by Trustpilot senior UK marketing director Steph Gillies

Between customers finding products and services via artificial intelligence tools, brands creating AI-focused experiences, and AI being mentioned far more often in customer reviews than previously, it’s clear that AI is rapidly reshaping retail.

However, research by the Centre for Economics and Business Research (Cebr), commissioned by Trustpilot, reveals that clunky or frustrating AI experiences put roughly £9bn of UK online retail sales at risk, representing about 6 per cent of the entire e-commerce market.

At a time when shoppers have an increasingly low tolerance for poor experiences, off the back of having more choice than ever and ongoing cost-of-living pressures, business leaders must realise that a bad AI interaction is no longer just an inconvenience, it is an expensive bottom-line mistake.

Trust begins at discovery

Retailers need to look beyond the basket and tills, as AI search is increasingly shaping customer confidence in all aspects – including customer experience. This starts at the discovery phase, as when consumers use tools like ChatGPT or Gemini to research products, trust and review websites frequently serve as the second most common citation source.

The impact of reviews  grows as consumers move through the buying journey, with the number of citations quadrupling from a casual discovery prompt to a high-intent buying prompt.

Because algorithms favour independent data and human experiences to formulate answers, ratings and reviews now appear much earlier in the buying journey than they typically did before.

This isn’t a coincidence, as rather than customer feedback simply appearing more often in AI responses, these trust signals are something that consumers are actively looking out for. For example, Trustpilot found that 57 per cent of people state they trust AI-generated recommendations more when they explicitly include reviews from Trustpilot.

AI should simplify purchasing journeys, not complicate them

The ultimate purpose of retail AI is to remove friction and enhance the retail experience as consumers move from stage to stage of their purchasing journey.

When executed well, shoppers love advanced features, with automated checkouts and virtual shopping assistants boasting satisfaction rates over 90 per cent. AI-generated summaries and AI product recommendations are also highly popular, used by 62 per cent and 57 per cent of consumers respectively, with fewer than 16 per cent reporting any issues.

The trouble starts when AI is poorly implemented, as 40 per cent of shoppers report negative experiences overall. Too often, brands rush to launch automated tools just to cut operational costs, making shopping a chore rather than a convenience. Evidently, using AI to speed up purchasing decisions means nothing, if usability or access to human support is compromised.

Take for instance customer service chatbots, which are frequently flagged as the weakest link in the digital journey. While nearly 80 per cent of UK shoppers have used one, the real issue can occur when businesses use automated systems to reduce pressure on human teams. When deliveries go missing or refunds are delayed, frustration arises as consumers seek real empathy, troubleshooting and nuance a human can provide.

These consumer frustrations and disconnect carry a real risk to retailers’ bottom lines; of the £9bn at risk, £4.9bn stems from immediate lost transactions when annoyed shoppers abandon their carts. The remaining £3.7bn reflects long-term reputational damage. On average, a dissatisfied shopper shares their bad AI experience with two other people, and 65 per cent of consumers will actively reduce their spending with a business after a single bad automated encounter.

Transparency and the human voice

As AI guides consumer buying decisions and determines retail success, transparency is no longer optional for retailers. Shoppers get highly uncomfortable when algorithms operate like a mysterious black box.

They need to understand how recommendations are made and what personal information is used. Openness turns an intrusive algorithm into a helpful, trusted personal shopper.

In a world of increasing automation, human feedback loops remain central to maintaining trust.  Trust is the currency that powers every transaction in an AI world, making it vital for businesses to get these experiences right – particularly with consumers making their voices heard via ratings and reviews.

With this in mind, I’d encourage brands to focus on inviting and responding to feedback, having this openly available, and most importantly – using customer insights to improve their offerings. The path forward for retailers is clear when it comes to the AI transition, it requires balancing smart technology with a human touch. To protect revenues, retailers must ensure AI genuinely improves usability, keep a clear path to live support, stay transparent about data, and anchor systems in real customer feedback.

Click here to sign up to Retail Gazette‘s free daily email newsletter