Something new under the sun A special report on innovation l October 13th 2007
Republication, copying or redistribution by any means is expressly prohibited without the prior written permission of The Economist The Economist October 13th 2007 A special report on innovation 1
Something new under the sun Also in this section
Revving up How globalisation and information technology are spurring faster innovation. Page 3
Can dinosaurs dance? Responding to the Asian challenge. Page 4
A dark art no more Like management methods before it, innovation is turning from an art into a science. Page 6
The love-in The move toward open innovation is beginning to transform entire industries. Page 8
The fading lustre of clusters The best thing that governments can do to encourage innovation is get out of the way. Innovation, long the preserve of technocratic elites, is becoming more Page 11 open. This will be good for the world, argues Vijay Vaitheeswaran
CRISIS is a terrible thing to waste, alent of 100 miles-per-gallon using alter- The age of mass innovation AVinod Khosla laments to Larry native energy. The charitable arm of Mr We are all innovators now. Page 13 Page. The two Silicon Valley luminaries are Page’s rm has already taken hybrid pet- chatting one evening at the Googleplex, rol-electric vehicles, like the Toyota Prius, the quirky Californian headquarters of and turned them into even cleaner plug Google. The crisis which Mr Khosla is con- in versions which can be topped up from cerned about is caused by carmakers’ ad- an electric socket. diction to oil and the consequent warming Mr Khosla believes clean cars, using ad- of the planet. The energy and car indus- vanced biofuels or other alternatives, will tries have not been innovative in many come about only through radical innova- years because they have faced no real cri- tion of the sort that Big Oil and Big Autos sis, no impetus for change, he insists. avoid. Risk and acceptance of failure are The two are plotting what they hope central to innovation, he argues, but the di- will be the next great industrial revolution: nosaurs typically avoid both. Big compa- the convergence of software and smart nies didn’t invent the internet or Google, electronics with the grease and grime of and much of the big change in telecoms the oil and car industries. Mr Khosla is also came from outsiders, he adds. kicking around his plans for getting chip Coming from almost anyone else such guys together with engine guys to de- talk would sound preposterous. But Mr Acknowledgement The author is grateful to the many velop the clean, software-rich car of the fu- Khosla and Mr Page are not ordinary busi- thoughtful people who shared their views with him. In ad- ture. Such breakthroughs happen only nessmen or armchair revolutionaries. Mr dition to those mentioned in the special report, he would like to thank John Denham, Kishore Mabhubhani, Dev when conventional wisdom is ignored Khosla helped to found Sun Microsystems, Patnaik, Ken Zita, Douglas Horn and experts at the World and cross-fertilisation encouraged; man- a path-breaking information-technology Bank, OECD and World Economic Forum. Special thanks go aged con ict , in his words. rm, and he went on to become a partner to the generous organisers of two distinctive conferences, "Sci Foo" and Technology, Entertainment, Design (TED), Mr Page, co-founder of Google, had ear- at Kleiner Perkins, a venture-capital com- which manage to capture the essence of innovation. lier hosted a gathering of leading environ- pany that was an early backer of Ama- mentalists, political thinkers and energy zon.com, America Online and many other A list of sources is at experts to help shape an inducement to get pillars of the internet economy. Mr Page’s www.economist.com/specialreports things moving: the Automotive X Prize, to Google is one of the internet’s biggest suc- be unveiled in early 2008. The organisers cess stories. At 34 he is a multi-billionaire. will o er at least $10m to whoever comes But these men are from Silicon Valley; An audio interview with the author is at up with the best e cient, clean, a ord- and Silicon Valley is not America. It is www.economist.com/audio able and sexy car able to obtain the equiv- tempting to dismiss such breathless talk of1 2 A special report on innovation The Economist October 13th 2007
2 revolution as just more hype from people concept of fast-food popularised by Mc- a decent chance of being turned into a pro- who are seeing the world through Google Donald’s, for instance, involved running a duct and brought to market. goggles. After all, go beyond the rare ed air restaurant in a di erent way rather than So why does the generation and han- of northern California and the rules of making a technological breakthrough. dling of ideas matter so much? We rmly gravity are no longer suspended. The well- However, innovation can involve plenty believe that innovation, not love, makes established industries which they mock of clever gadgets and gizmos. the world go round, insists John Dryden still move at their usual but reliably glacial One way to arrive at a useful de nition of the OECD. Corny perhaps, but studies pace, right? is to rule out what innovation is not. It is do show that a large and rising share of Well no, actually. Rapid and disruptive not invention. New products might be an growth and with it living standards over change is now happening across new and important part of the process, but they are recent decades is the result of innovation old businesses. Innovation, as this report not the essence of it. These days much in- (see chart 2). Innovative rms also tend to will show, is becoming both more accessi- novation happens in processes and ser- outperform their peers. We’re not discov- ble and more global. This is good news be- vices. Novelty of some sort does matter, al- ering new continents or encountering vast cause its democratisation releases the un- though it might involve an existing idea deposits of new minerals, Mr Dryden tapped ingenuity of people everywhere from another industry or country. For ex- adds. Indeed, the OECD’s experts believe and that could help solve some of the ample, Edwin Drake was not the rst man that most innovation has been caused by world’s weightiest problems. to drill for a natural resource; the Chinese globalisation and new technologies. The seditious scene from the Google- used that technique for centuries to mine Analysis done by the McKinsey Global plex also captures the challenge this salt. But one inspired morning in 1859, Col- Institute shows that competition and inno- presents to established rms and de- onel Drake decided to try drilling for oil in vation (not information technology alone) veloped economies. For ages innovation Titusville, Pennsylvania. He struck black led to the extraordinary productivity gains has been a technology-led a air, with gold and from his innovation the modern seen in the 1990s. Those innovations in most big breakthroughs coming out of oil industry was born. technology as well as products and busi- giant and secretive research labs, like Xe- ness processes boosted productivity. As rox PARC and AT&T’s Bell Laboratories. It The men in white coats productivity rose, competition intensi ed, was an era when big corporations in de- The OECD, a think-tank for rich countries, bringing