Mobile Attribution & Marketing Analytics for eCommerce
A Practical Guide on How to Let Data Work for You Mobile Attribution & Marketing Analytics for eCommerce
Introduction
Mobile commerce is on a tear. There’s no other way to put it. To name just one study from the many out there — PayPal-Ipsus research paper covering 22 countries found that mCommerce is growing at three times the rate of eCommerce.
Between 2013 and 2016 the compound annual growth rate for mCommerce is projected at 42 percent vs. 13% for eCommerce. To put growth in actual sales, mCommerce sales are expected to eclipse $290 billion worldwide by 2016.
With such numbers, it’s no wonder that usage of shopping apps is the fastest-growing app category, with sessions on shopping apps on iOS and Android devices increasing by 174% year-over-year, including 220% on Android alone.
That’s up from 77% last year as more and more retailers are building apps on top of their mobile-optimized sites despite the hefty price tag of developing and maintaining a commerce app.
www.appsflyer.com 2 Mobile Attribution & Marketing Analytics for eCommerce
Mobile Browser vs. App Funnel Metrics
Mobile Browser App
Products viewed per user x3.3 Products viewed per user
Basket x1.2 Basket
Purchase x0.8 Purchase
Conversion Rate x3.0 Conversion Rate
Source: Criteo, State of Mobile Commerce, Q2 2015
Apps also generated almost 50% of mobile transactions for retailers who have made their app experience a priority. In fact, mobile apps perform better than any other channel, including desktop.
Why are apps becoming a must-have touchpoint for retailers? Beyond the fact that mobile is dominated by apps (86% of time spent on mobile is on apps), there are several factors at play here: speed, optimal native mobile experience, ability to use mobile payments and push notifications, not to mention the impact on loyalty and lifetime value from a user who actually made a commitment to the brand by downloading its app.
www.appsflyer.com 3 Mobile Attribution & Marketing Analytics for eCommerce
Today’s savviest retailers (online & offline) are embracing big data as a major facilitator of business growth. Consumer information from anonymous shopping patterns to cross-channel loyalty programs is collected, sliced and diced to provide a more relevant and seamless consumer shopping experience - whether from a brick & mortar location, the website, mobile site or app.
A key communication channel merchants have with consumers on mobile is advertising, where relevancy and personalization play a big part. However, unlike the web, the ability to track, measure and ultimately optimize app install and retargeting campaigns is a challenge because of the mobile world’s inherent fragmentation (different OSs, different environments, different devices).
The good news is that advertising attribution and analytics platforms utilize sophisticated technology to connect the dots and provide retail app marketers with rich data to enable smart marketing decisions across the omni-channel user journey. This guide will cover the latest of what is possible in this area and how it applies to retail marketing.
Getting Started with Mobile Advertising Analytics? We recommend reading this first:
THE BEGINNERS GUIDE TO MOBILE ADVERTISING ANALYTICS
http://appsflyer.com/begin
www.appsflyer.com 4 Mobile Attribution & Marketing Analytics for eCommerce
A View From Above
Setting Goals
Defining Key In-App Events to Track
Relevant/ Smart Personalized Consumer Consumer Acquisition Communication
Retention & Cohort Analysis
The Bottom Line
Retention Uplift LTV Uplift ROI Uplift
www.appsflyer.com 5 Mobile Attribution & Marketing Analytics for eCommerce
Let’s Dive In
Defining Rich In-App Events that Matter to Your Business
A retail app, as you’re well aware, is one of the most complex apps to develop. The diversity ORDER NOW of in-app actions that can be taken by consumers is vast. The ability to track user behavior on a granular level and then view aggregated analytics is the basis of making smart marketing and advertising decisions across all acquisition and engagement channels.
It is therefore important that you clearly define your goals and pinpoint the in-app events that are directly related to these goals.
