Retail media's AI gap: 88% use AI for analysis, but just 16% for product and ad ranking
While 88% of retail media professionals polled during Particular Audience's AI in Retail Media Town Hall said their businesses use AI to analyse data or write reports, just 16% use it to choose product or ad rankings. The findings reveal a gap between the industry's adoption of AI and its willingness to use the technology to make decisions that directly affect advertising performance.
The session, led by Particular Audience (PA) Founder and CEO James Taylor in Amsterdam, attracted 89 participants and generated 953 live contributions, with between 52 and 74 people responding to individual questions. PA was a headline sponsor of the summit.
Traditional targeting is still widely used. Among 68 respondents, 50% use manual keywords or categories and 54% manual audience segments. Only 22% use AI that predicts relevance from live shopping signals. Respondents could choose more than one option.
"AI is already doing plenty of work in retail media, but there is a vast gap in using it for the decisions that actually drive performance," said James Taylor, Founder and CEO at Particular Audience. "The bigger opportunity is in deciding which products and ads are most relevant to a shopper, how they should be ranked and when an offer makes commercial sense. Retailers need to own those decisions, set the commercial rules and let AI optimise within those boundaries."
Also Read: Madhive Selects AdImpact to Bring Competitive Spend Data To Local Media
Willingness to delegate depends on the task
83% would delegate campaign reporting now or within 12 months, and 65% product or ad ranking. That falls to 42% for bids and budgets and 41% for supplier-funded offers. Trust and control were the most cited barriers, alongside accuracy, transparency and proof of performance.
Auction pricing draws the most concern
Asked whether AI would make decisions safer or riskier, on a scale where 1 is "much riskier" and 5 "much safer", attendees scored AI product and ad ranking at 3.4 and AI deciding what advertisers pay at 2.8 (66 responses).
The commercial value of better targeting
Particular Audience's research demonstrates how moving beyond traditional targeting can improve product discovery. In a grocery search test involving over 700 real customer queries, its Adaptive Transformer Search technology identified 80% of the relevant product catalogue within the first 120 search results, compared with 48.5% using traditional keyword search, a 65% improvement.
Rather than relying on fixed keywords or audience segments alone, AI can use live signals, including what shoppers are searching for, browsing and adding to their baskets, to predict relevance as their intent changes. Combined with automated ranking and decisioning, this allows retailers to make better use of their inventory and advertising placements while keeping the shopping experience relevant.
Giving AI more control, with the right safeguards
Trust remains a major obstacle to wider AI adoption. While 83% of respondents would delegate campaign reporting now or within the next 12 months, just 42% would hand over bids and budgets and 41% supplier-funded offers. Trust and control were the most frequently cited barriers, alongside concerns about accuracy, transparency and proof of performance.
As AI takes on a greater role in retail media, the priority is to build confidence in the decisions it makes. That means clear commercial rules, transparent decision-making and measurement that proves incremental value. Retailers that establish these guardrails will be better placed to capture the benefits of automation while retaining control of product selection, ranking and monetisation.
Polling note: responses came from a self-selected audience at the 7 October 2026 session. Response numbers varied by question, and the findings are directional, not representative of the wider industry.
About Particular Audience:
Particular Audience is an AI-native retail media and personalisation platform that unifies proprietary large language powered search, hyper-personalisation, and sponsored product ad decisioning into a single ranking system.
Founded in 2017, PA helps retailers maximise profit-per-pixel-balancing relevance, conversion, margin, and ad yield in real time. Its multimodal machine-learning platform replaces legacy keyword and rules-based approaches, enabling automated, intent-aware retail media without degrading the shopper experience.
Headquartered in London, Sydney, and Vancouver, Particular Audience supports global enterprise retailers and contributes to open standards such as the Model Context Protocol (MCP), enabling agentic commerce on retailer-controlled infrastructure.