fresh waves of innovation, the veloped countries accounted for most says innovation can be de ned as new institute explains. R&D spending. products, business processes and organic That is why innovation matters. With North America still leads the world in changes that create wealth or social wel- manufacturing now barely a fth of econ- research spending (see chart 1), but the big fare. Richard Lyons, the chief learning of- omic activity in rich countries, the knowl- labs’ advantage over their smaller rivals cer at Goldman Sachs, an investment edge economy is becoming more impor- and the developing world is being eroded bank, o ers a more condensed version: tant. Indeed, rich countries may not be by two powerful forces. The rst is global- fresh thinking that creates value . Both hit able to compete with rivals o ering low- isation, especially the rise of China and In- the nail on the head, and will serve as the cost products and services if they do not dia as both consumers and, increasingly, de nition in this report. learn to innovate better and faster. suppliers of innovative products and ser- According to popular notion, innova- But even if innovation is the key to vices. The second is the rapid advance of tion is something that men wearing white global competitiveness, it is not necessar- information technologies, which are coats in laboratories do. And that’s the ily a zero sum game. On the contrary, be- spreading far beyond the internet and into way it used to be. Companies set up verti- cause the well of human ingenuity is bot- older industries such as steel, aerospace cally integrated R&D organisations and tomless, innovation strategies that tap into and carmaking. governments fussed over innovation poli- hitherto neglected intellectual capital and What is innovation? Although the term cies to help them succeed. This approach connect it better with nancial capital can is often used to refer to new technology, had successes and many companies still help both rich and poor countries prosper. many innovations are neither new nor in- spend pots of money on corporate re- That is starting to happen in the develop- volve new technology. The self-service search. But rms are growing increasingly ing world. 7 disenchanted because the process is slow and insular. A global study across indus- Investing in ideas 1 tries by Booz Allen Hamilton, a consul- Productive innovation 2 R&D spending by region, 2006, % tancy, even concluded that higher R&D US productivity growth spending doesn’t ensure better perfor- Output per hour, 1954=100 North America Europe 43.7 28.9 mance in terms of growth, pro tability or 300 shareholder returns. Now the centrally planned approach is Innovation Labour 250 Total: giving way to the more democratic, even $478bn Capital joyously anarchic, new model of innova- 200 Other tion. Clever ideas have always been every- 0.5 where, of course, but companies were of- 150 China & India ten too closed to pick them up. The move 0.6 to an open approach to innovation is far Rest of Asia Japan more promising. An insight from a bright 100 4.8 21.5 spark in a research lab in Bangalore or an 1954 60 70 80 90 2004 Source: Booz Allen Hamilton database avid mountain biker in Colorado now has Source: Boston Consulting Group The Economist October 13th 2007 A special report on innovation 3
Revving up
How globalisation and information technology are spurring faster innovation
F YOU want a motorcycle, go to Chong- start-ups, venture-capital rms, large inde- Iqing. Although this dusty central Chi- pendent software rms and some of its nese city of drab o ce buildings and per- most important customers. petually grey skies is better known as the Innovation is also changing the gateway to the enormous Three Gorges pharmaceuticals industry. Small biotech- Dam, it is also the two-wheeler capital of nology rms, using networked ap- the world. Led by the region’s pioneers, proaches, are getting ahead of Big Pharma. China now makes half the world’s motor- This too opens the way for Asian competi- cycles. But more important than the num- tors, like Ranbaxy and Dr Reddy’s Labora- bers produced is the way these motorcy- tories. These rms were once copycats, cles are made especially the way trampling on Western patents to make designers, suppliers and manufacturers cheap generic versions of drugs. But have organised themselves into a dynamic increasingly they are shifting to process in- and entrepreneurial network. novation and even new drug discovery. Unlike state-run rms, the city’s priv- Such innovation can arise out of neces- ate-sector upstarts, such as Longxin and sity. Entrepreneurs in China must compete Zongshen, do not have big foreign partners with privileged state rms with access to like Honda or Suzuki with deep pockets cheap credit as well as the local arms of and proven designs. So they came up with multinationals. That makes China’s third a di erent business model, one that was models, like the one used by Chongqing’s sector , as Messrs Seely Brown and Hagel simpler and more exible. Instead of dic- motorcycle-makers. Other examples in- call it, extraordinarily resourceful in trying tating every detail of the parts they want clude the design networks established by to reach global markets. India has been from their suppliers, the motorcycle-mak- Taiwanese contract-producers in the tex- less integrated into the world economy, so ers specify only the important features, tile industry. Groups of innovative just-in- many of its innovative rms have initially like size and weight, and let outside de- time suppliers abound in Asia, feeding concentrated on reaching bottom of the signers improvise. Western fashion and consumer-goods pyramid consumers. For instance, Selco, This so-called localised modularisa- companies. They are often managed by an Indian solar-energy pioneer, found that tion approach has been very successful supply-chain experts, like Hong Kong’s Li because many of its customers were living and delivered big cost reductions and & Fung. Unlike Japan’s keiretsu, which in remote areas, it had to set up local net- quality improvements, says John Seely bound companies and their suppliers to- works of trained technicians to sell, install Brown, an innovation expert who used to gether with interlocking shareholdings, and repair its products, and provide cus- head the legendary Xerox PARC research these rms are free to leave their alliances. tomers with small loans. centre. It is one example of the sort of busi- They stay together only if they continue to ness-model innovation which he insists is learn and pro t from the experience. In Bigger names far more radical than conventional pro- some ways they resemble the nimble net- Most of these Chinese and Indian innova- duct or process innovation. works of rms that underpinned Silicon tors are not well known, but it is only a Valley’s success. matter of time before some will be. Frans China moves ahead Low labour costs may have given such van Houten, chief executive of NXP, a Examples of these business-model inno- rms a head start, but that is a transitory European semiconductor rm, is con- vations are now bubbling up from de- advantage. India’s software innovators vinced of that. He says there are now over veloping economies to threaten the estab- were once sni ed at as merely low-cost o - 