, more than 850 mil-

EU-27 in 2011 and 2020, billion € By providing the right privacy controls lion smartphones will run either Apple‘s iOS Creating trust means that certain prac- and options, organisations can also gain a com- or Google‘s Android by the end of 2012.1 Both 682 443 tices, still applied by some companies today, need 138 to be reconsidered – practices like keeping the 275 methods of collection and use hidden; using weak -65% Internet browser 306 data encryption and careless employee data ac- market shares 553 cess schemes; offering consumers few, if any, pri- Global desktop market shares (%), July 2012 Global mobile market shares (%), July 2012 407 vacy controls; and giving them a choice to “take 190 315 it or leave it,“ which often means no choice at 53 all. Personal data is a valuable asset for consum- Other ers and must be treated with care, in ways that Other 262 364 inspire trust, not break it. 8 36 29 23 2 46 22 14 18 Digital Growth Value Remaining To unlock the full value potential, organi- identity value potential at risk value 2011 2020 sations need to embrace a new paradigm for dig- 0 20 40 60 80 100 0 20 40 60 80 100 ital identity applications, one that rests on three Source: Company reports; EITO; EIU; Gartner; Eurostat; OECD; IDC; Forrester; IAB; BCG economic value modelling key pillars: First, they must handle personal data Source: NetApplications; StatCounter; W3Counter; BCG analysis

1 Source: Gartner (July 2012); BCG analysis

111 | 112 to the future ofA pra dcigititionetalr‘s i dguiende tity

Part 5

platforms enable third-party apps to track highly has won impressive funding from several pre- won‘t abuse their data) to “Class E“ (terms three elements are essential to ensure a sustain- sensitive personal data like location information. mier venture capital funds, indicating the po- of service that raise serious questions). One able flow of data: First, the benefit a consum- This model has already raised privacy concerns; tential of this business model. Its most recent challenge ToS;DR will need to resolve is how er receives has to exceed the “cost“ of sharing consumers and policy makers alike are worried financing round, in July 2011, brought in $41 to make its ratings widely visible to consum- the data. Second, there needs to be transpar- about data misuse by apps, particularly as the million, bringing total funding to $68 million.2 ers, as “Class E“ websites will obviously not ency on how the data is used. Third, privacy number of available apps continues to grow rap- publish their low ranking, and requiring users controls should be available and easy to use. idly. Currently, neither Apple nor Google provide • qiy.com. A Dutch initiative, Qiy is a platform to research ToS;DR‘s own site would seriously Dialogue and interaction with consumers – in much in the way of user control. How they that aims to connect individuals with partici- limit its reach. an ongoing process – is critical to getting these address the issue – whether they give users tools pating organisations so that the exchange of elements right. to manage their digital identities within the app personal data is better controlled by the indi- But it is not just start-ups and novel ecosystem, or take no action – will have impor- vidual. The idea is that people should decide initiatives that are working on ways to boost • Take accountability for a trusted flow of tant implications in terms of both privacy and what information goes out, and to whom. To responsibility, transparency and user control. data. Every business and public agency that competitiveness. this end, Qiy is working to establish a trusted Established companies – including some of depends on digital identity applications should framework, centred around the individual, for the digital world‘s biggest and most influential define guiding principles for how it works managing access and usage rights and hand- names – are taking steps, too. Google, for ex- with personal data. These principles should be Innovative approaches to building trust ling personal information. Further, Qiy intends ample, offers users an “Ads Preferences Man- communicated clearly both within and out- to give users access to all of the personal data ager“ that shows which interests behavioural side the organisation to ensure unmistakable However, it is not yet clear what optimal about them that is stored by the organisa- targeting has derived about them – interests that accountability and commitment. At the same transparency and controls for consumers will tions taking part. are applied to deliver online advertisements that time, an organisational structure and processes ultimately look like for each digital identity ap- are supposed to be more relevant for the user. that hardwire compliance with those principles plication. Intense experimentation is currently • tos-dr.info. Terms of Service; Didn‘t Read Through Google‘s tool, users not only see what should be established. being undertaken among established as well as (ToS;DR) is a not-for-profit project created to interests have been stored about them, but can very young organisations to find solutions that enhance transparency for consumers. Terms change them or opt out entirely. • Increase data security in order to safeguard can boost consumers‘ trust – and their confi- of service and privacy policies are documents digital identity. As the examples in this report dence in sharing their data. Below are some of in which websites disclose how they collect demonstrate, data breaches cause damage on the most innovative examples: and apply the personal data of their users. Guiding principles to unlock the value multiple fronts: Consumers see their personal These documents tend not to be read. They of digital identity information compromised; organisations see • reputation.com. This US-based company‘s are lengthy and often written in hard-to- trust erode and their brand and reputation suf- mission is to help individuals and small busi- understand language packed with legal terms. Of course, the challenge now is for or- fer. There are the direct costs, as well, to both nesses monitor their online reputations and So individuals usually have no idea what a ganisations to adopt the new paradigm. To help the consumer and the organisation. The level manage how they are publicly represented site actually does with their data. ToS;DR in- guide them on their way, we have developed the of security – not just technical measures, but on the Web. It continuously scours the Web tends to change that by analysing the terms following principles: also policies and access control – must reflect for a client‘s personal data and will remove it of service and privacy statement and pre- the high value of personal data. from people-search sites. The service employs senting them in plain, easily comprehensible • Engage customers for sustainable personal “reputation advisors“ who work with users to language. Each website is assigned a rating – data usage. Organisations should involve and • Build a data-driven organisation, not just a remove or suppress any negative content based on its policies regarding anonymity, educate consumers on personal data applica- data-driven IT department. Consumers aren‘t about them that shows up high in Inter- tracking, cookies and other privacy-related cri- tions, as well as on privacy management. And the only ones who benefit from transparency. net search results. While reputation.com‘s teria – ranging from “Class A“ (terms of service they should do so in a carefully thought-out, Organisations do, too. They should know what subscriber base is still limited, the company that treat users fairly, respect their rights and well-structured way. As we show in this report, data is available both inside and outside their

