Particular Audience Particular Audience
Talk to us

The retail decisioning layer

One system decides what each shopper sees. Most retailers run four or five separate tools.

Search in one tool, recommendations in another, pricing in a third, ads bolted on top — nobody owns the order products appear in, and the margin leaks between them. Particular Audience replaces that patchwork with a single decision: a Unified Decision Engine that scores relevance, inventory, margin and sponsored demand in one pass, on every surface, including the AI assistants now doing the shopping.

Intersport Austria — Intersport launched onsite retail media — and proved brand g | Particular Audience

Sponsored Products and Sponsored Display generated 536,653 impressions, 5.81× SKU ROAS and 17.3× brand-halo ROAS in the first five weeks live.

Intersport launched onsite retail media — and proved brand growth in five weeks

Sponsored Products and Sponsored Display generated 536,653 impressions, 5.81× SKU ROAS and 17.3× brand-halo ROAS in the first five weeks live.

  • 17.3× Brand-halo ROAS
  • 5.81× SKU ROAS
  • 536,653 Impressions
  • 1.30% CTR

First five weeks live · 13 Jul – 16 Aug 2026 · McKINLEY & Active Rebel · from IAB Slovenija Final deck. Bayesian share-of-category used sequential periods, not a randomised holdout; the next phase is proper holdouts.

Intersport Austria moved from build-or-buy decisions to a live onsite retail media network on intersport.at, with Sponsored Products ranked into listing and product pages and Sponsored Display delivered through feature teasers and image-text banners.

The first five weeks ran with two owned brands: McKINLEY in Outdoor and Active Rebel in Fitness & Running. The campaign delivered 536,653 impressions, 6,988 clicks, 1.30% CTR, 17.2% cost of sale, 5.81× SKU-level ROAS and 17.3× brand-halo ROAS.

Share of voice rose from 0.58% to 7.37% for Active Rebel and from 6.72% to 22.60% for McKINLEY. Together, the two brands reached 30% of site impressions during the campaign.

The Bayesian share-of-category analysis compared sequential periods rather than a randomised holdout. It reported P(test > control) of 100% for both brands; the next phase is proper holdouts.