400 rms designing chips in China. So far lished global giants. In a report with John shoring and back-o ce operations. But they produce very pragmatic, t for use Hagel, of Deloitte, a consultancy, Mr Seely rms like Infosys, Wipro and Tata Consul- designs, but he has no doubt they will Brown argues that the activity of private tancy Services (TCS) have become world quickly become world-class innovators. entrepreneurs means China is rapidly leaders in business-software services. S. One company the big carmakers are emerging as the global centre of manage- Ramadorai, TCS’s chief executive, says his watching closely is India’s Tata Motors, ment innovation, pioneering manage- rm sees innovation as a key enabler of which is developing a people’s car that ment techniques that most US companies its productivity edge . He points out that might radically change the process of de- are struggling to understand. his rm has been investing in R&D for 25 sign, manufacturing and distribution to The emergence of Asian world-beaters years and holds several dozen patents and achieve its target price of no more than exempli es the two forces driving innova- copyrights. Navi Radjou of Forrester Re- $3,000. If successful in India, Tata will pro- tion. Globalisation and the spread of in- search, a technology consultancy, ap- duce a version of the car for export. formation technology allow the creation plauds TCS’s global innovation ecosys- As the knowledge component of indus- of unexpected and disruptive business tem which brings together academic labs, tries continues to grow, it will lower even1 4 A special report on innovation The Economist October 13th 2007
2 further the barriers to entry in many busi- nesses. Yet the same democratisation of in- novation that empowers the new rms can be used to generate much greater inno- vation from within established compa- Can dinosaurs dance? nies. Some multinationals are already do- ing this in Asia to keep up with their local Responding to the Asian challenge competitors (see box). The e ects of the growing knowledge- RE consumers in India and China too Shakti, which provides Indian women’s component of innovation have become Apoor to a ord high-quality Western self-help groups with business education increasingly clear in heavy engineering. goods? That used to be the old idea of do- and the chance to earn a living selling Reinhold Achatz, of Siemens, claims the ing business in these countries as rms cheap sachets of Unilever products. The German giant has undergone a hidden o ered watered-down versions of their e ort has proved so successful that Uni- electronics revolution. We have more products at reduced prices. Mr van Hou- lever introduced a high-tech element: the software developers than Oracle or SAP, ten, of chipmaker NXP, says Indian and Shakti entrepreneurs now run kiosks but you don’t see this because it is embed- Chinese consumers are forcing multina- with personal computers which villagers ded in our trains, machine tools and fac- tionals to design sophisticated products can rent to send e-mails and browse the tory automation, he says. Mr Achatz cal- that more closely meet their needs, and web for things that can make a big di er- culates that as much as 60% of his rm’s this is making rms operating in Asia bet- ence to their lives, like market prices. sales now involve software. Some 90% of ter innovators. Alan La ey, who ran some of Procter the development in machine tools is in By recruiting ingenious local engi- & Gamble’s Asian businesses before get- electronics and related hardware, and the neers and designers in places like Banga- ting the top job at the American com- gure is similar for cars. A BMW, he says, is lore and Beijing, and paying close pany, says many Asian rms began now actually a network of computers. attention to trends and practices in the imitating what foreign ones did but are market, rms are coming up with pro- now very innovative, especially with Flat out in Germany ducts and services that can be sold in business models . That may seem like an exaggeration until other parts of the world too. Nokia’s engi- Mr La ey sees Indian rms shaking you step into the sleek new Hydrogen 7 neers are nding that many Chinese and up the way foreign companies operate, BMW saloon. Push the pedal to the metal Indians access the internet mainly and not only with back-o ce services on the autobahn and the car responds as through their mobile handsets. Such cus- where many began. Hours after he ut- every BMW should; cylinders growling en- tomers’ requirements of their handsets tered those words, Wipro, an Indian pio- thusiastically as the ultimate driving may therefore be quite di erent to those neer of software services said it would machine races past slower vehicles. But of Western users, many of whom have open a new development centre in At- this car is not like any other made by computers at home and at work. lanta, Georgia, that will report to its head- BMW. Press a button on the steering wheel GE’s research lab in China has come quarters in Bangalore. and it seamlessly switches from burning up with a simpli ed magnetic-resonance This is forcing P&G to innovate in petrol to hydrogen. imaging machine that costs a fraction of other ways too. Mr La ey uses the exam- The key to this advance, says Ulrich the one it sells in rich countries. The rm ple of detergents in China, where the Weinmann, of BMW, is smart software. now plans to sell it worldwide. Wenda, a company is using a low-cost manufactur- Electronics have been in cars for decades, question-and-answer knowledge com- ing method which he likens to Coca- but those were isolated dumb systems , munity product developed by Google in Cola’s syrup model, which supplies a he adds. Now cars are crammed full of net- China to help overcome a lack of local concentrate to local bottlers. P&G pro- works of computers with smart software content, was launched in Russia in June. vides secret, high-value performance controlling and monitoring things. New Unilever has long had a strong distri- chemicals to Chinese partners, who add BMWs can even synchronise with Apple’s bution network in India, but it has ex- basic ingredients and packaging before iPhone, and download maps and direc- panded its e orts with a division called distributing the products. tions from Google while you drive. The steady conversion of engineering into yet another knowledge-based indus- who is challenging incumbents in not one contractors, are working furiously to com- try forces the pace. We are a quite mature but two old-time industries. Mr Musk mercialise space. In September the X Prize industry, but customers now expect made his fortune during the internet Foundation and Google decided to fuel the change faster, adds Mr Weinmann. The boom by selling PayPal, an online pay- re by announcing a $30m prize for the demand for change is fastest in Asia. Sev- ments system, to eBay for $1.5 billion. He rst private-sector team to land and oper- eral hundred new mobile phones are now heads Space Exploration Technol- ate an unmanned rover on the moon. Peter launched every year in China, and cus- ogies, known as SpaceX. This is a start-up Diamandis, the foundation’s chairman, tomers there now expect their new BMWs o ering private space launches. Earlier this believes the old guard is no longer able to to be able to synchronise perfectly with year, it red a rocket into space, the rst to innovate. Real breakthroughs require risk each new handset, he sighs. be designed, paid for and launched en- and the ability to absorb failure, and large New competitors are emerging from tirely with private money. organisations are incapable of such risk unexpected quarters, which makes things SpaceX is the vanguard. Many private- taking, he says. di cult for established rms. One of them sector newcomers, fed up with the over- Mr Musk is not waiting to win any is Elon Musk, a 36-year-old entrepreneur bearing ways of NASA and the big defence prizes. Besides SpaceX, he has also started1 The Economist October 13th 2007 A special report on innovation 5
