Overview

Attribution

In Topsort, we specialize in doing transparent and precise advertising, that makes the advertisers feel not only empowered by their ad campaigns, but also safe that their corresponding metrics are reliable and trustworthy. As an integral part of this, Topsort has an advanced Ad Purchases Attribution system that is in charge of detecting which purchases are due to customers interacting with the ads powered by Topsort. This is a fundamental piece of the Topsort ads ecosystem as well, since it provides a feedback loop for both the advertisers and the Topsort Bidless algorithm that is in charge of optimizing the ads campaigns on behalf of the advertisers.

Topsort offers flexible attribution logic, allowing marketplaces to define how they want to track and measure the effectiveness of their campaigns. Marketplaces can choose between different attributing events, such as clicks or impressions, to indicate user interactions that are tied to a specific campaign. While clicks and impressions are directly linked to ads, Topsort typically uses attributable events —such as purchases— that may not be explicitly tied to a campaign but are crucial for understanding conversion outcomes. Through Topsort’s APIs, advertisers can track and link both attributing and attributable events, ensuring a complete view of the campaign’s performance and its role in driving conversions.

A key aspect of Topsort’s flexibility is that marketplaces can determine their own attribution logic, deciding whether clicks or impressions serve as the attributing event, and defining the attribution window — the time frame in which user actions, like purchases, can be attributed to a campaign. By offering this level of customization, Topsort empowers marketplaces to tailor attribution models to fit their unique needs and performance goals, whether they want to optimize for awareness, engagement, or direct conversions.

Attribution types

Click attribution

Click attribution is one of the most commonly used mechanisms in retail media. The main idea is that in order to attribute a purchase to an ad, there must have been a click to that ad. The purchase must have occurred after the click to the ad, and within a predefined window of time (normally, at most 14 days after the click).

Impression attribution

Similar to click attribution, impression-based attribution requires the customer to have seen an ad before making the purchase in order to attribute this to the ad campaign. This is a much more flexible model that can report higher gains for the advertiser. This model is particularly suitable for banners, that are often displayed on the homepage of websites and apps.

Halo attribution

The two models described above rely on linking interactions between the customer and an ad of the product that was ultimately purchased. While this seems the most intuitive way of doing attribution, they not always capture the value delivered to the advertiser to its full extent.

Suppose for instance that a brand makes an ad campaign about their brand new product. This attracts people attention, but they are not necessarily willing to try this new product; instead they opt for the old ones that they know and like. While the ad did not directly create a lift of the sales of the advertised product, the advertiser did experience a lift on their sales.

A similar situation occurs in marketplaces where sellers have their shops. Having an ad for a specific product is very often a gateway for customers to visit and explore their whole shops, creating a lift in revenue that is not directly seen through the specific advertised product. These gains are not always captured by Click/Impression-attribution.

Considering these scenarios is that we have developed a Halo Attribution model: with this model, we are able to capture the value provided by ads to the vendor through the purchase of any of their products after the customer interacts with the ads.

The Halo Attribution mechanism is able to report to the advertiser exactly what are the products that experience revenue lifts from the ads, despite them not being exactly the target of the ads seen by the customer.

Configuration

The default configuration is Click attribution but this can be modified depending on marketplaces’ needs. The default attribution window is 14 days, but this parameter can also be configured by the marketplace.