Retailers should consider owning their AI decision layer if they don’t want to lose ad revenue to LLMs.
Retail media is enjoying huge success at the moment, but this is threatened by the looming presence of the large language models (LLMs) like ChatGPT, Perplexity and Gemini and their ability to overpower data and profits.
As such, retailers have just a two-year window to protect themselves by ‘owning their AI decisioning layer’, according to a whitepaper from ParticularAudience.
Retailers that don’t take action to impose a decisioning layer are “effectively handing over their margins to big tech”. This is because they’d be leaving their entire catalogues exposed to third-party LLMs and AI scrapers allowing them to become “inventory in someone else’s auction”, as consumers increasingly turn to AI agents for their shopping needs.
Tech standards like the model context protocol (MCP) allow LLM agents to trawl through sites and browse products – but they are commercially blind, the research explained, and can’t enforce retailers’ inventory limits, profit margins, or sponsored ad modifiers.
James Taylor, chief executive and founder of Particular Audience, said: “For advertisers, AI offers improved efficiency and validated ROAS through predictive targeting and automated campaign management.
“For retailers, the prize is higher-margin yield and monetisation potential, while customers benefit from greater relevance and utility.”
To protect themselves and their profit margins, retailers must develop an AI decisioning layer within its architecture – built or bought – to ensure it is in control of its own monetisation, margin, and management.
“As LLM providers gain influence over how products are surfaced and ranked, the strategic question is: who controls the decisioning layer in commerce?” added Taylor.
“Retailers that own or meaningfully influence this layer retain commercial control of their monetisation surface. Those that cede it – by exposing catalogues directly to third-party LLMs without an intervening decisioning layer – risk becoming inventory in someone else’s auction.”
Agentic commerce is a fast growing industry. Global consulting firm Bain & Company has forecast the US market alone could be worth between $300-500bn by 2030, representing 15-25% of total e-commerce sales.
By the end of the decade, research from Morgan and Stanley suggests almost half of all online shoppers will use AI agents by 2030 adding adding $115bn into the US economy. It also found the speed of the adoption by consumers means retailers must act now to protect themselves.