2 Tesla Motors, which has devised an elec- 2010 experts think it could approach half. by a large plaque in the lobby which says: tric sports car capable of accelerating from Thanks to recent advances in batteries and Incrementalism is innovation’s worst en- zero to 60mph (100kph) in four seconds power electronics (made possible by inno- emy! We don’t want continuous improve- and has a top speed of over 130mph. More vations to power mobile phones and lap- ment, we want radical change. These are impressively, thanks to its advanced lith- top computers), Mr Musk thinks most cars the words of Sam Walton, the rm’s foun- ium-ion batteries and lightweight carbon- will become electrically powered. In 50 der. And to his credit, Walton did radically composite construction, the Tesla Roadster years, we’ll look back on the internal com- change the general store with his innova- has a range of 200-250 miles from an over- bustion engine and see it as a giant anach- tive approach to low cost, high-volume su- night charge. The rst cars to be produced ronism, like the steam locomotive. permarket retailing. But ask Linda Dill- on a larger scale are expected to hit the And again Asia could take the lead. man, a senior o cial at the rm, about road next year. They will cost a pricey Robert Lutz, GM’s head of product de- innovation at Wal-Mart today and she con- $100,000 or so, but Mr Musk’s rm has al- velopment, says investment in and enthu- cedes that radical thinking was easier ready started work on a new factory to pro- siasm for clean technologies in Asia is so when the rm was young. Meg Whitman, duce a family car which the company great that cars powered by fuel cells eBay’s boss, says the same. She concen- hopes to sell at half the price. (squeaky-clean devices that use hydrogen trates on incremental improvement with- Larry Burns, in charge of R&D for Gen- to make electricity) are likely to take o in in the online auctioneer while looking out- eral Motors (GM), is impressed by Tesla’s China before they do in the United States. side to acquire radical ideas by buying technology, but points out that the rm’s There is reason to believe he might be start-up companies, including ones in initial output of just a few hundred cars is right. Just as villagers in Africa and Bangla- other markets that imitate eBay. tri ing compared with global car produc- desh have gone straight from no phones to tion of some 60m vehicles a year. mobile phones, developing countries Ideas at double speed He is right, but Tesla and others are the could leapfrog with other innovations. Many executives feel the heat is on and thin end of what could be a big wedge. Be- Developing countries already have that they must innovate faster just to stand sides the innovative process lowering the higher levels of early stage entrepreneur- still. One reason is that product cycles are barriers to entry, much key intellectual ship, with more people engaged in things undeniably getting shorter. Gil Cloyd, property involved in carmaking nowa- such as starting new ventures often be- chief technology o cer at Procter & Gam- days is no longer guarded in-house by the cause the necessity for doing so is greater ble (P&G), the world’s biggest consumer- likes of GM and Ford: they now outsource (see chart 3). Tim Jones of Innovaro, a Euro- products rm, studied the life cycle of con- most aspects of making a new car (except pean innovation consultancy, points out sumer goods in America from 1992 to 2002 engines) to global parts suppliers and out- that Africa is about to take the lead in using (before the internet’s full impact was felt) side rms that put together large sections, mobile phones for payments and remit- and found that it had fallen by half. That, or modules . That makes it much easier tances, thanks to the introduction of he concludes, means his rm now needs to for newcomers to buy any bits they need. schemes like the M-PESA money-transfer innovate twice as fast. service introduced by Vodafone and Citi- 3M, an American company famous for Manufacturing is integrating group in Kenya. These allow people to inventing the Post-it sticky note, also be- Even more important, the cost of launch- send money using text messages. lieves the world is moving much faster. ing a new car company has dropped dra- Some people reckon that, as the nature Andrew Ouderkirk, one of the rm’s cele- matically. When Toyota and GM of innovation changes, so it is speeding up. brated inventors, thinks that is in part be- launched, respectively, Lexus and Saturn But that’s not obvious. Other periods have cause many things that his company used as semi-independent new companies, seen bursts of dramatic technological pro- to do in-house are now done by outsiders. they had to spend billions of dollars. Now gress: the arrival of the telegraph, for in- To keep up, 3M carries out concurrent de- an upstart would need just a few hundred stance, was just as disruptive as the in- velopment , which involves talking to cus- million. Paul Horn, the outgoing head of ternet is today. tomers much earlier in the process to try to research at IBM, believes that the car in- Visit Wal-Mart’s headquarters in Ben- shorten development times. dustry is but one example of the decom- tonville, Arkansas, and you will be greeted Even the rm that laid down the rst position of the vertically integrated busi- long-distance telegraph lines thinks to- ness model: car rms were once very day’s innovation frenzy is unprecedented. integrated, but now don’t make anything The necessity of invention 3 Mr Achatz, of Siemens, is adamant that in- they’re integrators in a ‘value net’. Early-stage entrepreneurial activity novation is happening much quicker and Mr Musk, mindful that he has yet to see that access to information is so fast now 20 a return on his investment, agrees the costs Indonesia that it allows much faster product-de- of entry are much lower, but gives warn- China velopment cycles. His rm is convinced 15 Thailand ing that it will not be easy to take on the in- Australia that there will be an explosion of medical Malaysia United India States cumbents: The last successful car start-up 10 know-how thanks to the advance of in- Argentina in America was 100 years ago. Even so, he Canada formation technology into medicine. Britain is convinced that the time has come to try Czech Republic Denmark Perhaps managers at rms everywhere 5 France Germany because of a fundamental technology entrepreneurial activity should be both far-sighted and paranoid in years involved in early-stage Japan discontinuity : the shift from the internal- % of population between 18-64 0 equal measure as they scan the horizon for combustion engine to electric drive. 0 1020304050 unexpected competitive threats. Some The proportion of electronics that GDP per person at PPP*, 2006, $000 companies are trying to organise them- makes up the cost of a new car has shot up *Purchasing-power parity selves with management techniques to from very little to perhaps a quarter. By Source: Global Entrepreneurship Monitor 2006 face just such a disruptive future. 7 6 A special report on innovation The Economist October 13th 2007
