AI is the new storefront and marketplaces are the new high street, says Romain Fouache, CEO of Akeneo. In a disintermediated commerce landscape, whoever controls product truth increasingly controls growth.
For much of the digital commerce era, product information sat quietly in operational domains, important, but rarely strategic. It powered PDPs, marketplace feeds, catalogues and campaigns, but it wasn’t central to competitive advantage.
That changes in 2026. As AI becomes the interface for shopping and commerce disintermediates beyond brand-owned channels, product data is becoming a control point for visibility, trust and revenue.
AI is rewriting the shopping journey and the 2025 holiday season made the shift visible. According to Adobe, Black Friday broke new digital records with $11.8B in US revenue, but more revealing was how people bought. Nearly a third of consumers used ChatGPT to complete a purchase and said they would do so again, citing both speed and reduced friction. Consumers asked AI to surface options, compare products, validate compatibility and explain trade-offs, functions previously spread across reviews, filters, FAQs, influencers and search.
AI is compressing discovery, comparison and checkout into a single conversational flow. The feedback loop was overwhelmingly positive: 84% of shoppers reported good experiences with AI-driven purchasing and more than a third completed at least one purchase based on AI recommendations.
And consumer expectations have shifted. They are saying, “If I describe what I want, the system should find it, prove it and help me buy it.”
However, AI fails fast when product data is weak. It does not inherently understand products. It interprets structured attributes, taxonomies, compatibility data, materials, dimensions and use cases to infer meaning and relevance. When product information is incomplete or inconsistent, AI fills gaps through probabilistic reasoning, a polite way of saying it guesses. The result in ecommerce is that incomplete data increases bounce rate while in AI commerce, incomplete data erodes trust.
Weak product information has nowhere to hide. It becomes the only information the model sees and the only information the shopper receives. 65% of consumers have already switched brands due to unclear or insufficient product information and one-third say inaccurate information directly damages loyalty. In an AI context, those consequences scale faster.
AI is powerful, but it is not magic. It cannot infer compatibility unless compatibility exists in the model. It cannot warn of allergens unless allergen data is structured. It cannot validate materials, sustainability attributes, returns constraints or substitutions unless the dataset contains them.
The second major shift is channel disintermediation. Brand websites are no longer the gravitational centre of digital commerce. More than half of consumers now plan to shop via third-party apps, marketplaces, social commerce environments, or AI-powered discovery layers. Retailers and brands are seeing storefront traffic decline not because demand is shrinking, but because demand is moving upstream into environments they don’t own.
In a disintermediated ecosystem, product information must be synchronised across every surface where buyers encounter it, whether the final conversion happens or not. The strategic implication is profound. When discovery and decision-making no longer happen on brand-owned property, product data becomes the only portable artifact of truth. In other words, success does not depend on having a storefront but on owning the product truth.
This is why product information is emerging as a board-level conversation. The questions leaders are asking have shifted from IT-centric to commercial. Are we syndicating the same truth across channels? Can AI agents understand and represent our products accurately? How many launches are delayed due to product data readiness? How much revenue leakage is tied to misinformation, returns, or friction? How fast can we extend into marketplaces, regions and new categories?
These are growth questions, not infrastructure questions. They sit at the intersection of revenue, trust and customer experience, precisely the domains boards are mandated to govern.
In 2025, even MarTech analysts marked a pivot in how PIM was evaluated: not just on functionality but on business outcomes like conversion, launch velocity, omnichannel consistency and AI readiness. Execution and usability increasingly differentiate leaders from laggards, because poor product data is a commercial constraint.
The role of product information itself is evolving. Historically, PIM was designed to manage and enrich data for accurate publication. In 2026 and beyond, the mandate expands to include syndication across a proliferating channel ecosystem; activation in AI-powered discovery and recommendation engines; enablement for sales, merchandising and partnerships; feedback loops integrating returns, questions and reviews; and governance ensuring consistency and trust at scale
Product information is therefore becoming a live system, continuously enriched, continuously activated and continuously synchronised.
The growth formula for 2026 recognises that three forces are converging: AI increases dependence on structured product data; disintermediation requires product truth to travel beyond owned channels; and trust becomes a commercial differentiator, not a UX nicety. The result is the following equation – AI + disintermediation + trust = product data as growth lever.
Most companies will invest heavily in AI tooling over the next 24 months. The differentiator will be those who invest in the foundation required to make AI commercially viable and commercially defensible – managed, accurate, consistent and governed product data.
AI is the engine while product data is the fuel, and boards will fund both.
