AI in eCommerce: Evolution or Extinction?
If you spend five minutes on LinkedIn, you’ll be told one of two things: AI will transform eCommerce beyond recognition, or it’s overhyped and nothing will really change.
The reality for UK retailers sits somewhere between those extremes.
AI isn’t new to eCommerce. What’s new is the visibility of it.
Long before ChatGPT went mainstream, tools like Nosto were personalising product recommendations, and platforms such as Algolia were refining on-site search with intelligent algorithms. We just didn’t call it “AI” every time.
Today, the label has changed. The capability has accelerated. But the fundamentals of retail — trust, margin, brand, experience — are still in place.
The debate isn’t whether AI has value. It’s about timeline and adoption.
Customer Service: Where AI Is Already Paying Off
If there’s one area where AI is genuinely delivering commercial value today, it’s customer service.
Retailers are using AI to handle high volumes of repetitive enquiries, surface order history instantly and provide 24/7 responses without scaling headcount at the same rate as order growth. According to IBM, organisations using AI in customer service report measurable cost efficiencies and performance improvements, including reduced support costs and improved response times¹.
For a growing eCommerce business, that matters.
But there’s a trap here.
Many brands rushed into chatbots years ago and created frustrating loops with canned responses, poor context, no clear route to a human. Large language models are significantly more capable, but they still require careful implementation. If your returns process is unclear or your data is messy, AI won’t magically fix it. It will simply automate confusion faster.
The retailers seeing success are those who treat AI as a layer on top of a well-designed journey, not a shortcut around it.
The Generational Shift You Can’t Ignore
Consumer behaviour is shifting too.
Younger shoppers increasingly prefer self-service. They would rather resolve an issue through chat, search or automation than pick up the phone. For them, speaking to a human is often the last resort.
That doesn’t mean human service disappears. It means it becomes premium.
We may see a future where:
-
AI handles the default tier of customer support
-
Human interaction becomes the enhanced tier
Just as vinyl didn’t disappear when streaming arrived, human interaction won’t vanish, but its role may evolve.
The Rise of AI Agents — and the Fear of “Browserless” Commerce
Where things get more disruptive is on the customer side.
We’re now seeing partnerships between OpenAI and Shopify, AI-native browsers like Comet from Perplexity AI, and payment providers such as Stripe exploring AI integrations.
The narrative is bold: customers won’t browse websites. They’ll ask an agent to find the best product, compare options and complete checkout.
For low-consideration purchases, that makes sense. If you’re buying batteries or a plug converter, speed matters more than brand storytelling.
But retail isn’t built solely on commodity products.
When customers buy:
-
High-ticket items
-
Configurable products
-
Luxury goods
-
Or anything connected to health or lifestyle
They don’t just want the cheapest option. They want reassurance. They want to feel understood. They want to trust the brand.
Websites do more than process transactions. They signal credibility, authority and identity. Removing that layer entirely assumes customers are comfortable delegating those judgements to AI. That level of trust isn’t universal yet.
Adoption Is the Real Variable
The loudest voices in tech often assume that once a tool exists, adoption is inevitable.
Retail history suggests otherwise.
Many brands still choose not to sell on Amazon because they want to protect margin, brand positioning or customer ownership. Those same considerations will apply to AI agent integrations.
There are also behavioural barriers. Even small UX inconsistencies can reduce trust. If customers hesitate over something as minor as payment logo presentation, full delegation to an AI agent won’t happen overnight.
The only thing genuinely up for debate here is timeline.
AI capability is accelerating quickly. Consumer trust and behavioural change move more slowly.
SEO vs AI Search: Less Disruption Than You Think
There’s growing discussion around “GEO” (Generative Engine Optimisation) replacing traditional SEO.
In practice, the foundations remain the same.
AI search is becoming an integral part of everyday shopping behaviour, with many customers saying they would use it to search for personalised product recommendations (49%)². If your technical SEO, authority and structured data are solid, you’re already well positioned.
What does change is the importance of data hygiene.
AI models pulling from product feeds and APIs mean hidden inconsistencies become visible. Inaccurate metadata, weak specifications or poor structured data will limit visibility in both traditional and AI-driven search.
The advice isn’t glamorous, but it’s practical:
-
Clean up your product data
-
Ensure structured data is complete and accurate
-
Maintain authority-building content
-
Continue investing in technical SEO
The retailers who treat this as a data discipline rather than a marketing trend will benefit most.
So What Should UK Retailers Do Now?
You don’t need a radical AI strategy overhaul tomorrow. But you do need awareness.
First, use the tools yourself. If you don’t understand how AI search or agents behave, you can’t optimise for them.
Second, modernise obvious gaps. Legacy rule-based chatbots and outdated search experiences will become more noticeable as competitors upgrade.
Third, protect what differentiates you. Brand, trust and experience still matter — particularly for mid to high-consideration purchases.
Finally, resist the extremes. AI is neither irrelevant nor the immediate death of the website.
What’s far more likely is gradual integration.
Platforms will embed AI features by default. Retailers will upgrade tools because they’re better, not because they’re labelled “AI”. Customers will use agents for convenience purchases but still browse for emotional or complex ones.
Evolution, not extinction.
And the retailers who win won’t be the ones chasing every headline. They’ll be the ones improving experience, cleaning data and adapting steadily as adoption unfolds.
Sources
- https://www.ibm.com/thought-leadership/institute-business-value/en-us/report/virtual-agent-technology
-
https://business.adobe.com/resources/digital-trends-report.html
