Mastering Mobile Attribution

"picture of person holding phone with network on it to indicate that they are using Facebook Audience Network"

In our March webinar, “Mastering Mobile Attribution,” Katana partnered with James Benoit, the Manager of Technical Partnerships at Kochava, to share actionable tips to help marketers accurately measure the return on their mobile marketing investments. Kochava, a third party independent mobile attribution platform, was recently announced as a Snapchat Audience Match API Partner.
In 2016, US mobile advertising spend reached $40 billion. With consumers increasingly relying on mobile devices to make purchase decisions offline, it only makes sense for marketers to want to quantify the influence that their mobile channels are making. Mobile attribution platforms allow you to optimize your expense to make sure you’re getting the most out of every dollar spent. If you plan to do any amount of user acquisition or have any amount of ad spend to put behind your product or service, attribution should be the first thing you’re thinking about.

The State of Mobile Attribution


In the mobile arena, attribution is defined as the measurement of click and impression signals from ads that are served on the web or in-app. Essentially, it’s the matching or attribution of engagements with an ad unit to an app install, or some post-install activity that’s being measured.

Online advertising attribution historically uses varying degrees of cookie matching, image pixel tags and custom URL parameters. However, the mobile advertising industry lacks an existing industry standard and has to account for both the iOS and Android operating systems. As a result, advertisers can choose to cater mobile attribution to specific operating systems or approach mobile attribution at a universal scale with the help of third party attribution vendors.
ANDROID MOBILE ATTRIBUTION: For Android devices, Google simplifies the attribution process by associating a referrer URL parameter in download links to the Google Play™ Store. This information is then transferred to the Google Analytics SDK (software development kit) for Android, and campaign data is automatically populated in Google Analytics for the advertiser’s app. Through Google Play, marketers can also take advantage of the Google Campaign Tracking URL Builder to generate a measurement URL that refers a user to a certain app. However, the primary method of attribution for Android is the Google Install Referrer parameter. Unfortunately, if Google Analytics is your main source of attribution analytics, then you won’t be able to send conversion data to advertising partners in real time (which poses a big problem because timely optimizations are contingent upon real-time data).

iOS ATTRIBUTION: Unlike Android devices, iOS operating systems lack any primary parameters to award attribution once a user clicks into the Apple® iTunes® App Store environment. Essentially, any action a user makes in the App Store becomes lost in a ‘black hole,’ consequently blocking marketers out from any insight into the performance of their app (creating what is known as a walled garden). In order to remedy this, marketers have to set up unique attribution mechanisms and integrate with third party attribution platforms to fill the void.

Marketers can track the data they want to track by using the Apple-regulated IDFA (ID for Advertisers) or by fingerprinting a user’s device with a combination of basic browser information and IP address data.


There are varying levels of sophistication in the way that companies or players in different verticals have evolved to use the data they have access to when using a mobile attribution platform. Depending on the vertical-specific industry, the engagement metrics measured will vary. For example:

Basic In-App Events to Measure

      • Install
      • Conversion Rate
      • Revenue
      • Average Revenue Per User (APRU)
      • Number of purchases

Vertical-Specific Industry In-App Events to Measure

Gaming is at the forefront of sophistication, focused on re-engaging users and maintaining engagement over time.

  • Gaming might procure the most ‘purchase events’ if we’re talking about free-to-play games.
  • In-App Events: tutorial completion, Facebook registration, achievement unlocked, level passed

Travel marketers should be focused on enriching and defining parameters of each in-app event in order to acquire new users and re-engage loyal consumers.

  • In-App Events: hotel/flight packages viewed or searched, wish-list items, registration, initiated checkout, hotel/flight packages booked or login

eCommerce is sophisticated in running engagement campaigns to measure a defined set of post-app install activity.

    • In commerce or travel, you may be talking about users that come back time and time again to make purchases.
    • According to a PayPal-Ipsus research study, mCommerce is growing at 3X the rate of eCommerce. Sessions on shopping apps on iOS and Android devices are increasing by 174% YoY.
    • In-App Events: product viewed, product added to cart, registration, product purchases, login

Why Marketers Should Be Concerned with Real Time Mobile Attribution Platforms

Many integrated partners, whether it’s a publisher, ad network or demand-side platform, rely on a signal that comes from a real-time mobile attribution platform to inform their system that attribution was awarded.

It is extremely important that a system maintains real-time status, so that all sources are notified in the moment when attribution has been awarded. This ensures that if you have a really strong source, it’s not impeded by data delays, or post-back delays. If you’re buying at scale, this is crucial because a lot of money can be wasted in a very short amount of time if there are interruptions to those signals.

Advancements to Technology for Mobile Attribution

Traditionally, fingerprinting technology suffered from mixed accuracy and match rates. According to Benoit, the flavors of fingerprinting have evolved from solely deterministic methods to more sophisticated methods that leverage statistical modeling.

Fingerprinting is a device recognition and identification method that leverages public parameters, such as device names, types, versions, IP addresses, platforms and/or carrier. These variables are then used to create a digital fingerprint ID that logically matches specific device attributes and features, including OS type, version and device model.

What to know about fingerprinting:

  • Fallback method: Since fingerprinting uses a probabilistic model, it’s only used when deterministic identifiers, such as Google Play Referrer or other device identifiers aren’t available for attribution.
  • Highly accurate in short term: The bigger the gap between a click and an install, the greater the chance that the user has time to change their fingerprint. The attribution window is usually short (within 24 hours.) In app install campaigns, the majority of clicks, installs and initial app opens happen within 1-2 hours.
  • Used primarily for iOS: The referrer method isn’t available for iOS, so fingerprinting is used more often on Apple devices.
  • Anonymous: User fingerprints are privacy compliant.

Wrapping Up

Mobile attribution is not just about measuring acquired users; it’s about measuring engagement and acquisition immediately and over time. Engagement isn’t something marketers should think about after six months has gone by and they’ve acquired 20,000 users.

With so many apps on the market, it’s important to think about the different insights mobile measurement offers. For example, a brand focused solely on acquisition might appear successful upon first glance at the attribution data. Once analyzed, mobile measurement can tell a different story, conveying paint points, such as poor user interaction, lifetime usage or a significant drop-off between channels.

Tips to Get Started with Mobile Attribution

  1. Ensure the platform your brand uses is in real time.
  2. In terms of mode, evaluate SDK (attribution analytics tools that track the sources of your mobile app paid installs or post-install paid events) load, and consider adding another third party indicator to that.
  3. Understand the different avenues of integration.
  4. Plan out the KPI’s for your app or product, taking into account which type of events your brand wants to measure.
  5. Know your tool set and product so that you can run effectively at scale.

If you want to learn more about how you can measure the impact of your mobile marketing efforts, contact or find the full webinar below.