Depending on which headline you read, AI will either solve all the world's problems or end humanity as we know it. With the news around AI and the tech itself developing at breakneck speed, it's becoming harder to separate fact from fiction. Retailers are facing their own version of this dilemma, with agentic commerce and the future of shopping dominating discussion. 

Artificial intelligence will transform retail. It is already having a huge impact on customer care, returns and logistics. I see the next big opportunity in customer experience, inspiration and discoverability. 

AI covers many technologies and applications. Treating them as one conversation makes it hard to remove the hype from reality. But retailers should be asking a much simpler question: what can AI do for customers and the business today? 

Start with the problem, not the prediction 

Despite the talk around AI, adoption is still in its early stages. My advice to retailers is to start with the problems you already face and ask how AI can help solve them.

Where does the shopping journey break? Which manual tasks are most labour intensive? Which product lines aren't getting the attention they deserve? 

These questions often point to a clear opportunity, helping shoppers find inspiration and the confidence to buy. At StoryStream, we connect visual content with relevant products and outfits. This makes products easier to discover and retail websites worth exploring. AI has a role in each step.

User-generated content on SweatyBetty.com showing products on real people in real-life situations

Discovery is changing  

For shoppers, discovery is about more than finding a product. Social media has changed how people browse, with user-generated content (UGC) showing products in real-life situations. Static product grids do not offer the same inspiration. 

Imagine you are shopping for a holiday. You could search for “summer outfits” and work through the results. Or you could ask an AI assistant: “What should I wear for warm evenings in Spain?” The difference is context. You are asking for inspiration that fits your plans, not just an endless and often unhelpful lists of products. 

AI can help turn that search for inspiration into product discovery. Shoppable UGC is one-way retailers are doing this, bringing approved creator, influencer, customer and brand content onto their websites as social-style galleries. Sweaty Betty saw a 12% uplift in average order value after implementing the AI-powered tool. 

Catering for the invisible audience 

Retail websites have always been designed to win over people. Now, they need to win over machines too. AI agents are increasingly sitting between shoppers and retailers, researching, comparing and deciding which products make the cut. 

A product page can look engaging and informative to a shopper yet leave important information unreadable to an AI tool. Product descriptions, prices and other content can be missing from the version an agent sees. A polished front-end is only half the job. 

That makes the foundations, product data, descriptions and metadata, essential. Images and videos also need to be connected to the products they feature and the context they provide. It is not enough for a shopper to recognise a product in an image; an AI agent needs to understand that connection too.  Our customers are already embracing this, using AI-powered tools like StyleGraph to link visual content with richer product information and context, so AI engines can better understand, surface and recommend their products.

AI-powered Shop The Look in action on Primark.com - StyleGraph AI automatically matches products to model imagery, matches outfits according to Primark's style guide, and manages in/out of stock inventory.

Make the visit worth their while 

Helping shoppers find products on your website is only part of the challenge. Someone thinking about buying a statement jacket wants to know what to pair it with. A complete look can give them the confidence to add another item.  The challenge is delivering this across a changing catalogue without more manual work. 

AI can help solve this. Recommendations must reflect the brand, merchandising rules and stock availability, not just visual similarity. An outfit is not much use if the suggested products have sold out. 

Primark offers a great example of what this looks like in practice. By using AI Shop the Look, they have seen a 5% uplift in average order value on those journeys. Outfits are automatically styled and tagged across more than 2,000 products each month, scaling the merchandising team’s expertise rather than replacing it. 

This is what excites me: AI solving real problems with measurable commercial results, without adding manual work. 

Build for today. Be ready for what follows 

AI is already revolutionising our industry, but the most exciting transformations are yet to come. My advice? Start with one opportunity in your own customer journey. Set a clear measure of success, test and build from there. Use AI to make your products easier to discover and your website worth visiting.  

And look beyond industry announcements to cut through the hype. Look at how your customers are using AI when they shop. Start with those everyday experiences, and you can build for the way people actually shop - not just the way the industry predicts they will.

Sources and further reading

• How community-led shoppable content raises Sweaty Betty’s AOV and keeps customers engaged

• Primark scales AI Shop The Look across 2,000+ products per month, lifting AOV by 5%