What is Attribution?
What is Attribution?
When you use any application — whether it's a mobile app or a website — you often come across ads or sponsored content. These ads are one of the primary ways applications and websites generate revenue. They allow platforms to offer their services for free or at a reduced cost while advertisers get visibility for their products or services.
Now, ads have often been criticized as an intrusion of privacy, and that may have been true a few years ago. But thanks to compliance policies like GDPR (General Data Protection Regulation), DMA (Digital Markets Act), ATT (App Tracking Transparency), and LQDN (La Quadrature du Net's advocacy efforts), the extent of tracking in advertising has significantly reduced. Today, you have a choice — you can either consent to tracking or refuse it. This control empowers you to protect your data. And if you prefer to see ads that are more relevant to your interests, you can consent to that as well. The key difference from the past is that the choice now rests with you, the user.
When ads are shown to you, your interaction with them is tracked through two major events:
Impressions — An impression is recorded when an ad is viewed on your screen. It doesn't require any action from you; simply seeing the ad counts.
Clicks — A click is recorded when you actively engage with the ad and are led to another page, such as a product listing or a landing page.
These impressions and clicks are sent to the backend and stored in the application's database. They form the foundation of understanding how users are interacting with an ad.
Now, suppose after seeing or clicking on an ad, you go ahead and make a purchase, sign up for a newsletter, or complete some other desired action. This event is called a conversion. A conversion is the most significant event in the advertising lifecycle because it represents the actual outcome the advertiser was hoping for. Conversions are also recorded in the backend, just like impressions and clicks.
Once this data is collected, an attribution algorithm is applied to it. The goal of attribution is to determine which impressions and clicks played a role in driving the conversion. In other words, it answers the question: which ad interactions deserve credit for the final outcome?
Attribution algorithms come in several types, each with a different philosophy on how to assign credit. Here are the most common ones:
All of the credit is assigned to the last event before the conversion. For example, if a user saw three different ads but clicked on the last one right before making a purchase, that final click receives 100% of the credit. This model is simple and widely used, but it overlooks all the earlier interactions that may have influenced the user's decision.
The opposite of last touch — here, the very first interaction in the customer journey receives all the credit. If the first ad a user ever saw for a product was a banner ad, that banner gets full attribution for the eventual conversion. This model is useful for understanding which channels are best at generating initial awareness, but it ignores everything that happened afterwards.
This is a multi-touch attribution (MTA) model, meaning a conversion is credited to multiple touchpoints rather than just one. In linear attribution, the credit is distributed equally across all touchpoints. If a user interacted with four ads before converting, each ad receives 25% of the credit. It's a fair and balanced approach, though it doesn't account for the fact that some touchpoints may have been more influential than others.
Also a multi-touch model, but here the credit is distributed based on how close each touchpoint is to the conversion in time. Typically, the touchpoints closer to the conversion receive more weight, with the weight decaying as you go further back in time. The logic is that more recent interactions likely had a stronger influence on the final decision. However, the reverse can also be configured depending on the use case.
Also known as U-shaped attribution, this multi-touch model gives the most credit to two key moments in the customer journey — the first touch and the last touch — while distributing the remaining credit among the touchpoints in between. A common split is 40% to the first interaction, 40% to the last interaction, and 20% shared equally among all the middle touchpoints. The reasoning is that the first touch deserves credit for introducing the user to the brand, and the last touch deserves credit for driving the final conversion, while the interactions in between played a supporting but less decisive role. This model strikes a balance between first touch and last touch attribution, making it a popular choice for campaigns where both awareness and conversion are valued.
This is the most advanced approach. Instead of relying on fixed rules, data-driven attribution uses AI or machine learning models to analyze patterns across large volumes of data and assign credit based on the actual impact each touchpoint had on the conversion. It adapts to the unique behavior of your audience rather than applying a one-size-fits-all formula. While powerful, it typically requires a significant amount of data to produce reliable results.
Attribution is a critical methodology for measuring Return on Ad Spend (ROAS). It connects the dots between the money spent on advertising and the outcomes that money generated.
By looking at the ratio of conversions to impressions and clicks, advertisers can draw meaningful conclusions about the effectiveness of their campaigns. For instance:
If the number of impressions and clicks is high but conversions are low, it likely means the ad is reaching people but not the right people — the targeting may need to be refined.
If impressions and clicks themselves are low, it suggests that the ad's reach is limited. This could be a signal to increase the budget to generate more visibility, or to revisit the ad creative and placement strategy.
Conversely, if a small number of impressions is leading to a healthy number of conversions, the campaign is performing efficiently and the targeting is likely well-calibrated.
In essence, attribution turns raw ad interaction data into actionable insights, helping advertisers make smarter decisions about where to invest their budgets.
There are different types of ads, and depending on where a user sits in the marketing funnel, the objective of the campaign changes. We'll explore that in upcoming posts — stay tuned!