AI traffic to UK retail sites has surged, but many small stores are invisible to it. We share how to get your site seen by AI shopping tools.
- Chatbot visitors are still a small share of sessions, but they’re growing fast; traffic to UK retail sites from AI tools has grown 1,439% in just two years.
- Chatbots can only recommend what they can read. Put specs, delivery times, and returns policies in plain text, and check you’re not accidentally blocking AI bots.
- Trust matters as much as the product page. Keep your details consistent across your page, feed and structured data, and build genuine reviews.
Your next customer might ask a chatbot what to buy instead of searching on Google.
Due to the steady rise of agentic commerce, traffic to UK retail sites from AI tools such as ChatGPT and Gemini has grown 1,439% in two years, and 37% of UK consumers have already used an AI assistant for online shopping.
However, there’s a catch for small sellers: chatbots can only recommend what they can read, and the average retail product page scores just 66 out of 100 for AI readability.
So, how do you make the AI shortlist? And what should you fix first? We spoke to ecommerce experts and small online sellers about what to change on your website, from product details AI can’t read to settings that may be blocking it entirely.
Why AI is already worth a place in your sales strategy
AI shopping may still be a small channel, but for some sellers it’s already showing up in the numbers.
Deesha Goswami, the founder of handcrafted jewellery and clothing business House of Falguni, saw ChatGPT send 334 visits to her site in 30 days, just two more than Google’s 332. That window produced only one order, but it shows how quickly chatbots can rival traditional search as a source of customers.
For others, the visitors are few but valuable. Christophe Renaud of digital strategy agency Arkeris says ChatGPT and Perplexity account for just 1-3% of sessions across the stores he audits, but those visitors convert around twice as well as search traffic.
You may not even notice AI sales when they happen. Founder of AI-native holding company Avenor, Brian Bigelow, says a customer found his brand through ChatGPT and bought a $1,400 sofa in one evening, but the store logged the sale as direct traffic. Without tagging your links, he tells us, you could be benefiting from AI and never know.
[Agentic commerce] is early, but very real. At Covelle we make non-toxic, made-to-order sofas, and we’ve already had a new customer find us through ChatGPT, comparing our collections and buying a $1400 loveseat in a single evening. The customer found us directly through ChatGPT, not through an ad or earlier visits.
While sales made from start to finish by automated AI shopping bots still represent a small fraction of UK transactions, that could change quickly. Research from Global Payments suggests that if shoppers embrace the technology, AI agents could account for around £99 billion of UK retail spending a year by 2031.
None of this means moving your marketing budget just yet. But with AI visitors already turning into customers, it’s definitely a trend worth tracking now.
Make your products readable: what experts say to fix first
Chatbots only recommend what they can read. So, your first job is making sure they can read yours. The good news is that most fixes are simple, and many are free.
Jonathan Solis, a solo Shopify seller and founder of online pet care business Whisker Bark, recommends starting with your product pages. “Treat your product page as the thing an AI agent reads, not a human skimming,” he told us.
“That means a plain title that says what the product is and who it’s for, a description that answers the questions a buyer would ask a salesperson (fits what size dog, what weight, what cars, how it installs), and complete structured data: price, availability, dimensions, weight, shipping,“ Solis adds.
It also pays to be specific about delivery, too. Yavuz Saka of ecommerce fulfillment company DSCP says AI tools respond better to details like “2 to 4 days to mainland UK” than to vague promises of fast shipping.
Another problem starts at the door. Rob Smith, founder and CEO of Lionfish Tech Advisors and a former Gartner analyst, points to a test of 200 retailers in which 29 blocked the AI shopping agent entirely. Of the 171 it could read, 78% had no findable delivery details, and 80% had no findable returns policy.
According to Smith, the same gaps turn up on nearly every small store he tested. “A machine only trusts what it can read on the page. Nothing else counts,” he says.
Your product page isn’t the whole story. Reputation counts too
According to the experts I spoke to, fixing your own site is only half of the battle. AI tools also draw on what the rest of the web says about you, and they’re quick to spot when the details don’t match.
Christophe Renaud, founder of the web agency Arkeris, tells us that genuine reviews on Trustpilot or Google are one of three fundamentals, alongside an accurate Google Merchant Center feed and clean schema markup. “An AI doesn’t take your word for it. It will often go and check what your buyers are saying about you behind your back,” he says.
That track record could also get harder to catch up on. Joshua Uebergang, founder of Shopify marketing agency Digital Darts, says early Shopify App Store reviews can build lasting advantages that newer entrants struggle to displace.
“If AI channels use evidence of real sales, customer reviews and independent mentions to recommend products, brands building that track record now could gain an advantage that cleaner product data alone may not overcome.”
AI won’t replace your marketing overnight, but the stores that get readable, open, and well-reviewed now will be the ones it recommends first. The good news is that the fixes are mostly simple and free. Here are five to start on today.
- Rewrite your product pages in plain text – Use clear titles, put sizes and specs in text rather than images, and list each variant separately.
- Spell out delivery and returns – State precise delivery times and your full returns policy where a bot can find them.
- Check you’re not blocking AI bots – Look at your Cloudflare and site settings to make sure “block AI bots” isn’t switched on by accident.
- Match your details everywhere – Keep price, stock, and returns identical across your product page, shopping feed, and structured data.
- Tag your links so you can see AI traffic – Add tracking to links you control so AI-referred sales don’t get logged as direct.