1 Source: http://www.crunchbase.com/company/reputation-com

113 | 114 to the future ofA pra dcigititionetalr‘s i dguiende tity

Part 5

own walls, and all parts of the organisation Once again, balance is key. Privacy protec- The most successful policies will put the A balanced approach, both in policy and should be encouraged to identify value crea- tions shouldn‘t be so stringent that they discour- consumer front and centre, but also enable exper- in the applications themselves, is essential to tion opportunities, experiment with new appli- age innovation, nor should they be so weak that imentation in, and development of, new uses of realise the full potential digital identity has to cations and develop analytical talent. they enable misuse. It also requires sometimes personal data. They will be firm regarding prin- offer. This challenge is not limited to Web 2.0 taking a step back from some of the emotion that ciples, but will allow for flexibility in implemen- communities or the digital economy, but is highly While safeguards are needed, can imbue discussions on privacy. Politicised and tation and enforcement. They will give organisa- relevant for the economy as a whole. Putting the the regulations that ultimately result hyped concerns about the privacy implications tions a certain amount of leeway, acknowledging consumer at the centre of the discussion is the should be flexible enough that con- of new applications – concerns voiced recently that different markets and regions have different right way to address digital identity. The guid- sumers can make their own informed in the case of Google‘s “Street View“ feature – historical, cultural and political legacies – and ing principles for capturing the benefits of per- choices about whether and how they can morph into widespread outcry as it did in different, constantly shifting attitudes towards sonal data applications should be based on how share data. Germany. That, in turn, could trigger blanket privacy and new uses of personal data. individuals view their digital identity and what prohibitions on certain data uses – which might influences their decision making. There is a lot at Striking the right balance in policy and reg- impact a wide range of applications beyond the stake here. Digital identity isn‘t just about data ulation targeted one. Ultimately, while well intended and and opportunities and growth. It will impact our designed to protect individuals, such prohibitions lives and our society. It will change the way we Across Europe, policy makers are ac- can actually hamper or delay innovations that interact with the world around us. It will shape tively discussing proposed rules on privacy. would benefit them – and would put domestic our future. While safeguards are needed, the regulations industries at a competitive disadvantage relative that ultimately result should be flexible enough to their international competition. Balance also that consumers can make their own informed means that sometimes a “highest“ level of pro- choices about whether and how they share tection that stalls novel new applications is less data. Individuals should be able to gen- desirable than a “high“ level that spurs them. erate value from their digital identity – if they wish to do so. At the same time, they should be able to rest assured that their data isn‘t being put to unintended or undesired purposes.