Here’s a list of key in-app events in retail apps:
• Login • Wishlist
• Registration • Add to Cart
• Content Viewed (product/ • Initiated Checkout category screens)
• Search String • Purchase
www.appsflyer.com 6 Mobile Attribution & Marketing Analytics for eCommerce
Diving deeper into the data involves one more crucial step: defining the right parameters of each in-app event - thus making it “rich.”
For example:
Purchase Content Viewed
Revenue Search String Content Type Price Content ID Currency Quantity Content Id Currency Quantify Customer User ID
So how can rich in-app events lead to the ability to meet specific business objectives? Well, if one of your goals is to increase the number of consumers who bought at least two electrical products, you’ll want to track the purchase event “enriched” with the quantity and content type parameters. If you want more female shoppers as active consumers, you can track the content viewed event in addition to content type, content ID, quantity and / or search string parameters.
www.appsflyer.com 7 Mobile Attribution & Marketing Analytics for eCommerce
Ultimately… by tracking these rich in-app events, you’ll be able to
Expand your knowledge about your users through deep 1 analytics
Set the path towards granular segmentation and enhanced 2 targeting with networks that are at the forefront of mobile targeting capabilities (those that are able to create advanced audience targeting campaigns after receiving rich in-app event data in real time from you or your tracking provider)
www.appsflyer.com 8 Mobile Attribution & Marketing Analytics for eCommerce
Smart Customer Acquisition
So far we’ve laid out our goals, defined the in-app events that matter to your business and understood why this data is so valuable.
Now let’s take a step back and think about what advertising attribution and analytics is all about from a user acquisition standpoint: The idea is to continue to measure what users are doing inside the app after they’ve installed it — particularly if these are rich in-app events.
The data is then aggregated to get an in-depth picture of your audience and which channels (paid or owned), ad networks, campaigns and/or creatives have led users you want — not just any users — to install the app.
Measuring Paid Acquisition
Here’s a table that shows which networks did well and which did not when aggregating rich in-app event data of users who added women fashion items to the cart.
Key takeaway: Networks 4 and 5 reign supreme, while Network 2 should be dropped.
www.appsflyer.com 9 Mobile Attribution & Marketing Analytics for eCommerce
Another example depicts Network 6 acquiring the largest number volume of users who logged-in and made at least two purchases within their first two weeks of activity.
Key takeaway: An interesting insight comes from Network 5: although it delivered the lowest number of users, in terms of the install to in-app action conversion rate, 8.48% is a great figure so an even bigger investment in this network makes sense.
Enabling Lookalike Targeting
By measuring in-app events you can get a clear picture of the types of consumers that generate the highest value to your brand. You can then use this knowledge to find users that “look like them” from a variety of sources by: 1) Increasing spend in a top performing network to maximize reach of its quality audience (make sure you don’t go overboard and end up targeting the same users) 2) Replicating the same targeting criteria used by the performing network with other networks 3) Utilizing the capabilities of the most advanced networks that support real-time data transfer of rich in-app events from a tracking provider and automatically turn it into a lookalike campaign
www.appsflyer.com 10 Mobile Attribution & Marketing Analytics for eCommerce
Measuring Activities on Owned Channels
Leveraging your own assets on owned channels, particularly emails and phone numbers of your existing segmented consumers who have yet to install your app, is something you’ll definitely want to measure, and compare to other channels — including paid ones.
Key takeaway: The advertiser's email campaign is doing very well compared to other channels with the highest conversion rate and the second highest average revenue per user.
www.appsflyer.com 11 Mobile Attribution & Marketing Analytics for eCommerce
Measuring the Consumer Journey with Multi-Touch Attribution
Another important layer to track in order to optimize user acquisition campaigns runs higher up in the funnel, and it involves the ability to go beyond last click attribution. Flawed as giving all the credit to the network that delivered the last click is, it’s the industry standard. However, it serves its purpose when looking from a pure billing perspective.
If a marketer is able to see the most common conversion paths that led users to install an app, he/she will continue to invest in these contributing networks. After all, these networks drove your users down the funnel and made them “sales-ready.”
