From Segments to Signals in Retail Media — recorded live on the Retail Media Stage at MAD//North, Manchester, 25–26 February 2026. AI, machine learning and the closing gap between retail personalisation and media targeting.

Speakers: James Taylor (CEO & Founder, Particular Audience); Beth Smith (VP Customer Success, Particular Audience); Dave Trolle (Marketing Director, Face the Future).

Personalisation is just good prediction. The same collaborative filtering that predicts your next song predicts the item a shopper needs right now — from clicks, dwell time, basket adds and semantic search, not demographic segments.

75% of intent is missed by exact-match keywords. The system learns what the user is trying to achieve right now, not who they were three months ago.

Face the Future case study: when ads were sold on a bid-only auction, ad scale rose 9× while retail media CTR fell 87.3% (overall CTR −82.5% desktop, −59.9% mobile). Switching off signal-led relevance cut clicks 44.72%, sales 55.62%, sitewide AOV 12.68%, units per transaction 10.52% and revenue per visitor 12.27%.

With/without comparison presented on stage by Face the Future and Particular Audience; directional.

Dave Trolle, Face the Future: "From a brand side, it's hard to let go of control. Advertisers like the 'comfort' of buying a specific keyword or audience segment because it's tangible and what we've known over recent years."

Beth Smith, Particular Audience: "Changing the mindset requires proving that 'black box' AI isn't actually a black box — it's just a faster, smarter way of processing thousands of data points at once."

Watch the session · Face the Future case study · Retail Media Cam podcast