115 | 116 Appendix

appendix

A117 | 118 APPENDIX APPAPPENENDIX DIX

Appendix The four exhibits on the following pages present additional results of the consumer research referred to in part 2 of this report, conducted exclusively for this study.

Assessment of privacy level Importance of reasons per data type How private do you consider the following data types? How important are the following reasons for you to disclose your private data? to disclose private data % 100 Very private – 89% 80% 63% 53% 28% 22% 27% 32% 24% 25% 10% 9% preferrably no % 100 6% 11% 14% 15% 15% 27% 28% 30% 32% 37% 40% 51% Not at all important 48% 40% one to know 80 7% 12% 63% 23% 13% 17% 53% 50% Somewhat private – 80 15% 58% 53% 48% only certain people 32% 32% 20% 60 or organisations 21% 31% 33% 21% 21% 60 51% 22% 25% 51% 41% 30% 40 28% Unimportant 42% 31% 30% 22% 32% 40 Not private – 39% 35% 34% 26% everyone can know 30% 25% 20 25% Neither important 18% 20% 20% 22% 20% 20 nor unimportant 10% 15% 20% 20% 13% 16% 16% 1% 2% 5% 6% 12% Important 0 9% 6% Age Opinion Name Interests Past Purchase Media Location Social Health Financial Credit card 12% 6% 6% 5% 5% 3% 3% 2% 2% 2% 2% 1% Extremely important group, on and purchases plans usage network info data data 0 To use To receive To have To save To register To get To get To share my To receive To get To get To partic- gender products e-mail posts online discounts the chance time at products a service access to thoughts targeted added more ipate in a shopping to win a the next to receive for free a paid ads by more comments, birthday par- price or visit of the better service people as “likes,“ ty via video Low sensitivity Medium sensitivity High sensitivity money website support “a friend“ messages telephony from others

Note: Numbers may not sum to total due to rounding. Note: Numbers may not sum to total due to rounding. Source: BCG digital identity survey (n = 3,107, August 2012) Source: BCG digital identity survey (n = 3,107, August 2012)

Awareness of data collection Individuals’ approval

by sector Are you aware of any of the following organisations collecting data you consider “very private” or “somewhat private”? Do you approve the following uses of private data by organisations? of data uses % 100 % 100 79% 77% 72% 65% 63% 58% 57% 50% 47% 32% 31% 18% 23% 33% 48% Yes 80 80 13% Strongly disapprove 20% 48% 41% 27% 21% 60 60 30% 35% No 31% 22% 28% 40 25% 40 2≠5% Disapprove 36% 21% 22% 14% 28% 17% 20 13% 20 22% Neither approve 11% 20% 20% 17% 18% Don‘t know 18% nor disapprove 14% 14% 14% 15% 9% 11% 11% Approve 0 6% 3% 3% 2% Strongly approve Banks Health Public Social Mobile Loyalty cards Online Cable Online search Retailers Car manu- 0 They use the data to deliver They use the data for tailored They allow other companies They allow other companies to insurances agencies networking phone shops network engines facturers a product or service. marketing purposes. to use the data anonymised. use the data linked to your name. sites operators providers

Note: Numbers may not sum to total due to rounding. Note: Numbers may not sum to total due to rounding. Source: BCG digital identity survey (n = 3,107, August 2012) Source: BCG digital identity survey (n = 3,107, August 2012)

119 | 120 APPENDIX APPAPPENENDIX DIX

Appendix The exhibit below provides an overview of how we assigned the „digital identity value share“ – the estimated share of digital identity value of an application‘s overall value generation – for each use case to reflect the particular importance of personal data to that application. The exhibits on the following pages depict the resulting digital identity intensity by sector and show how we defined the sector size for the purposes of this study.