A dark art no more
Like management methods before it, innovation is turning from an art into a science
HAT matters gets measured. be found in management. from their normal jobs contributes more to WThat is one of the basic tenets of One reason why bosses might not want the rm than it costs. corporate strategy taught at business to be too obsessive about creativity is that It is a question that even Eric Schmidt, schools. As driving growth through inno- generating ideas is the easy part. Exploit- Google’s chief executive, cannot answer. vation is today at the top of corporate ing them has always be harder. As Thomas Surprisingly, he declares that trying to agendas you would expect to nd manag- Edison, one of America’s greatest inven- measure his rm’s innovation process ers treating it like a science. After all, manu- tors, put it, genius is 1% inspiration and 99% would choke it o altogether. Tim Brown, facturing philosophies such as total-qual- perspiration. But many managers are re- head of Ideo, a design consultancy, con- ity management (a process of continuous luctant to take the same hard-nosed ap- curs: A lot of innovation is anti-Six Sigma. improvement) and Six Sigma (which proach they use in other parts of their busi- You want a lot of variance. uses statistical methods to eliminate varia- ness and apply it to fragile creative types. tions and defects) were quanti ed and If any rm has an analytical approach Fuzzy logic widely deployed a long time ago, often to innovation it should be Google. After Not surprisingly, Je rey Immelt, chairman with good results. all, the rm’s superstars are its software en- of GE, strongly disagrees. His rm has long Yet innovation remains a frustratingly gineers. It is so obsessed with data that it been a champion of Six Sigma. Mr Immelt fuzzy notion. Many bosses think it is essen- posts nerdy tip sheets on statistical-quality reckons that operational excellence is tially a creative process. Some anoint measurement above the urinals at the the crucial part of innovation, not the chief innovation o cers , bring in con- Googleplex. And yet managers sound like fuzzy ideas-generation bit. He suggests that sultancies or set up secret skunk works mumbling teenagers when they are asked passion and vision might make up just to tease out the ideas they fear their own how they approach innovation. 20% of the process. bureaucracy might squash. One senior ex- Marissa Mayer, the company’s am- Larry Keeley of Doblin, a innovation ecutive maintains that innovation simply boyant head of user experience , declares consultancy, has followed this debate cannot be de ned exactly, but that like that Google is not merely a search engine closely for decades and insists the answer pornography I know it when I see it. but an innovation engine that needs is clear: Creativity is maybe 2% of the in- constantly to reinvent itself just like novation process. It’s a vanishingly small The wrong measure Macs and Madonna . As 3M and some component, and it’s the part you can ac- Jorma Ollila, non-executive chairman of other rms do, Google grants its engineers quire from outside the rm. both Nokia and Royal Dutch/Shell, argues permission to spend 20% of their paid time Despite di culties trying to de ne it, that it is a mistake to measure innovation on pet projects unrelated to their daily job. the innovation process is steadily becom- by the number of patents issued by a com- She points to a few examples of new pro- ing a practical science to be measured, pany or the extent to which new technol- ducts that have emerged this way, such as taught and managed. Clayton Christen- ogies are introduced. He suggests that the Gmail, but cannot provide any real evi- sen, a professor at Harvard Business most fertile area of innovation today can dence that allowing sta to take time o School and an expert on the subject, insists1 The Economist October 13th 2007 A special report on innovation 7
2 that innovation simply isn’t as unpredict- ventions but, to use Mr Immelt’s words, that given today’s accelerating pace of able as many people think. There isn’t a in turning $50m ideas into billion-dollar global innovation rms need really harsh cookbook yet, but we’re getting there. ideas. His way of doing that is a highly discipline to weed out ideas quite The Haas business school at the Uni- structured process that involves a mix of quickly we are working at fast failing, but versity of California at Berkeley has al- management training, increased exposure are not there yet. He thinks his own com- ready gone so far as to revamp its entire to outside ideas (for example, his rm is pany’s legacy as a hardware manufac- curriculum to concentrate on innovation starting a venture capital fund to get early turer a capital-intensive and slow-mov- management. Berkeley is home to some of visibility of clever inventions) and con- ing sector compared with software or the leading experts on the subject, includ- tinuous funding for the development of services is holding it back. ing Henry Chesbrough (who popularised new ideas. He also emphasises that the ac- Turf wars are another obstacle to fast the notion of open innovation ) and An- ceptance of failure is an integral part of the failing. Employees in one part of a com- naLee Saxenian (whose recent book The e ort, as long as it is fast failing . pany often reject ideas and advice from a New Argonauts analyses Silicon Valley It is the last bit of Mr Immelt’s process di erent part. Mark Little, GE’s head of re- and related innovation clusters). Richard that points to one of the biggest thoughts search, confesses that getting his bo ns to Lyons, now of Goldman Sachs, led the re- emerging from innovation research in re- kill o unviable projects is the hardest task vamp at Haas in his previous job. He is cent years: neither idea generation nor exe- he faces: Like a dog with a bone, people convinced that all managers can be taught cution is as important or as tricky as the l- don’t want to give them up. how to nurture innovation. Even if rms can overcome the stigma The rough outline of how this might be of failure, how exactly are bosses to know done is emerging. But there is no one-size- which potential innovations to kill? Mr ts-all strategy. Bosses have to appraise the Christensen, author of The Innovator’s strengths and weaknesses of their rms Dilemma , believes he has cracked the honestly and continuously to take account code. He says it can require unlearning of rapidly evolving competitive threats. some of the things that managers often ac- But cut through the clutter of PowerPoint cept as golden rules. The chief one is the presentations and faddish slogans, and a belief in listening and responding to the number of things become clear. needs of your best customers.
