Mobile Attribution Model

What is a mobile attribution model?

A mobile attribution model is the accepted methodology in an organization of attributing paid installs to ads.

The chosen model greatly impacts which ads, channels, and networks get credited with a paid install.

There are a few different attribution models:

  • Last-touch attribution: gives the last ad the user clicked before installing 100% of the credit.
  • Multi-touch attribution: assigns some of the credit of the installs to all ads that a user was exposed to during his journey to installing. Usually, less credit is given to earlier ads than the last ad (which gets the majority of the credit for the install).
  • View-through attribution: assigns a value to an ad even if a user didn’t click on it but just viewed it.
  • First-touch attribution: assigns 100% of the credit for the install to the first ad the user interacted with. Although it only tells part of the story, it points out the early ads that led to an install.
  • U-Shaped/positioned-based attribution: in this model, 40% of the credit is usually assigned to the first-touch ad, 40% to the last-touch one, and 20% is divided between all the ‘middle’ touch points. This reflects the belief that the first and last touches are the most important.
  • W-shaped attribution: similar to the U-shaped model, it assigns the most weight to the last and first touches, but it also assigns significant value to a middle touchpoint where the user became a lead. This usually happens somewhere in the middle of the journey (let’s say a user signed up to a newsletter after engaging with an ad).

In the mobile ecosystem, with the exclusion of Facebook and Google that use view-through attribution (perhaps a result of their attribution bias), the de-facto standard is last-touch attribution. This is because of (among other reasons) the difficulty of using other more complex models.

Why a mobile attribution model is important?

The choice of a mobile attribution model has a huge effect on the “signals” it produces to the marketing team. At the end of the day, the goal of attribution is to provide the best insights for the marketing budget makers to allocate their budgets efficiently.

After attributing installs to specific channels and ads, a clear picture emerges that shows the quality of these channels and ads in the form of user-quality metrics such as retention, lifetime value, and ROAS.

Mobile attribution model and App Store growth

Mobile attribution has a clear and large gap: organic installs. The mobile attribution model is dependent on tracking user identifiers after they clicked on an ad and eventually installed an app.

In the cases in which the user wasn’t tracked (didn’t click on an ad, turned off ad tracking, etc.) the user will be categorized as organic on the attribution platform. Thinking about the goal of attribution (to generate actionable insights to the marketing team), this solution doesn’t solve the organic side of attribution; hence, there are no actionable insights for increasing organic installs that stem from these models in the current mobile ecosystem.

You May Also Like…

Influencer Marketing

Influencer Marketing

What is influencer marketing? Influencer marketing is the practice of identifying, approaching, and (usually) sponsoring category influencers to promote a certain app or game. In certain categories, there are influencers that hold a very high share of voice for that...

Monthly Active Users (MAU)

Monthly Active Users (MAU)

What are monthly active users (MAU)? Monthly active users (MAU) are users that launched the app at least once over the course of a calendar month. Why monthly active users (MAU) are important? As growth practitioners, the ultimate goal isn’t just app installs but app...

Daily Active Users (DAU)

Daily Active Users (DAU)

What are daily active users? Daily active users (DAU) are the number of users who launched an app at least once in a given day. Why daily active users (DAU) are important As growth practitioners, the ultimate goal isn’t just app installs but app growth, usually...