When it comes to mobile advertising, there is always a question regarding attribution methodology. In the mobile arena, attribution refers to the measurement of user events (such as an app install, repeat app launch, level completion, or in-app purchase) that are a result of marketing activity.
Every mobile marketer wants to measure the effectiveness of their marketing campaign, especially as it involves spending money on advertising. However, attribution is a problem particularly for the mobile arena because there is no existing standard for measurement methodology. So how is a mobile developer/marketer to accurately measure the performance of their app/campaigns without well-trusted mobile app attribution methods?
Traditional Online Advertising Attribution
In the online advertising world, marketing attribution already has its tried-and-true methods: cookies, image pixel tags, and appending custom parameters to the URL. Therefore, it is fairly easy to set up integrated marketing campaigns to determine which visitor came from which source.
Cookies have been around since the dawn of the web and are simply pieces of code that web servers use to put information in client browsers (the web servers then retrieve that information at a later time for various uses, such as maintaining session state for users).
However, using cookies does have some limitations including:
- Some browsers set restrictions on cookies (such as a byte limit of 4096 bytes)
- Users can disable their cookies in their browser
- Cookies expire, which can lead to inconsistencies in attribution
- Some sites limit the number of cookies allowed per domain
Pixel tags are small blocks of code embedded in a web page that are invisible but allow the logging of user activity. When users view a web page, the browser downloads the pixel image and other information. This download involves the browser requesting the pixel image from the server that is hosting/storing/serving it, causing the server to take notice of (and log) the download request. As a result, the organization running the server knows when users view a web page.
Conversion measurement with pixel tags involves placing an iFrame or image pixel on the confirmation page for a campaign. When users see an ad and click on the measurement URL for the campaign, the measurement URL records the click, sets a cookie on the client computer, and redirects the user to the campaign URL. If the user continues, buys something, and proceeds to a confirmation page, then the web server displays the iFrame or image pixel, the client browser sends the cookie information to the Attribution Analytics platform, which logs a conversion.
Appending Custom Parameters to a URL
Through the measurement URL of a campaign conversion page, statistics on users who click on the measurement URL and complete a conversion (such as purchase, sign-up, page view, or lead) can be collected and viewed in your reports.
For example, you can run multiple campaigns (such as a Promoted Tweets on Twitter and Google AdWords campaign) on Google with the goal of driving all visitors to download a new white paper document on your web site.
To determine which visitors came from which source, you simply append parameters to each measurement URL to log the requests from each channel (or campaign). So to attribute users from Twitter, you send them to: http://www.example.com/resources/?source=twitter.
By appending the custom parameter value pair "source=twitter/facebook/google", you can easily see how many visitors came from each source and also what their additional interactions (or engagements) are with your entire web site. Therefore, by appending additional parameters to your URL, you can easily collect more information (and specify what in particular to collect) about your referral traffic to gain deeper insight into how your campaigns are performing.
As you can see, the accessibility and ease-of-use for measuring online marketing activities makes attribution relatively simple for online marketers. Marketers can attribute every dollar spent to specific return-on-investment (ROI) for that dollar.
Mobile App Attribution
Because the mobile advertising industry lacks an existing standard, most mobile marketers face a monumental problem: how to measure the effectiveness of their marketing campaigns across the mobile space! To make matters more difficult, the two main device platforms (iOS and Android) vary slightly in the attribution methods used and the sequence in which those methods are executed.
Google makes life for the mobile app advertiser easier by using a referrer URL parameter in download links to the Google Play™ Store. The Google Analytics SDK for Android uses this parameter to automatically populate campaign information in Google Analytics for your app, which enables the source of an install to be recorded and associated with future page views and events (for assessing the effectiveness of a particular ad for your app). For information about the Google Install Referrer, please visit Main Methodologies for Attribution.
In addition, for any campaign run through Google Play, you can take advantage of the Google Campaign Tracking URL Builder, which generates a measurement URL (based on your own app info) that refers users to your specific app and the subsequent referral information is later available in your Google Analytics report. For information about the Google Campaign Tracking URL Builder, please visit Campaign Tracking URL Builder.
The Google Install Referrer parameter is the primary (and therefore standard) method of attribution for all Android devices. However, if you rely completely on Google Analytics for your attribution analytics, then you can't send conversion data to your advertising partners (publishers). And if your partners don't receive any attribution data, how are they to optimize their campaigns for your app? Answer is, they can’t, and this is a big problem!
If you're advertising a mobile app for the iOS platform, then after users click into the Apple® iTunes® App Store environment, their actions essentially enter a 'black hole' where you no longer have any insight into the performance of your app.