All that jazz Siren songs For a start the debate over creativity versus This seemingly sensible strategy can be a execution should be put to rest: rms need dangerous siren song, Mr Christensen ar- to do both. But that does not mean they gues. His in uential book shows how have to do it all themselves. On the con- even successful rms can get into trouble trary, the double act is best managed with by trying to please their best customers. a loose and open approach during the Because there may be only a handful of wild and woolly idea-generation phase, highly pro table, high-end buyers who and a tighter, more concentrated one to want and can a ord more features and bet- turn ideas into products or services. John ter performance, rms can invest heavily Kao, author of Jamming: The Art and Dis- in trying to deliver what this elite group cipline of Business Creativity , likens the wants even though the resulting products process to playing jazz: there is no xed may end up beyond the reach of the ma- score in any given improvisation, but that jority of their customers. does not mean there are no underlying That, argues Mr Christensen, allows principles either. tering process that links the two. Harold upstarts to enter the market and o er infe- P&G is a good example of an inward- Sirkin, of the Boston Consulting Group, is rior (although perfectly adequate) technol- looking rm that has embraced creativity the co-author of Payback , a book on in- ogies and products at much cheaper prices and openness with some success. But Mr novation strategy. He sco s that rms and push incumbents into ever smaller La ey, its chairman, makes clear this is no have too many ideas and too much em- niches and ultimately out of business al- mystical process. He argues that even a phasis on creativity more ideas merely together. He cautions this disruptive in- process that is open to fresh thinking from choke the funnel even more. In fact, the novation is not the same thing as radical the outside, as P&G’s is, can be run the more ideas a rm comes up with, the more or breakthrough innovation, although same way as a factory: It is possible to important it is for bosses to decide early on the notions are often con ated. In his measure the yield of each process, the which of them to kill o . This is to avoid view, personal computers disrupted IBM’s quality and the end product. heading down countless and costly dead mainframe computers and Digital Equip- On the ip side, a rm known for ends. As Ron Adner of Insead, a French ment’s mini-computers, as did Nucor’s emphasising execution over creativity is business school, puts it, Innovation is a highly e cient mini-mills to US Steel’s GE. Its focus on the practical application of loser’s game, as we know most initiatives blast furnaces. new ideas, rather than invention itself, fail. But the truly innovative companies Now Chinese and Indian rms are goes all the way back to its founder, Edi- know how to deal with losing. poised to disrupt established companies son. Indeed, he commercialised but did That is why failing fast and learning everywhere in much the same way, he ar- not invent the light bulb. from those failures is so important for com- gues. Their impact, he says, will be even GE’s strength is not in breakthrough in- panies. Niklas Savander, of Nokia, argues more traumatic because both countries1 8 A special report on innovation The Economist October 13th 2007
2 have a large pool of domestic customers sions, but teenagers who never had radios treated as part of an executive’s job. many of whom have only just begun con- loved these cheap devices. Is there a risk that with too many rules, suming and do not have the same high He also thinks it is better to make things rms could lose out to serendipity? Ask Mr expectations as Western customers typi- simpler and easier for the bottom and mid- La ey how he intends to keep P&G’s edge cally have. Chinese and Indian companies dle of the market, as personal computers if innovation becomes less ad hoc and he can practise on their domestic customers did, rather than add needless bells and immediately points to Toyota’s embrace of while they improve quality to the point whistles for the handful of top customers total-quality management as a model. they can begin to export. South Korean who can a ord and demand them. And he Many rms have studied the Japanese car- rms have already gone through much the says companies should act decisively to maker’s legendary methods, as P&G’s ri- same process with consumer-electronics co-opt or pre-empt disruptive ideas them- vals are even now studying its innovation and cars and in the process have fright- selves, even if it threatens their core busi- model, but none has really been able to ened many of their Japanese rivals. nesses in the short run. copy it. That is because Toyota’s real edge is Executives at USSteel, a traditional inte- the strong culture which drives its unre- Snap, and it’s too late grated steel- rm nervously eyeing the lenting quest for quality. In a sense, Mr Christensen’s management threat from new mini-mill technology, Bill Reinert, a senior Toyota o cial myths echo a sentiment expressed by Ed- nearly built a cheap and cheerful mini-mill based in North America, explains it thus: win Land, the inventor who founded Po- themselves to compete against the upstart What’s discontinuous about our rm is laroid. People who seem to have had a Nucor. However, recounts Mr Christen- the very long view of management. That new idea have often simply stopped hav- sen, those aspiring innovators within US vision has pushed us from being a closed ing an old idea, he said. Alas, his succes- Steel were forced to halt the pro table pro- company to one with continuous informa- sors at Polaroid did not pay attention. The ject by bean counters, who argued that it tion ows, both into the company and rm stuck by its successful old idea for was cheaper just to produce more steel within it, about market, regulatory and lm-based instant photography and stub- from the rm’s existing blast furnaces geopolitical trends. bornly ignored the disruptive potential of (since their capital costs had been paid for A symbol of this is Toyota’s Prius hy- digital imaging until it was too late. Polar- and steel could be produced for merely the brid-electric car. It was a risky bet on an oid went bust in 2001. marginal cost of cranking out an extra unproven technology, but it has been a Mr Christensen’s alternative innova- tonne). That short-term thinking scup- huge success. It was a long-term vision, tion strategies include watching out for pered the giant rm’s best chance for re- says Mr Reinert, that overcame the rm’s new technologies or new business models inventing itself. innate caution. And in the future the com- which are designed to attract customers Peter Drucker, an eminent manage- pany is going to have to make similar bets who may not be using your product today ment guru, argued decades ago that inno- again. We are convinced that we are en- because it too expensive or too compli- vation and entrepreneurship are pur- tering a disruptive future, and we want to cated. Sony’s early transistor radios were poseful tasks that can be organised are in be ready for it, he says. He is not alone in tinny compared with RCA’s big home ver- need of being organised and should be taking that view. 7 The love-in
The move toward open innovation is beginning to transform entire industries
