Retail’s Hidden AI Blind Spot

Over the past year, artificial intelligence's role in the shopping journey has moved past the experimentation phase, becoming a baseline expectation for consumers. Tools such as AI agents, AI-powered search and recommendations, and AI-powered checkout have become an integral part of consumer experiences in a way that goes beyond just optimizing experiences.

AI tools are also becoming the gateway to product discovery, which can be transformational if the technology works as intended. However, results can often come up short, with MIT reporting 95 percent of generative AI pilot programs failing due to poor and fragmented data foundations, ultimately leaving consumers with poor recommendations. To prevent this and ensure shoppers are getting real value from AI, brands and retailers must ensure they have accurate, consistent and complete product information.

The Trust Gap: Why AI Fails Without Accurate Product Information

As AI in commerce has continued its exponential growth, one primary element that often slips through the cracks is a focus on strong product data. To deliver the personalized recommendations consumers expect, AI works with the data it's provided, and can’t interpret any additional context or intent beyond structured data. If current product information has inconsistencies in the types of data or missing insights, AI is bound to hallucinate to fill in the gaps when it isn’t built to do so in a reliable manner.

Weak product information doesn’t just risk poor recommendations in one-off scenarios; it can turn into a customer loyalty issue if not addressed. Recent data shows that unclear product records cause 65 percent of consumers to switch brands. Furthermore, 33 percent claim inaccurate product information diminishes brand loyalty, showing just how significant the risks of not having updated data are.

Setting Up for Success in the AI Shopping Era

To ensure brands and retailers are set up for success in implementing AI shopping experiences, they need to make sure their product information maintains a few key elements, including:

• Having a single source of product truth: AI relies on consistency to deliver strong outputs. Centralizing and governing product information in one location allows data to remain consistent across all channels, eliminating conflicting answers that AI tools may use if data sprawl exists.
• Maintaining structured and enriched product attributes: To ensure there’s no room for AI to hallucinate its own interpretations, brands and retailers’ data must be comprehensive. Whether it’s product dimensions, materials or regulatory information, the more insights a product’s data can include allows AI to understand what its intended use is.
• Continuous management for data quality: The current state of a product will most likely not be its end state, as user reviews and feedback may dictate how a product evolves. Product data must do the same alongside it. Incorporating feedback loops gives brands and retailers one less thing to worry about updating. Consistent feedback ensures AI always has the latest data to give consumers the best recommendations.

In addition to preparing product information for the AI era, brands and retailers should always be sure to keep transparency at the forefront of shopping experiences to drive success. With only 45 percent of consumers trusting that AI can deliver reliable results, maintaining a transparent business model is crucial to win trust. If brands and retailers are going to partake in AI-driven shopping, they must stand behind AI’s recommendations with full confidence that it will provide a better customer experience.

At the end of the day, the most impactful AI strategy isn’t adopting new tools, it’s centralizing the foundation to fit AI’s needs. The retailers that act on updating their product information now will be among the first to see better AI performance, higher conversion rates, and increased loyalty. Getting a head start is the best thing that can be done to come out on top in the AI shopping era, and that means focusing on product data.

Romain Fouache is the CEO of Akeneo, a product experience (PX) company and global leader in product information management (PIM).