THE CUSTOMER IS CHANGING. THE SEARCH IS TOO.
MONDAY BRIEFING
What if one of the biggest changes coming to fashion retail isn't a new competitor, a new marketplace or even a new generation of social platform, but the way your customer searches?
For more than 20 years, retailers have built digital strategies around a relatively simple principle: get the customer to search, get onto page one of Google, win the click and convert the visit. SEO, paid search, product feeds, landing pages and increasingly sophisticated personalisation have all been built around that model. But the search journey is now being fundamentally rewired. Instead of typing a few keywords, scrolling through pages of results and deciding which products are relevant themselves, customers are increasingly able to describe what they actually want in natural language and let AI do the narrowing down.
That is a much bigger change than simply putting a chatbot on a website. It changes the role of search from finding products to understanding intent.
Imagine a customer saying: “I've got a summer wedding in Italy. I want something under £250, I don't want to look too formal, I need something I can wear again afterwards and I need it by Friday.” Traditional search would force that customer to translate those needs into keywords, apply filters and work through a long list of products. An AI-led shopping journey can potentially interpret the occasion, budget, destination, versatility and urgency together, and move towards a much more relevant shortlist.
And this is no longer simply a prediction about the future.
This week, Anthropic launched its Claude Commerce Agents blueprint, giving retailers reference implementations for shopping and merchant agents that can search catalogues, understand customer preferences, compare products, build carts and support customers through the shopping journey. Anthropic has also reported early results from retailers using shopping agents on Claude, including larger carts and higher purchase completion rates. Those are company-reported figures, not independent industry benchmarks, but the direction of travel is significant.
That's an important signal. Because AI is no longer simply being used inside the retailer to improve forecasting, customer service, content creation or operational efficiency.
It is increasingly becoming part of the customer journey itself.
THE SIGNAL: SEARCH ISN'T DISAPPEARING. IT'S BEING REWIRED.
Google is building the infrastructure for the same shift. Its Universal Commerce Protocol is designed to allow AI agents and merchant systems to work together across the shopping journey, from discovery and product information through to checkout and post-purchase support. Google is already bringing agentic commerce into AI-powered shopping environments, moving the technology beyond simply answering questions towards helping customers actually transact.
And that changes the question retailers should be asking.
For years, the question was:
“How do we rank?”
Then it became:
“How do we get the click?”
Then:
“How do we convert the traffic?”
The next question could be:
“How do we get recommended?”
We are already seeing retailers experiment with what that might look like. Frasers has launched Ask Frasers, an AI shopping assistant designed to let customers describe what they are looking for naturally, refine their preferences and receive recommendations based on product information. Frasers has reported early indicators showing conversion rates up to 25% higher than traditional search experiences. That is a Frasers-reported early result rather than an independently audited benchmark, but the strategic signal is clear: AI is beginning to move from answering questions to actively influencing the product decision.
The customer doesn't necessarily want 500 search results. They want the right answer. And that is particularly important for fashion because customers often don't know the exact product they want. They know the occasion, feeling, style, budget or problem they are trying to solve.
“I need an outfit for a wedding.”
“I want a jacket that looks expensive but is under £300.”
“I need something for work that doesn't feel corporate.”
“I want trainers I can wear with jeans and also travel in.”
“I need a dress for a holiday that works during the day and can be dressed up at night.”
These are not traditional product searches.
They are intent statements.
And the retailers that can structure their product data, content and customer experience around that intent may have an advantage over those that continue to optimise primarily around keywords. This is also why the implications extend beyond search.
If an AI system is going to recommend a product, it needs to understand what that product actually is. Product descriptions, imagery, reviews, availability, pricing, fit information, brand positioning and other product signals become increasingly important because the AI needs reliable information from which to make a recommendation.
Google is explicitly developing its commerce infrastructure around this interaction between AI agents and merchant systems. And we are beginning to see the shift in the UK market too.
John Lewis has said that AI-agent-driven shopping now accounts for 2.5% of its product searches, up from 0.3% a year earlier. The retailer has responded by increasing investment in content creation, recognising that the way products are discovered is changing. That number is still small. But the rate of change is what matters. Going from 0.3% to 2.5% in a year is not evidence that traditional search has disappeared. It is evidence that a new discovery channel is developing quickly. And that creates a much bigger strategic question for fashion.
Because if a customer asks an AI:
“Find me a great jacket under £300.”
They may not receive ten pages of search results.
They may receive three recommendations.
Suddenly, the question isn't:
“How do I get the customer to click on my website?”
It becomes:
“How do I make sure AI understands my brand well enough to recommend me?”
That could change the economics of digital discovery.
For years, brands have competed for visibility by buying keywords, improving rankings, increasing media spend and fighting for attention. Now there may be another layer.
AI recommendation.
And if AI increasingly determines which products make the shortlist, then brand visibility may become partly dependent on how well a business can be understood by machines as well as customers. That makes product architecture, data quality, reviews, availability, pricing, imagery, brand authority and customer experience increasingly strategic, not simply operational. But there is an even bigger issue underneath all of this.
WHO SHAPES THE CUSTOMER DECISION?
If AI increasingly sits between the customer and the product, does the customer relationship belong to the retailer?
The brand?
The marketplace?
Google?
Claude?
Or some combination of all of them?
This matters because the value of a customer isn't simply the transaction. It is the data, the behaviour, the preference, the loyalty, the next purchase and the relationship that comes afterwards. AI may become extremely powerful at discovery and recommendation. But retailers will still want to own the relationship. So the future may not be about AI replacing the retailer. It could be about who controls the layer between the customer and the retailer. And that is why this matters to fashion now. The search box is becoming a conversation. The results page is becoming a recommendation. And the recommendation could increasingly determine what the customer ever sees.
THE INSIDER'S PERSPECTIVE
The biggest mistake fashion businesses could make is treating this as another technology project. It isn't. This is a customer-discovery shift. The businesses that benefit won't necessarily be those with the most sophisticated AI, they will be the ones with the clearest proposition, strongest product data, best customer understanding and most relevant product architecture. If an AI agent is increasingly helping decide what a customer sees, then being visible is no longer enough. You need to be understandable, credible and recommendable. The question fashion leaders should be asking now is simple: if an AI was asked to recommend a product in our category today, would it understand why our brand should be in the answer?
That is the strategic challenge. Because the customer is changing. The search is changing. And the businesses that understand that shift early could have a very different relationship with the next generation of fashion discovery.
THE SIGNAL
The next generation of fashion search may not be about giving customers more products to choose from. It may be about giving them fewer, better answers. And if AI increasingly controls that shortlist, the next battle for the fashion customer may not be for the click.
It may be for the recommendation.
THE FASHION THINK TANK
INSIDE FASHION. NOT OBSERVING IT.