Without a unique attribution mechanism similar to the Google Install Referrer URL, the Apple iTunes App Store requires you to rely on other attribution methods if you want any insight into the performance of your app.
Universal Mobile App Attribution Methods
If users are operating devices other than Android, or you simply want to use a universal attribution method, then implement one of the following attribution methods.
Unique Identifier Matching
Attribution Analytics performs unique identifier matching by matching the unique identifiers from the install to a click. Unique identifier matching is an automated and real-time method of comparing clicks to installs instead of manually performing the match at the end of a week or month using Excel spreadsheets. Attribution using unique identifiers enables 1:1 accuracy of matching clicks to installs (when identifiers are passed app-to-app).
For more information about identifier matching, please visit Main Methodologies for Attribution. For more information about unique identifiers for attribution, please visit Unique Identifiers for Attribution.
Device fingerprinting pulls basic (and not always 100% unique) information (such as IP address) from mobile device headers to connect a user from ad-click to app install (or some other conversion). Device fingerprint matching works by redirecting users through a measurement URL and collecting the publicly available HTTP headers about the device. Attribution Analytics uses this information to create a unique fingerprint about the device (whose user clicked a measurement URL). When a user installs the mobile app, the Attribution Analytics SDK collects the same data points from within the mobile app and sends it to the Attribution Analytics platform, which searches for clicks with the matching fingerprint of the install. Attribution Analytics attributes the install to the device that had the last click with a matching fingerprint. While you use identifier matching for app-to-app attribution, you use device fingerprint matching for web-to-app attribution because the publishing app in app-to-app scenarios cannot actually pass device identifiers. For more information about device fingerprint matching, please visit Main Methodologies for Attribution.
Open URL with Click ID
In lieu of an unique identifier, like Apple’s IFA, we support attribution based on a Tracking ID. Conversion Attribution with ID from Click (a technique borrowed from our cookie-less web site measurement) works by making a conversion request with a Tracking ID generated on click. By passing in the Tracking ID generated on click into the conversion request for an install or event, the conversion is then attributed to the advertising partner (publisher) associated with the click. Several of our top integrated advertising partners notify us of clicks and/or impressions server-side. However, for those of you that use server-side measurement, waiting for us to respond with our Attribution Analytics click ID to then include in the Google Play install referrer for attribution is not an option, nor is it possible to include our Attribution Analytics click ID in the open URL of both Android and iOS apps for attribution of events with re-engagement. Instead of relying on a Tracking ID (Attribution Analytics click ID) that we generate, our attribution process supports using a third-party Tracking ID. To take advantage of this, simply pass your own value into our tracking_id parameter.
The Open URL with Click ID method provides 1:1 accuracy for attribution of events. It works for both app-to-app and web-to-app scenarios to attribute on a click-through basis by using the URL (deep link) that initiated the “app open”. This method does not work for install attribution because on the first “app open” after the install, the deep link is inaccessible.
For more information about using an Open URL with Click ID, please visit Main Methodologies for Attribution.
Unified Attribution Solution
As the mobile space experiences growing pains, mobile app developers and advertisers are looking to implement analytic technologies that can support several attribution methods (to ensure that all sources of your paid media and advertisements are accurately measured and attributed). Using a single technology that provides this multi-faceted approach is key to making your marketing spend as effective and efficient as possible.
Attribution Analytics combines these different attribution methods into a single unified solution that not only seamlessly attributes app-to-app installs, but also mobile web-to-app installs.
With our unified attribution solution, partners can rely on and trust Attribution Analytics to attribute installs accurately based on a wide range of available data. The Attribution Analytics technology supports a broad range of unique identifiers that it can use for attribution by unique identifier matching.
You can specify one or all of these unique identifiers on click and on conversion, depending on the app platform (such as Apple iTunes App Store, Google Play Store, or Amazon Appstore for Android™). Attribution Analytics evaluates each request and determines the most accurate method to attribute a conversion. If attribution based on unique identifiers is unsuccessful or no unique identifiers are specified on click, then Attribution Analytics automatically uses device fingerprint matching. For apps downloaded through the Google Play Store, an install referrer string (in our case, a tracking ID) is passed on click and later, gets passed back to use for conversion attribution.
Because Attribution Analytics completes the entire mobile attribution process (measuring of the initial click, attributing the click to install, and notifying partners of installs and events through server postback URLs), partners can pass their ID (that references their ad clicks) into our “ref_id” parameter. This reference id value can be unique to the partner (such as an impression id, click id, or tracking id). On install, when Attribution Analytics attributes the install to the partner, Attribution Analytics can include the reference id for the partner so they can easily match the install to its corresponding click for the reporting and optimization in their system.