Application of digital identity Development of digital Digital identity value share Assessment per digital identity value share value shares in quantification identity intensity model per sector Very low = 20% Digital identity is minor value contributor and majority of value add is provided by other elements Sector size 2020 in billion€€ Digital identity value for Digital identity organisations 2020 in billion€€ intensity Example: Digital identity only used for secure authentication and payment in e-commerce, 2011 2020 but major value represented by purchased goods 1 Traditional production 2,980 1-2% 63 20-24% 0% 2%

Low = 40% Personal data contributes value to product/service, though other elements responsible for most value contribution 2 Retail 1,070 2-3% 13 5-9% 1% 1%

Example: Binding customers with rewards is major goal of loyalty programmes, but insight generated from personal data for demand planning and optimisation 3 Financial services 1,420 3-4% 42 19-23% 1% 3%

4 Medium = 60% Personal data is important value contributor and product/service can not be delivered Telco & media 270 -1-1% 19 14-18% 2% 7% without it, other elements remain important

5 Public services / health 8,470 2-4% 120 33-37% 0% 1% Example: Personalised medicine not possible without personal health data; but IT capabilities, R&D and medical treatment also important within process 6 Web 2.0 communities 10 20-25% 7 22-26% 71% 72% High = 80% Personal data is key enabler for value generation and delivers most value of service/product 7 E-commerce 310 10-15% 44 18-22% 9% 14% Example: In social networks, personal data is the product and key for monetisation via adver- tisements through third-party applications, other apps such as games contribute small share of value 8 Online info / entertainment 100 10-15% 20 14-18% 17% 20%

Very high = 95% Personal data represents (almost) entire value of product/service, remainder of value retrieved from minor supportive activities CAGR (%) Digital identity intensity CAGR ‘11-‘20 ‘11-‘20 > 15% 5-15% 0-5% Example: Companies selling contact details of consumers in various demographic and behavioural segmentations based on acquired permissions Note: Digital Identity intensity = estimated share of digital identity value of overall sector size Source: Company reports; EITO; EIU; Gartner; Eurostat; OECD; IDC; Forrester; IAB; BCG economic value modelling

121 | 122 APPENDIX APPAPPENENDIX DIX

Sector size definition to assure Liberty Global commissioned The Boston Consulting Group to author a study on the topic of digital comparability of digital identity identity in the context of the European economy. The objective of this work is to contribute to a debate intensity across sectors Sector size 2011 CAGR Sector size Sources currently high on the agenda of industry, policy and regulation with a quantitative angle on the value in billion € ‘11-’20 definition of personal data applications, empirical evidence on consumers‘ actual data-sharing decision making and a holistic perspective. The study reflects BCG‘s thoughts on the topic of digital identity, supported Total sector Euromonitor 2,700 1-2% revenues minus by industry analyses as well as case studies and company examples based on publicly available informa- 1 Traditional production International, EIU retail gross margin1 tion. In the process of writing the study, over 15 European industry managers and policy makers were Euromonitor, 2 850 2-3% Retail gross margin1 interviewed whose expert contribution is reflected in this work. The study provides a basis for discussion Retail “Einzelhandel” for key stakeholders across public and private sectors on a broad set of topics related to digital identity Revenues of retail BCG analysis, 3 Financial services 1,050 3-4% banking, asset manage- Cerulli Associates, developments and future strategic, policy and regulatory priorities. ment and insurances SwissRe

EITO Industry 4 Telco & media 280 -1-1% Total sector revenues For further information or additional perspectives, please contact: Report Total government2 5 Public services / health 6,630 2-4% and health care Eurostat John Rose Olaf Rehse expenditures Senior Partner & Managing Director Partner & Managing Director 6 Web 2.0 communities 1 20-25% Total sector revenues BCG analysis BCG New York BCG Düsseldorf [email protected] [email protected] Forrester research 2011, E-commerce 100 10-15% Total sector revenues 7 Eurostat Björn Röber 8 Online info / entertainment 30 10-15% Total sector revenues BCG analysis Project Leader BCG Düsseldorf [email protected] Note: Digital Identity intensity = estimated share of digital identity value of overall sector size 1 In order to make retail sector size comparable, only the retail gross margin is considered to account for the high share of purchasing costs; this margin is then deducted from traditional production (sector from which retail purchases goods) to avoid double counting 2 Including intermediate consumption; employee compensation, subsidies, consolidated property income, other consolidated current transfers, social benefits, consolidated capital transfers, gross capital formation

123 November 2012 Published by Liberty Global, Inc. with permission of The Boston Consulting Group, Inc. 2012. All rights reserved.