Skip to main content

Retail Media

Particular Audience's Retail Media product offers a flexible and robust interface for on-site sponsored products and display advertising, suitable for any e-commerce platform or application.

All on-site sponsored placements are measured in impressions and engagement to effectively determine return on investment and revenue incrementality.

On-site inventory is managed and controlled by configuration in the PA Discovery OS Application. Once configured, Advertisers can target this inventory, originating Ad Sets in the Discovery OS app, or via a 3rd party ASM, DRP or Retail Media aggregator.

Getting Started

To begin using Retail Media, first configure Widgets and Routes for your website in Discovery OS. These elements are powered by Artificial Intelligence (AI) algorithms and are essential for supporting sponsored products and display ads. For detailed instructions on setting up these components, please refer to the Personalisation Overview documentation. Follow the step-by-step guides to configure widgets and routes to manage your media inventory effectively.

Once Widgets and Routes are configured, you can build Retail Media Placements in Discovery OS. Placements are your on-site inventory to display Ad Sets. Each Placement is configured at a slot level, meaning that you can configure multiple placements for the same widget, each with a different Retail Media strategy.

APIs for Integration

Integrating with PA’s On-Site Retail Media API is straightforward. The following APIs are essential for functionality:


Products, their attributes, and their relationships to each customer and session are crucial to the Particular Audience Relevancy AI algorithms in the PA Discovery OS Platform.

The Products API is used to push a product catalog to Particular Audience, which is essential for building Ad Sets for on-site sponsored product advertising. Product attributes are utilized by our Relevancy AI for Natural Language Processing (NLP) and Computer Vision (CV) machine learning to enhance product relationships.

Product attributes also aid in creating automated customer segments and affinity groups, such as targeting customers who purchased a specific brand in the last 30 days or viewed certain categories and brands within 90 days.


The Recommendations API calls for Sponsored Product slots based on the page being viewed, returning specific fields in the recommendations payload essential for subsequent events.

  • adSetId: Supplier Ad Set ID created from the retail media ad set.

  • adSetVersion: Tracks the current Ad Set version, used in analytics.

  • costPerClick: Slot CPC for calculating and validating the HMAC field.

  • hmac: An autogenerated and encrypted string shared between the recommendations API and event ingestion, ensuring click events are genuine and preventing fraudulent budget decrement.

  • hmacSalt: A string of 16 random characters generated by the recommendations API for HMAC validation.


Tracking Customer Behavior

The PA Events API captures customer behavioral signals, enabling highly relevant on-site sponsored product placements for each customer.

  • View Products Event: Tracks each product view, influencing retail media relevancy algorithms based on current, recent, or historical views.

  • Purchases Event: Gathers all purchase data, crucial for analyzing customer purchasing patterns and essential for retail media attribution and ROI analytics.

  • Add To Carts Event: Monitors when products are added to a cart, crucial for offering real-time relevant sponsored product ads as a customer's cart changes.

Analytics and Attribution Tracking

Properly tracking analytics and attribution is vital for measuring retail media impact and ROI.

  • Slot Impressions Event: Tracks all product recommendation slot impressions, important for understanding share of voice and key metrics like CTR.

  • Clicks Event: Monitors clicks on product recommendation slots and banners, essential for managing CPC budgets and providing engagement and attribution data.

  • Checkouts Event: Tracks customer checkout events, a critical step in the purchase process.