AI has already moved beyond powering product pages and now determines visibility, pricing, recommendations and ultimately revenue, making it a board-level commercial asset, explains Benoit Jacquemont, CTO at Product Experience (PX) Company Akeneo.
Product data now determines whether products are discovered by AI, how they are priced, whether they appear in recommendations, how effectively they convert and increasingly, whether they are purchased at all. As commerce becomes more intelligent, autonomous and conversational, product data is therefore evolving from operations into the infrastructure that underpins revenue generation.
The organisations that embrace this shift will be better positioned for the next era of commerce while those that continue to view product information as an administrative overhead risk becoming invisible.
Much of the conversation around artificial intelligence focuses on models, algorithms and automation, yet AI is only ever as effective as the information it receives. This is becoming increasingly apparent as AI moves from experimentation into core business operations. AI has moved well beyond experimentation. McKinsey’s latest global research shows that almost nine in ten organisations now use AI in at least one business function, while most are deploying it across multiple functions. Businesses are also reporting measurable commercial returns, particularly in marketing, sales, software engineering and customer service, although enterprise-wide transformation still depends on the quality of underlying data and governance.
Companies are getting these returns because some have invested in the data foundations that allow AI to make better decisions. The problem is most product information was never designed for this world. In many organisations, product data still sits fragmented across ERP systems, spreadsheets, supplier databases and disconnected applications. Information is duplicated, incomplete or inconsistent. And while humans can often compensate for those gaps, AI cannot, which means that as AI increasingly becomes the interface between buyers and sellers, poor product data becomes a direct commercial liability.
Visibility therefore depends on structured product intelligence. Traditional digital commerce rewarded brands that invested heavily in websites, search engine optimisation (SEO) and user experience. Agentic commerce doesn’t work that way. Increasingly, consumers are asking AI assistants for recommendations rather than browsing category pages themselves. Large language models (LLMs) are beginning to compare products, evaluate suitability, explain differences and ultimately complete purchases on behalf of users.
In this scenario, AI never experiences the company website but it does experience product data from all known and trusted sources. Every attribute, specification, image, taxonomy, compatibility field and sustainability credential becomes part of the information AI uses to determine whether a product deserves recommendation. If product information is incomplete, inconsistent or poorly structured, AI cannot confidently recommend it. The kicker is that this can mean better products may lose to better data.
Looking ahead, optimised product data will enable a whole range of improvements in pricing, merchandising, personalisation, cross-selling, localisation and customer support, which in turn cut return rates and improve inventory levels as well as conversion performance.
Going further, existing product information platforms will be able to make these changes continuously, which means issues are identified before they become problems, and enrichment activities are coordinated quickly, both of which enable teams to operate at AI speed. So, rather than expecting teams to manually identify missing attributes, resolve syndication errors or enrich thousands of products individually, intelligent and managed agents can coordinate these activities continuously, surfacing recommendations, automating repetitive work and allowing employees to focus on higher-value decisions.
The objective is not to replace people but to enable them to spend less time managing product data and more time improving commercial outcomes. However, one thing to bear in mind, as AI becomes embedded across commerce, governance becomes a competitive advantage rather than a compliance exercise. The faster AI operates, the greater the need for trusted product data, clear approval workflows and human oversight. Akeneo’s UK research found that governance and compliance are among the biggest barriers to scaling AI, even as organisations grow more confident in their AI readiness.
Commerce has now entered an era where every recommendation, every search result and every autonomous purchasing decision begins with trusted product information. And the technology will get better and better but the key to success is for organisations to treat product data as commercial infrastructure rather than operational administration.