Let’s explore the following scenarios:
Network A Network B Network C Network D 1,500 view click click last click installs
Network A Network B Network D 900 view click last click installs
Key takeaway: Network C plays a key role in driving installs
www.appsflyer.com 12 Mobile Attribution & Marketing Analytics for eCommerce
Measuring (True) Omni-Channel LTV
The proliferation of mobile devices, on top of web and brick & mortar touchpoints have created an amazing opportunity for omni-channel retailers. However, it has also generated a massive challenge of connecting the dots to provide a seamless and consistent user experience.
By using a server-to-server integration you can measure the actions acquired app users take across touchpoints (assuming they are logged in to enable matching via their consumer ID), giving you the true value generated by the acquiring ad network.
U.S. Cross-Device Share of eCommerce Transactions
% of users who used multiple User completed purchase on: devices in path to purchase 41% Smartphone
37% Desktop
43% Tablet
Source: Criteo, State of Mobile Commerce, Q2 2015
www.appsflyer.com 13 Mobile Attribution & Marketing Analytics for eCommerce
Let’s look at the following example to illustrate this point:
"I clicked on an ad that led me to install an app after which I've made 3 purchases"
Disconnect Connected
"When "Sadly, my it's all LTV is only connected I'm $50 worth $350!
Key takeaway: The value an ad network provides omni-channel retailers is often much higher than they think. Best of all, you can track it to see it.
www.appsflyer.com 14 Mobile Attribution & Marketing Analytics for eCommerce
Measuring the Impact of TV Ads
As app marketers drive user acquisition through DSPs, SSPs, Google AdMob, Facebook, and a whole slew of mobile channels, they’re realizing that mobile as a channel is becoming saturated, overly competitive and even limited in reach. One channel they are turning to for help is television.
The fact is TV is still very much a part of the fabric of our lives. According to Nielsen’s 2014 Q4 Report, there were 285 million TV viewers in the US so running a TV ad campaign guarantees access to a mass audience. Digital video platforms have a long way to go to catch up.
The good news is that even while watching TV, a viewer’s smart device is only an arm’s length away. In fact, Nielsen found that 84% of US viewers are watching TV with a second screen in hand. Sixty-four percent of viewers are using their smart device simultaneously. It’s a great opportunity to capture their attention to download an app on the spot or incentivize inactive users to take action.
View ad for 'Famous Footwear' mobile app
www.appsflyer.com 15 Mobile Attribution & Marketing Analytics for eCommerce
How does it work? Basically TV syncing providers offer detailed information on the exact time a TV ad was aired. A tracking provider correlates install data which is provided in real time. Credit is attributed to the appropriate TV network, assuming it occurred within a predetermined time frame after the ad ran (usually within minutes to a couple of hours at most). Advertisers have the ability to pre-define the attribution window for a TV ad campaign.
www.appsflyer.com 16 Mobile Attribution & Marketing Analytics for eCommerce
Measuring Shopping via Social Networks
A new form of online shopping that’s creating a lot of buzz as of late has to do with ‘Buy’ buttons, on social platforms — including Facebook, Instagram, Pinterest, Twitter and Google — have all announced plans to offer this service.
How does it work? A user enters his/her credit card number into the social app, along with his/her shipping details. When viewing an item of interest, he/she can click on the “Buy” button, which then informs integrated retailer partners of the shipping destination, he can click on the “Buy” button, which then informs integrated retailer partners where to ship it.
This form of one-click shopping has great potential and offers another way in which retailers can acquire new customers — although probably not new app users as it doesn’t involve any app install. However, the merchant can always encourage them to install the app once they are in his/her database.
In theory, when and if this functionality takes off, measurement is rather straightforward. The social app can send an in-app event to the analytics company (usually requires a special integration of its own to select providers), which is actually a deeplink with metadata on the actual purchase.
www.appsflyer.com 17 Mobile Attribution & Marketing Analytics for eCommerce
Relevant & Personalized Consumer Communication
The digital age consumer expects nothing short of a seamless experience as he/she constantly moves back and forth across online and offline brand touchpoints. Any disruption and you risk losing her to a sea of competitors eager to take your place.
