By Gary Carlile, EVP at nShift
AI is rapidly becoming the new front door to commerce. In the UK, the shift is already visible. Search is becoming more conversational; AI-generated overviews are reshaping how consumers evaluate options, and tools like ChatGPT are becoming part of the path to purchase.
According to Ofcom, Google Search is used by 82% of UK adults, around 30% of searches now show AI overviews, and ChatGPT recorded 1.8 billion UK visits in the first eight months of 2025, up sharply from 368 million in the same period a year earlier.
This matters because ecommerce is highly important for the UK economy. According to the Office for National Statistics, online sales accounted for 28.3% of total retail spend in December 2025, with online sales values up 11.1% year over year.
So, the question for retailers now isn’t whether AI will influence shopping. It already is. The harder question is whether the delivery operation supporting the customer experience is strong enough to support the expectations AI is helping create.
As Jurgen Leijdekker, CEO of nShift, explains, “AI will reshape ecommerce, but it will not remove the operational realities underneath it. If anything, it will make them more visible.
The retailers that benefit most from AI will be the ones with the strongest delivery foundations, connected data, and platforms that can turn intelligence into execution fast”.
AI is changing commerce from the start of search and down
The shift is much more than improved search. McKinsey’s 2025 work on agentic commerce argues that AI agents are beginning to move from answering questions to shaping and executing transactions, estimating that agentic commerce could mediate as much as $1 trillion in US B2C retail revenue by 2030, with global potential reaching $3 trillion to $5 trillion.
For retailers, this is both exciting and uncomfortable. Exciting because AI can compress discovery timelines, reduce browsing friction, and help consumers make decisions faster.
Uncomfortable because smarter discovery does not remove the operational burden that follows the click. It raises the stakes. Why? Commerce is not won when a customer finds a product.
It is won when the retailer keeps the promise. And that promise depends on delivery options, carrier performance, tracking visibility, exception handling, and returns.
AI can improve the promise, but it cannot fulfil it
This is where the next wave of AI adoption is likely to separate the strong operators from weaker ones.
AI can make the top of the funnel search feel cleaner and more intelligent. It can improve recommendations, merchandising, service interactions, and even parts of checkout.
But it cannot make a fragmented delivery operation reliable. It cannot connect disconnected carrier data by itself. It cannot make poor returns processes feel seamless.
If the weaknesses already exist, AI won’t eliminate them. It will make them more obvious. This is a particularly important point for UK ecommerce leaders.
As the buying journey gets shorter and smarter, customer tolerance for friction does not expand with it. It contracts. So while AI may improve how shoppers arrive at a purchase decision, it will not change the basic physics of delivery.
If the promise at checkout is disconnected from reality in execution, the gap becomes visible faster. If post-purchase communication is poor, if the experience breaks in places that customers will remember, if returns are clumsy, the intelligence at the front end starts to look cosmetic.
The real AI story is underneath the interface
The broader evidence on AI adoption points in the same direction. McKinsey’s 2025 State of AI research found that workflow redesign has the biggest effect on whether organisations see EBIT impact from generative AI.
More than three-quarters of respondents said their organisations use AI in at least one business function, yet more than 80% said they are yet to produce a tangible enterprise-level EBIT impact. This gap matters because it suggests many companies are still adding AI without rewiring the business systems that enable real value creation.
BCG’s 2025 research is even more direct. Only 5% of firms in its study are generating significant value from AI at scale, while 60% are seeing little to no material value.
The difference is not enthusiasm alone. It is whether AI is being embedded into operating models, data structures, governance, and workflows. Retail is especially vulnerable to getting this wrong because the customer-facing layer is easy to showcase.
A slick assistant, faster search, and personalised recommendations look impressive in a demo. But the necessary operational work underneath remains the difficult part.
Connecting delivery promises to carrier capability is hard. Making shipment data usable across systems is hard. Handling post-purchase exceptions at scale is hard. Returns orchestration is hard. None of this becomes less important because the interface gets smarter.
What this means for UK retailers now
The practical takeaway is not to slow down on AI. It is to get much more serious about what AI depends on.
If discovery is becoming more intelligent, checkout would benefit from becoming more accurate. If shopping journeys become more dynamic, fulfillment can become more dependable. If more of the experience is automated, then tracking, exceptions, and returns should reflect what is actually happening, not what the brand hopes is happening.
This is why this is not just an AI story. It is a platform story.
The retailers most likely to benefit from AI in the coming years are unlikely to be the ones with the most visible AI features. They are likely to be the ones with operational infrastructure strong enough to support these features. They will have the systems that connect the customer promise to delivery execution, translate data into usable decisions, and keep the post-purchase journey from becoming the weak link.
In this sense, AI is not lowering the bar for ecommerce. It is raising it.
The next phase of ecommerce will be less forgiving
AI will make commerce smarter. It will make journeys more adaptive, discovery more efficient, and customer interactions more responsive.
But it will not rescue weak delivery operations. It will reveal them faster, because the distance between promise and performance will become easier for customers to see and harder for retailers to explain away.
Strategically, retailers can look at not where AI can be added, but whether the commerce and delivery foundations underneath the experience is strong enough to carry it.
In the next phase of ecommerce, AI is transforming how people shop — but execution can determine who wins. AI gets attention. Execution earns loyalty.
Published 26/03/26