ERKELEY seems like a tting place to (see chart 4 on next page). from less than a fth to around half. That B nd the godfather of the open-innova- To see why travel to Cincinnati, Ohio has boosted innovation and, says its boss, tion movement basking in glory. The Cali- which is about as far removed culturally Mr La ey, is the main reason why P&G has fornian village was, after all, at the very from Berkeley as one can get in America. been able to grow at 6% a year between heart of the anti-establishment movement The conservative mid-western city is 2001 and 2006, tripling annual pro ts to of the 1960s and has spawned plenty of home to P&G, historically one of the most $8.6 billion. The company now has a mar- radical thinkers. One of them, Henry Ches- traditional rms in America. For decades, ket capitalisation of over $200 billion. brough, a business professor at the Univer- the company that brought the world Ivory IBM is another iconic rm that has sity of California at Berkeley, observes soap, Crest toothpaste and Ariel detergent jumped on the open-innovation band- with a smile that this is the 40th anniver- had a closed innovation process, centred wagon. The once-secretive company has sary of the Summer of Love. around its own secretive R&D operations. done a sharp U-turn and embraced Linux, Mr Chesbrough’s two books Open In- No longer. P&G has radically altered the an open-source software language. IBM novation and Open Business Models way it comes up with new ideas and pro- now gushes about being part of the open- have popularised the notion of looking for ducts. It now welcomes and works with innovation community , yielding hun- bright ideas outside of an organisation. As universities, suppliers and outside inven- dreds of software patents to the creative the concept of open innovation has be- tors. It also o ers them a share in the re- commons rather than registering them come ever more fashionable, the corporate wards. In less than a decade, P&G has in- for itself. However, it also continues to take R&D lab has become decreasingly rele- creased the proportion of new-product out patents at a record pace in other areas, vant. Most ideas don’t come from there ideas originating from outside of the rm such as advanced materials, and in the pro-1 The Economist October 13th 2007 A special report on innovation 9
2 cess racks up some $1 billion a year in li- stay ahead of the competition. censing fees. People power 4 Even in developed markets, the accel- Since an army of programmers around Most significant sources of innovative ideas eration of innovation is making patents the world work on developing Linux es- % of respondents selecting up to three choices less relevant. What is more, say brand ex- sentially at no cost, IBM now has an ex- perts at P&G (which claims not even to 01020 3040 tremely cheap and robust operating sys- count patents any longer), the dizzying tem. It makes money by providing its Employees pace of change today confuses consumers clients with services that support the use Business partners with a ba ing array of choices. Such rms of Linux and charging them for it. Using are increasingly turning to trusted brands open-source software saves IBM a whop- Customers to simplify things for their customers. An- ping $400m a year, according to Paul Horn, Consultants drew Herbert, head of Microsoft’s research until recently the rm’s head of research. Competitors centre in Cambridge, England, puts it this The company is so committed to openness Associations, trade way: Our brand hides a tremendous that it now carries out occasional online shows, conference boards amount of innovation. Internal sales and jam sessions during which tens of thou- service units Open innovation also appears to keep sands of its employees exchange ideas in a corporate bureaucrats on their toes, mak- Internal R&D mass form of brainstorming. ing companies better at competing. The Mr Chesbrough, of course, heartily ap- Academia combination of exciting new technologies proves. He gives dozens of other examples Source: IBM “The Global CEO study 2006”, based on and juiced-up management processes has, of rms doing similar things, ranging from interviews with 765 CEOs and business leaders according to Mr La ey, helped P&G to re- Clorax, a household products rm to Air duce its rate of failed product-launches Products, an industrial gases company. Mr GE’s Mr Immelt observes that his rm is from eight out of ten to just half. Chesbrough reckons that IBM and P&G a leader in a number of elds, such as mak- Unilever’s David Duncan insists that have timed their shift to a high-volume ing jet engines and locomotives, which re- his rm one of P&G’s biggest competi- open-business model very well and that quires doing things that almost nobody tors is much better connected to its cus- if their competitors do not do the same else in the world can do and where intel- tomers than it was. Twelve years ago, they will be in trouble. lectual-property rights and a degree of se- when I joined, we were very closed, verti- Not everyone is impressed. Kenneth crecy still matter. Mark Little, his head of cally integrated and owned most of the Morse, head of MIT’s Entrepreneurship research, is even more sceptical and says value chain even the chemicals and soft- Centre, sco s at IBM’s claim to be an open outside ideas don’t really stick well here . ware we used, he says. Now it is much company: They’re open only in markets, He professes great satisfaction with the more receptive to ideas and services from like software, where they have fallen be- output of GE’s own research laboratories. the outside, even posting challenges on the hind. In hardware markets, where they We’re pretty happy with the hand we’ve internet for people to come up with new have the lead, they are extremely closed. got, he adds. ideas. But he too confesses that there can Horses for courses, perhaps. Boosters be di culties: it’s like the rst time you Open costs of open-innovation agree that there are used Google; it was scary and a bit tricky, David Gann and Linus Dahlander, of Lon- perils. One of them is that it is not easy to but soon you see that it’s great. don’s Imperial College, are also sceptical. work with outsiders. Corporate cultures So how do you know if open innova- They argue that rms have always been can sometimes clash and some outsiders tion will work for a particular company? It open to some degree and that the bene ts are not used to working in a business envi- may well depend not just on what a com- di er depending on their line of business. ronment. For example, P&G has a co-in- pany does but also on how it is perceived Those using older technologies, for in- vention lab with BASF, a German chemi- in the market. Hal Sirkin, of the Boston stance, may bene t less. They also point cals giant with its own strong Consulting Group, suggests that rather out that the costs of open innovation, in corporate-culture. Bo ns from the Ameri- than see rms like P&G and IBM as truly management distraction or lost intellec- can government’s prestigious Los Alamos open innovators, it is better to view them tual-property rights, are not nearly as well national laboratory also sit in on some of as beacons . They have enough world- studied as its putative bene ts. P&G’s research-planning sessions. The class experts working for them to attract Yet another critique comes from capi- consumer products rm believes that the outsiders who have brilliant ideas. Such tal-intensive industries, where products bene ts of working with people from such rms are open in the sense that they are take a long time to develop and remain on diverse organisations are worth the e ort. now casting a very wide net in their search sale for years. Toyota’s Mr Reinert laughs For one thing, patents are becoming for ideas. However, once they have cap- when asked about open innovation. With much less important nowadays than tured the essence of those ideas, argues Mr the billions of dollars his rm spends on brands and the speed at which products Sirkin, they control them and the process research and on equipping its factories can be got to market. It is true that some of of getting them to market. not to mention a ve-year product-de- the rising stars in developing economies velopment cycle he suggests it would be are beginning to take out more patents, but At your service foolish to open up and allow rivals to steal many of their innovations are still kept On a summer day in east London, a ware- its edge. Eventually even Google will quiet as trade secrets. So uid are their house was taken over by a company eager have to make something tangible, and markets, and so weak the historical patent- to make a splash. It was decked out to look when they do they will protect it just like protection in them, that bosses often prefer like a cool New York loft. The Ministry of Tesla Motors, which does not have an to keep things in the dark and come up Sound, a London nightclub, was hired for a open model, he adds. with the next innovation as necessary to party afterwards. The event was packed1 10 A special report on innovation The Economist October 13th 2007