The good news is that this consumer also perceives behavior— based communication as a service rather than an annoyance and is comfortable with sharing information in exchange for personalization, according to a newly-released and comprehensive mybuys report.
I'm comfortable having my shopping interests and behaviors used by retailers...
60% In order to expedite my shopping experience on site 60% So they can deliver relevant offers throughout the year 58% In return for a more personalized experience 58% As I believe I will receive a better overall experience
To coordinate a better shopping experience 50% across the channels I use
The report also found that 52% of consumers spend more money when they receive a truly personalized and consistent shopping experience.
www.appsflyer.com 18 Mobile Attribution & Marketing Analytics for eCommerce
Mobile, which is already playing a huge role across the consumer journey (both in the research and purchase stages), is no exception. Whether through email, coupons, push notifications or advertising, personalized and relevant communication on mobile devices is paramount to success.
www.appsflyer.com 19 Mobile Attribution & Marketing Analytics for eCommerce
Retargeting Enablement, Attribution and Analytics
First things first: a must-have element in a retargeting campaign on mobile is deeplinking as you’ll want to direct users to specific app screens. It’s a no-brainer.
Second, there’s the issue of transferring data to a retargeting network based on the level of integration you have or your analytics provider has with the network, as follows:
Static:
Data on standard in-app events can be transferred ‘manually’ to lists (for example, a list of IDFAs of users who viewed products but didn’t buy). Once a list is created, however, it cannot be changed.
Dynamic:
Here we have two types:
Basic integration: Send standard in-app events in real time to enable segmented retargeting. For example, users who added a 1 product to cart but didn’t buy. In most cases, these integrations can only support a name parameter (but at least a user in that segment is ‘dynamic’ and will automatically be part of a different segment once he/she purchases).
Advanced integration: Send rich in-app events in real time to enable personalized retargeting on a ‘segment of one’ (e.g. a user with 2 IDFA 1234 who viewed product 777 can be shown a creative with product 777 and relevant messaging; you can even bid higher for that user if, for instance, he/she has a high LTV). Currently, the number of networks that can support such an advanced integration is still limited but more and more networks are following suit.
www.appsflyer.com 20 Mobile Attribution & Marketing Analytics for eCommerce
When it comes to retargeting attribution, the mechanism is no different than an install. It simply means an advertiser defines a desired in- app event (e.g. app open for a dormant user, in-app purchase for an existing user) for which the attribution provider would credit a media source that delivered the last click that led to the desired action.
Retargeting attribution is often divided between “re-attributions” and “re-engagements.” The former represents users who re-installed the app after interacting with a retargeting campaign (assuming the IDFA/GAID remains the same), while the latter describes users who have the app installed and engaged with a retargeting campaign.
Once tracked, aggregated data of all users will help you pinpoint the best performing retargeting networks, campaigns and creatives in the same way that was illustrated in the acquisition campaigns above.
www.appsflyer.com 21 Mobile Attribution & Marketing Analytics for eCommerce
Measuring Location-Driven Retargeting Campaigns
According to a Google-Ipsos report, 4 in 5 consumers want ads customized to their city, zip code or immediate surroundings. The report also found that 50% of consumers who conducted a local search on their smartphone visited a store within a day while 18% actually bought a product. That’s why targeting consumers while on-the-go has tons of value.
Tracking location is possible by Lat/Long parameter. The advertiser can add these parameters to each in-app event which can be sent to the networks that support rich in-app events. You’ll then be able to understand where consumers take specific types of actions.
www.appsflyer.com 22 Mobile Attribution & Marketing Analytics for eCommerce
Measuring Bluetooth-Based Beacon Technology
With beacon technology, retailers can pinpoint a user’s location and send relevant content based on where consumers are. Since bluetooth connections actually work indoors unlike GPS, it captures an audience that’s highly likely to engage with the messaging while in the immediate vicinity of a store or, unlike GPS, in store.
