John Lewis Says AI Shopping Has Surged Eightfold in a Year — and Retail Is Starting to Change for Good
The AI Personal Shopper Arrives
When the Algorithm Goes Shopping
A quiet revolution is starting to change how Britons shop online. John Lewis says searches reaching its products through artificial-intelligence agents have risen from just 0.3% a year ago to 2.5% today — an increase of more than eightfold.
The number is still small compared with conventional online shopping, but the speed of the rise matters far more than its current share. Customers are increasingly asking AI systems to find, compare and recommend products rather than manually working through search results themselves, opening a new battle over which retailers and products machines choose to put in front of them.
AI Shopping Has Moved From Experiment to Measurable Behaviour
John Lewis managing director Peter Ruis described the growth as exponential and said it was appearing across age groups rather than being confined to younger technology enthusiasts. That matters because one of the biggest questions surrounding generative AI has been whether conversational shopping would become a genuine consumer habit or remain an interesting novelty.
At John Lewis, the direction is becoming harder to dismiss. Moving from 0.3% to 2.5% means the proportion attributed to AI agents has increased by roughly 8.3 times in only a year.
The conventional online shopping journey is familiar. A customer opens a search engine or retailer website, enters a few keywords, looks through dozens of products, changes filters, compares reviews and prices and eventually chooses what to buy.
An AI agent can compress much of that process into a conversation. A shopper might instead ask for a coffee machine below a particular price, suitable for a small kitchen, easy to clean and highly rated, before allowing the system to narrow hundreds of possibilities to a handful of recommendations.
The critical difference is that the retailer's website may no longer be where that journey begins.
John Lewis Has Already Bet on the Shift
John Lewis had seen this change coming. In March it announced a significant investment intended to make its products more visible when customers seek recommendations through AI platforms including ChatGPT and Google Gemini.
Its ambition goes beyond appearing in AI-generated recommendations. The retailer said customers would ultimately be able to transact through third-party AI applications when the necessary technology becomes available in Britain, reducing the distance between asking an AI what to buy and actually purchasing it.
That investment forms part of John Lewis's broader £800 million multi-year transformation programme. The company has been trying to place its products wherever consumers increasingly discover them rather than assuming shoppers will always begin on JohnLewis.com or walk into one of its stores.
The same strategy has included social commerce and a greater emphasis on digital content. John Lewis has now opened a content studio at its Oxford Street flagship where creators can produce material, alongside its own celebrity-led video programming.
That may look separate from AI shopping, but the two trends are becoming increasingly connected. AI recommendation systems need information about products, brands and consumer preferences to decide what is relevant. Retailers therefore face growing pressure not merely to advertise to humans but to make their products intelligible and discoverable to machines.
The Search Bar Faces a New Competitor
For more than two decades, digital retail has largely revolved around search. Google helped people find a retailer, while the retailer's own search engine and category pages helped them find a product.
Agentic commerce potentially inserts another layer between the customer and the shop.
OpenAI has already developed shopping experiences that allow people to describe what they want conversationally, compare products and, for eligible merchants and markets, move towards completing purchases inside ChatGPT. Google is simultaneously developing its own agentic-commerce infrastructure across Search and Gemini, including technology designed to allow AI systems, merchants and payment services to communicate.
The destination therefore becomes less important than the recommendation.
That represents an enormous opportunity for retailers whose products are frequently selected by AI systems — and a serious threat to businesses that become effectively invisible to them.
Traditional search-engine optimisation has spent years trying to answer one question: how does a business appear near the top of Google?
Retail could increasingly face another one: how does your product become the answer an AI gives?
AI Could Change Who Holds Power in Online Retail
The consequences go beyond convenience. Product discovery has always been valuable because whoever controls discovery can influence where billions of pounds in consumer spending eventually flow.
Department stores traditionally controlled that discovery through shop floors, displays and knowledgeable staff. Search engines later became enormously powerful digital gatekeepers. Marketplaces then built recommendation engines of their own.
AI agents could become the next layer.
Instead of consumers examining 50 products themselves, an AI system might examine hundreds and return three. For the shopper, that could mean less time wasted. For a retailer or manufacturer excluded from those three recommendations, it could mean losing a customer before they ever visit the website.
It also raises new questions about how recommendations are generated, how current product information remains, whether sponsored placements can be distinguished from independent recommendations and how much influence individual AI platforms could eventually exert over the purchase funnel.
The technology companies themselves are already building infrastructure around that future. Google's Universal Commerce Protocol is intended to create a common framework through which AI systems and retailers can exchange product, checkout and customer information, while OpenAI's Agentic Commerce Protocol is designed to connect conversational product discovery with merchant systems.
The foundations for AI-mediated retail are therefore being built before most shoppers would necessarily describe themselves as users of "agentic commerce".
John Lewis Still Has to Win in a Difficult Economy
The technology push arrives at a difficult moment for British retailers. John Lewis has warned of challenging economic conditions as households remain cautious about discretionary spending.
Its department-store business nevertheless entered the year with some momentum. John Lewis reported sales of £4.9 billion for the year to January 2026, up 3%, while adjusted operating profit increased to £58 million.
Digital commerce is already central to the business. Around 60% of John Lewis orders now take place online, meaning even a relatively small change in how consumers discover products can become commercially significant if the trend continues.
Leadership is changing too. Ruis is due to leave the managing-director role on September 6, with Will Kernan succeeding him, while the John Lewis Partnership is scheduled to publish its half-year results on September 10.
Those numbers will provide a broader picture of how the retailer is performing. The AI statistic, however, offers an early glimpse at something potentially much bigger than one company's trading figures.
The Bigger Battle Is Over Who Chooses What We Buy
AI agents do not need to replace conventional ecommerce overnight for the economics of online retail to start changing. They merely need to capture enough of the discovery stage that retailers can no longer afford to ignore them.
John Lewis's jump from 0.3% to 2.5% remains an early-stage signal rather than proof that traditional online shopping is disappearing. Most customers are still arriving through other channels, and it remains unclear how quickly AI-driven discovery will convert into large-scale purchasing behaviour in Britain.
But eightfold growth in a year is difficult to treat as background noise.
If that trajectory continues, the most important digital shop window of the next decade may not belong to a retailer at all. It could be the conversation consumers have with an AI before they ever see a shop, advertisement or product page — with algorithms increasingly deciding which handful of products get the chance to compete for the sale.
