Attribution in retail media advertising identifies and credits the marketing and advertising efforts that influence a consumer's purchase decision. Epsilon Retail Media measures sales attribution to pinpoint the parts of your retail media activity that led to purchases by the targeted shoppers. Our attribution methodology includes:

• User matching: This connects purchases made by shoppers who interacted with your ads, effectively linking ad interactions to later purchase behavior.
• Comprehensive attribution: Achieve precise attribution by matching purchased product SKUs with customer ad views. This unified process guarantees accurate credit for your product-specific campaigns when purchases align with them.
• Lookback window: Purchases made within a specific period after an ad is clicked or seen are attributed accordingly. The standard look-back window varies depending on the campaign type. It could vary between 14 to 30 days. For more information, see Types of Attribution Models in Epsilon Retail Media.

Attribution setup components

A successful attribution setup consists of the following four components:

Catalog setup

Description: Catalogs serve as directories listing all products at the retailer product code level. Each product entry includes essential attributes such as brand, category, and image URL. For more information on setting up a catalog, see Introduction to Catalogs and Products.

Purpose: This data enables the accurate identification and classification of products during purchase events. When creating campaigns, advertisers select the exact products they wish to advertise. Properly structured catalogs facilitate precise attribution by linking purchases to specific product information.

Campaign setup

Description: Campaigns are used to display product advertisements on a retailer's website. These campaigns are configured with targeting criteria and bidding strategies to optimize ad exposure and performance. For more information on setting up a campaign, see Create a Campaign for Retailers.

Purpose: Clicks and impressions generated by these ads are crucial metrics for calculating attribution, helping to determine the effectiveness of ad placements.

Ad impressions and clicks

Description: These are interactions recorded when an end user views (impressions) or clicks on an advertisement, creating measurable touchpoints.

Purpose: These touchpoints are key components in attribution models, as they track the user's interaction path leading to a purchase, enabling marketers to understand the customer journey.

Purchase event stream

Description: This component involves a continuous flow of transaction data sent to Epsilon Retail Media's purchase event endpoint. The system aims to match these events with corresponding touchpoints from the same user.

Purpose: By correlating purchase events with prior ad interactions, this stream facilitates the attribution process.