Sending a push notification is ideal if the user has the app installed. If not, an SMS will also work well. Sending an ad while in-store can work by offering a user an immediate discount for installing the app (requires sending of real time attribution data to enable opening of the specific campaign app screen after deferred deeplinking).
Measuring a beacon-driven campaign - whether using Apple’s iBeacon or Google’s brand new platform-agnostic Eddystone platform and recording events in your server, - involves defining in-app events and reporting them to your analytics provider. You can then use cohort reports t o measure which group of users from which campaign had the highest engagement with beacon messages.
www.appsflyer.com 23 Mobile Attribution & Marketing Analytics for eCommerce
Measuring Direct Marketing Efforts
Direct communication with consumers mainly over owned channels is the key to a strong consumer-brand relationship. All you have to do to measure these campaigns is set up a tracking link and then compare your different channels — paid or owned.
Email Campaigns: According to research by Forrester, 42% of retailers' email messages were opened by consumers on their smartphones and 17% were opened on tablets. So you’ll definitely want to measure email engagement and more importantly what follows. With proper deeplinking in place, you can direct users to specific app screens from the mobile web or email app.
Push Notifications: Push has gone mainstream. That’s the key takeaway from the following highlights of a study by Urban Airship: • Retailers sent 34% more push notifications in 2014 compared to 2013 • Consumers’ engagement rate with push notifications during the 2014 holiday season doubled compared to 2013 • More than one-third of all retail app opens were the result of a push notification
Coupons: The number of adults who redeem coupons via a mobile device is expected to jump from 78.7 million to 104.1 million between 2014 and 2016, according to eMarketer. With such volume, it is important that retailers offer the best overall experience to optimize usage.
www.appsflyer.com 24 Mobile Attribution & Marketing Analytics for eCommerce
Creating a Unified User Experience for Optimal Results — Ideal for Direct Messaging It’s often challenging to create seamless communication in the fragmented mobile space with both prospects and customers. It can be achieved with these five components:
1) Smart Links: Advanced technology can make your life much easier by allowing you to configure a single tracking link per campaign for all sources. This smart link automatically detects where the user is and where he/she needs to be directed as he engages with the campaign. So instead of configuring at least five tracking links for iOS, Android, Windows, Amazon and web, all you need is one. That means there’s a lot less room for error — be it broken links or links that end up taking your users to the wrong place.
2) Deeplinking: Connecting the web-to-app and app-to-app environments, while supporting the opening of a specific app screen.
3) Deferred deeplinking: The ability to identify whether the user has the app installed on his/her device or not and if not, to take that user to the app store where he/she can download the app.
4) Real-time attribution data: Enables the opening of a specific landing page after deferred deeplinking and upon first app launch.
5) Customer User ID: Utilizing this this information to re-engage with the consumer in way that reflects his/her specific actions across the different channels.
www.appsflyer.com 25 Mobile Attribution & Marketing Analytics for eCommerce
Let’s explore two examples:
1) Email offer to a consumer who has the app installed:
Smartphone Gmail app iPhone product page in smartphone app Single Smart Link Tablet web iPhone product page in tablet app
Desktop Email promotion for web new iPhone on sale iPhone product page in desktop app
2) SMS to an existing consumer who does not have retailer’s app installed:
NEW COLLECTION! Download App Launch
SMS with deal on Redirect to the Specific landing new collection right app store page of new collection deals
www.appsflyer.com 26 Mobile Attribution & Marketing Analytics for eCommerce
How to Keep Consumers Coming Back for More
Retaining consumers to foster a lasting consumer-brand relationship and ultimately maximize their lifetime value is perhaps the most important goal for a retailer. The problem is that it’s also the most challenging… So what can you do to get retention right? There are two main strategies here: first, product-related. It goes without saying that your product must provide a top notch user experience. Any unnecessary friction and you risk losing the valuable consumers you worked so hard to acquire. Second, advertising-related, which is what we’ll focus on. Specifically we’re talking about retention and cohort analysis.
www.appsflyer.com 27 Mobile Attribution & Marketing Analytics for eCommerce
Retention Analysis
The more straightforward analysis involves the retention report. The retention value is the unique number of users who were active on a specific day/week out of the total number of unique users who first launched the app in the selected timeframe.