2 with journalists. At last the stars took to the Mr Von Hippel thinks that rms that are GM to innovate. Now OnStar can check if a stage a group of besuited Nokia execu- close to their lead users can come up with car is working properly, open the doors for tives there to announce a dramatic change much better designs for new products and a driver who accidentally locks the keys in- in corporate strategy. get them to market faster. This advice ap- side and even locate the nearest pizza Nokia, a Finnish company, makes mo- pears to contradict what Harvard’s Mr place. GM’s Larry Burns believes OnStar bile-phone handsets which are used by Christensen says, but in fact the two theses helps to improve his rm’s brand loyalty nearly a billion people around the world. are compatible. Mr Christensen’s point is because it keeps the company in constant However, it now wants to be a services that rms should not uncritically cater to touch with its customers. rm. Why? Niklas Savander, of Nokia, ar- the demands of their most pro table cur- Richard Lyons, of Goldman Sachs, of- gues that the mobile-phone business is rent customers. They must question those fers the most compelling argument for moving so rapidly, thanks to the democ- demands or they could end up doing little rms to think hard about recruiting users ratisation of the internet, that we must in- more than gold-plating their current o er- to speed up and improve their innovation novate or die. Providing people with de- ings; like Mr Von Hippel, he thinks rms e orts. In rich countries about four- fths vices alone is not enough, the company should keep a closer watch on new and of economic activity now involves ser- has concluded. dissatis ed users, who are much more vices, but pro t margins are eroding. He ar- With half of the value and most of the likely to be the source of disruptive ideas. gues in a new paper that commoditisa- innovation in a mobile-phone handset tion often occurs even faster in services now made up of software, the leap to ser- Invented on Facebook than in physical products , because inno- vices is not so great as it seems, he adds. Mr Von Hippel adds that networks of vations are easier to copy, patents can pro- Nokia is now rolling out Ovi, a branded hyper-critical users can even help rms vide less protection, up-front costs are service o ering users networked gaming, quickly lter out bad ideas and thus en- lower and product cycles are shorter. music downloads and other services from courage the process of fast failing. The For a business that uses open and net- their handsets. craze for social networking sites, like Face- worked innovation, it matters less where Visionary companies need to do even book and MySpace, could be useful. Sinan ideas are invented. Managers need to focus more than that, argues a report by C.K. Aral, of the Stern business school at New on extracting value from ideas, wherever Prahalad and Venkataram Ramaswamy, York University, argues that how people they come from. Unfortunately govern- two academics at the University of Michi- relate to the products they use (something ment planners, who are often obsessed gan. They think rms should cultivate a often discussed on such sites) reveals a so- with national innovation policies and the network that includes consumers in which cial structure and preferences. That can need to create clusters like Silicon Valley, personalised, evolvable experiences are help rms understand more about their have not learnt this lesson. History also the goal, and products and services evolve customers and how to market products shows that countries that come up with as a means to that end. That sounds u y more e ectively. new technologies are often not the ones enough to have come straight from the User networks operate in many busi- that commercialise or popularise those in- Summer of Love. nesses. OnStar, a mobile-information sys- ventions. Richard Halkett, of Nesta, a Brit- Yet despite the dangers, some compa- tem widely launched by GM in 2000, was ish research body devoted to innovation nies have successfully brought consumers initially meant only to provide safety and policy, jokes that the right policy for gov- and others into the innovation process. emergency services for drivers. But motor- ernments should be never invented Lego, the Danish maker of children’s ists wanted it to do more, and they pushed here . He may be right. 7 building blocks did it; and it helped revive the company. In uenced by research done at MIT on how children learn, Lego launched Mindstorms, a robotics kit that allows people to design their own robots and other devices. Numerous websites have popped up as users including many adults come up with all sorts of ways of putting together the kit to make things ranging from intruder alarms, sorting ma- chines and even the controls for small un- manned aircraft. Eric Von Hippel, of MIT, has long advo- cated user-driven innovation. He says you can see it all around you. Users who feel passionate about certain products often ddle around with them because they fail to provide exactly what they want. It might be a mountain bike, a kayak or even a car. He reckons open innovation misses the point if it is not inspired by users, because companies are then just talking about a market for intellectual property rights, it’s still the old model. The Economist October 13th 2007 A special report on innovation 11
The fading lustre of clusters
The best thing that governments can do to encourage innovation is get out of the way
HE scene in Salzburg earlier this year Twas one that Joseph Schumpeter, an economist obsessed with innovation, and Mr Drucker, the management expert, would surely have approved of. Several dozen government o cials and academ- ics from around the world gathered at Schloss Leopoldskron, a spectacular ro- coco palace located on the shores of an idyllic lake. They came not for the fresh Al- pine air, hearty Austrian fare or even the hills alive with music. It was for a confer- ence organised by the Salzburg Global Seminar, a non-government organisation, to discuss what they could do to turn their economies into innovation powerhouses. Holding such a meeting in the heart of Europe seemed only tting and not just globally disruptive innovation in de- only about 6.4 billion ($9 billion) in the because the two great theorists of innova- cades although Skype, an internet-tele- EU, while their American counterparts tion both hailed from the region. After all, phony rm that is now part of eBay, once splashed out some $45 billion on new ven- it was a European, France’s Georges Do- looked like it might qualify. tures. The link between venture capital riot, who invented venture capital during Europe’s innovation malaise is the re- and innovation is a strong one. Samuel his time teaching at Harvard. And it was sult of a complex mix of factors. Some Kortum and Josh Lerner, two American another Frenchman, Jean-Baptiste Say, places, like Ireland, Finland and parts of academics, have shown that a dollar of who coined the word entrepreneur two Scandinavia, do better than others. And venture capital could be up to ten times centuries ago to describe the plucky up- Cambridge, England, can reasonably more e ective in stimulating patenting start who shifts economic resources out claim to have created Europe’s best inno- than a dollar of traditional corporate of an area of lower and into an area of vation cluster, albeit one that falls far short R&D