From search to discovery
Justin Thomas, VP Sales, EMEA North for Akeneo explains why product data is now the front line of AI commerce.
Optimise the page, improve the ranking, buy the sponsored slot, refine the funnel; in digital commerce, this has been the operating model for the longest time. That’s no longer enough because AI has changed how people search, specifically how products are interpreted, compared, recommended and, increasingly, chosen.
Shoppers’ discovery is becoming more intent-led, conversational and mediated by intelligent systems. A customer may no longer ask for a “waterproof trail running shoe;” they may ask an AI assistant what to wear for a wet 10k run across muddy ground, with ankle support, within a certain budget, from a retailer that can deliver tomorrow. The winning product will be the one whose data gives the agent enough confidence to recommend it.
That is the new reality for brands, manufacturers, distributors and retailers. In an agentic commerce environment, product data is the product experience. This does not mean every purchase will be handed over to AI overnight. Consumers are comfortable using AI to research, compare and focus on options, but trust still matters at the point of purchase.
Shopping and buying are different behaviours. One is exploratory, the other involves confidence, brand preference, price, availability, fulfilment, returns and control. But this distinction makes product data more important, not less. AI may influence the journey before the customer reaches a product page, before they see a campaign, and sometimes before they know which brand they prefer.
That creates a profound discoverability challenge. AI agents and conversational interfaces do not evaluate products in the same way a traditional search engine ranks pages. They parse, summarise and reason across structured information, reviews, specifications, attributes, availability, content quality and contextual relevance. If the data is thin, inconsistent, incomplete or trapped in disconnected systems, the product may rank lower or may not be surfaced at all.
This is why the old separation between product information and commercial performance is breaking down. Now, Product Information Management has become the foundation for visibility, trust, conversion and growth.
There are three imperatives for success. The first is velocity. Markets move quickly, trends emerge quickly, and AI-led discovery can change quickly. Brands cannot afford product data processes that take months to adapt, they need the ability to enrich, localise, govern and activate product records at the speed of market intent.
The second is comprehensiveness. AI discovery rewards depth. Basic descriptions and technical specifications are no longer enough. Product records need to answer a much wider range of buyer intents; use cases, compatibility, sustainability requirements, materials, care instructions, comparisons, constraints and reasons to believe. In other words, brands must move from thin product pages to rich, agent-ready product stories.
The third is trust. AI can accelerate enrichment, translation, classification and optimisation, but unmanaged AI also creates risk. Content that is inaccurate, off-brand, non-compliant or generic can damage customer confidence very quickly. The answer is not to avoid AI. It is to govern it. Businesses need AI that works within clear data models, validation rules, approval workflows and brand guardrails.
An agentic-first approach is essential with all data in a product cloud designed to govern and orchestrate product data from supplier to customer, and combining an adaptive data model with specialised AI agents that support the product data lifecycle. These agents can help teams centralise, enrich, activate and optimise product information, while keeping humans in control of strategy, governance and quality.
The role of the team changes as a result. Instead of spending the majority of their time correcting spreadsheets, reformatting attributes and repairing channel errors, people can become strategic orchestrators. They define the rules, supervise the agents, identify opportunities and focus on the product experiences that drive growth.
But product experience is no longer only about what a product is. It is also about how it performs commercially. That is why pricing is becoming part of the PIM conversation. Product data explains what is being sold, why it matters and where it should appear, while pricing helps determine how that product competes, converts and contributes to margin. In AI-mediated commerce, those signals cannot sit in isolation.
An agent recommending a product will increasingly need to understand not only features and fit, but value. Is the price competitive? Is the offer relevant? Does the product meet the customer’s intent at the right balance of quality, availability and cost? For brands and retailers, this means pricing is very much a part of the experience itself.
The next phase of commerce will see businesses build on a foundation of product data that is trusted, structured, complete and commercially intelligent, data that can be understood by humans as well as machines.