Depending on how deep the granularity offered by the analytics provider is, you can check the retention of users from a media source, campaign, ad group, country, city, OS version, device brand, device model etc.
Let’s explore the following case study of a leading retailer, broken down by campaign:
Key takeaway: • Campaigns 1 and 4 performed well across the board, retaining 12-14% of users on day 12. • Campaigns 2 and 3 underperformed from day 1, retaining only 5-7% of users on day 12.
www.appsflyer.com 28 Mobile Attribution & Marketing Analytics for eCommerce
Optimization performed on Campaigns 2 and 3:
• Budgets reduced • Targeting and messaging modified to better reflect Campaign 4 • Low performing ad sets within the campaigns were shut down
The data from a couple of weeks later clearly shows the advertiser’s efforts paid off:
www.appsflyer.com 29 Mobile Attribution & Marketing Analytics for eCommerce
Cohort Analysis
Retention is an indicator of how many users who launched the app during a set timeframe reopened the app in the days that followed. Cohort goes much deeper. Let’s take a step back to explore this strategy.
A cohort report enables you to group users with common characteristics and measure specific KPIs over different timeframes. It offers a really good indication of the quality of the average customer and whether it’s increasing or decreasing over time. The metric — whether sessions, revenue, or any other defined in-app event — is calculated by different timeframes, which represent the first X activity days per user, and then accumulated among all users (that’s why the graph never drops).
Here’s an example of how a cohort analysis can boost retention and lifetime value by increasing the average number of sessions per user. In this case, the cohort was grouped by the media source that acquired the user (other examples include campaign, publisher, ad group, geo etc.)
www.appsflyer.com 30 Mobile Attribution & Marketing Analytics for eCommerce
Cohort graph as seen on May 1:
Average Sessions Per User A B
C D
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Days
Key takeaways: • Network A starts off with a bang, producing high quality users that are highly active in the first 14 days after which the growth rate begins to plateau • Network B is also delivering quality users showing a modest yet steady climb throughout the period • Network C is slowly growing until day 7 but then flattens • Network D performed poorly - not only did it acquire low quality users that barely engaged with the app, it also did not show any improvement over time.
www.appsflyer.com 31 Mobile Attribution & Marketing Analytics for eCommerce
Optimization performed: • Having noticed the drop on day 14, the retailer began retargeting its users at that time to encourage them to continue engaging with the app • The budget of Network B was increased hoping it would generate more users who would only increase their engagement over time • Network C was broken down to the campaign and ad level and those that underperformed were removed • Network D was removed
Cohort graph as seen on June 1: A B
C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Days
Key takeaways: • Network A retargeting worked well as consumers continued engaging with the app at a growing rate beyond day 14 • Network B continued to deliver quality users (the increase is not seen in this graph as it shows the average sessions per user, but it can easily be viewed in the report that measures quantity) • Network C optimization worked well as the engagement growth rate improved compared to the previous month
www.appsflyer.com 32 Mobile Attribution & Marketing Analytics for eCommerce
To Sum Up
s more and more retailers are embracing mobile as a central if not the most important component of their strategy, they Aare moving beyond mobile-optimized sites to develop native apps. Despite the fragmentation and the challenges it creates, the mobile world is rapidly evolving (and that’s an understatement), catching up with the web when it comes to measurement, analytics and ad optimization capabilities. Retailers who are ready to take on this brave new world will reap the rewards.
www.appsflyer.com 33 Mobile Attribution & Marketing Analytics for eCommerce
Visit us at www.appsflyer.com
Copyright © 2015 AppsFlyer
www.appsflyer.com 34