By Heywa Labs

Your customers don’t speak catalogue. They’re planning a walking holiday, finding a gift for someone difficult to buy for, or trying to make a small living room work. Sometimes they know exactly which product they want. Often, they need help working out what to buy.

Yet the familiar ecommerce journey starts with the retailer’s organisation of its stock. Choose a category. Apply some filters. Scroll through products. The customer has to translate their situation into the language of the catalogue, then do much of the thinking themselves.

My view is that AI gives retailers an opportunity to change that starting point. A shopper should be able to describe what they’re trying to achieve, and get an experience that helps them achieve it. That has implications for the whole interface, from the first question to the final selection.

Start with the shopping mission

Consider someone preparing for their first camping weekend. They may know they need a tent, but have little idea what distinguishes the options, what else is essential, or how to spend a limited budget.

“Two nights camping with my daughter in May, travelling by car, with £300 to spend” gives a retailer a useful brief. A good experience could ask how many people need to sleep in the tent, explain where it is worth spending more, and assemble a suggested kit. The shopper could swap an item and see how that affects the total.

The mission connects products across categories and gives recommendations a reason. A sleeping mat belongs in the selection because it helps complete the trip. That is a more useful basis for a suggestion than simply knowing which product page someone opened.

This approach is particularly relevant where customers need advice or are buying several things that must work together. Exact product search still has a clear role. Someone reordering a familiar moisturiser should be able to do so immediately.

Let the interface help people choose

Understanding a request is only part of the work. Retailers also need to decide how to present the answer.

Conversation helps a shopper explain their needs and refine them. But imagine choosing a sofa through paragraphs of text. You want to see the sofas, compare their dimensions, check the fabrics and understand what each option means for your room.

An effective journey might start with a question about space and budget, then show a visual shortlist. A comparison could highlight dimensions and practical differences. The customer could request something smaller, or choose a favourite and explore similar options, without starting again.

I use “Generative UX” to describe this idea: the interface adapts to the task and the customer’s context. The retailer defines reliable components and rules, while AI helps select and populate the views that are useful at each stage.

For a wardrobe, that might be a collection of pieces shown together. For a coffee machine, it could be a comparison organised around daily use and maintenance. The format should make the decision easier.

Bring retail expertise into discovery

A retailer’s expertise can become part of this journey. Which products work well together? What should a beginner prioritise? Which tradeoff would an experienced adviser explain before recommending a purchase?

That knowledge needs to be supported by dependable product information. Stock, prices, dimensions and compatibility must come from reliable sources. If the system cannot establish whether an accessory fits, it should make that uncertainty clear. Confident wording cannot compensate for missing facts.

Commercial priorities matter too. Retailers will want to account for availability, margin and merchandising commitments. Those choices need to operate within the shopper’s brief. Recommending an unsuitable product undermines the advice, however attractive its margin.

There is an opportunity here to express a brand through the help it gives. A specialist outdoor retailer and a department store might approach the same holiday brief differently, reflecting their assortment and expertise. The resulting experience should feel like shopping with that retailer.

Begin with one journey worth improving

For retailers considering this approach, I would start with a specific mission where the existing journey asks too much of the customer. Preparing for a new hobby or furnishing a compact room gives teams a concrete problem to work on.

Map the decisions involved. What does the shopper need to tell you? Which information is essential to a recommendation? Where would a visual collection help, and where is a comparison more useful? Ask only questions that materially change the advice. A lengthy interview can become another barrier to shopping.

Keep the experience connected to the rest of the website. Customers should be able to inspect product details, edit a selection and add items to their basket without losing the context they have already provided.

Then test it against the existing journey. Conversion and basket value matter, but so do the suitability of purchases and the cost of serving each mission. Watch for returns, latency and abandonment. Time spent talking to an assistant is not, by itself, evidence that shopping has improved.

The requests can also reveal gaps in an assortment or in product information. Repeated questions about whether furniture fits through a doorway, for example, suggest information the retailer should make easier to find across the site.

Build around the customer’s goal

I believe the next chapter of ecommerce will be shaped by experiences that can understand a customer’s situation and help them work through a purchase. Conversation opens that possibility; visual, interactive journeys make it practical.

For retailers, a useful starting question is: which shopping mission could we help someone complete much better than we do today? Answering that gives an AI initiative a clear purpose and something meaningful to measure.

At Heywa Labs, we’re putting these ideas into practice with Heywa Studio, helping brands build visual, interactive shopping experiences around their customers’ goals. If you’re exploring what this could look like for your brand, we’d love to hear from you. Come talk to us.

Published 02/11/2026