Retail wants to become an AI powerhouse. But first, it needs to get its data house in order.

The UK retail, brands and eCommerce market is moving quickly towards AI.

But behind the excitement sits a much less glamorous challenge: getting the data foundations right. 

Over the last year, one theme has come up repeatedly in conversations with Data, AI and Technology leaders across retail: the appetite for AI is enormous. Retail businesses are like excited puppies when it comes to the possibilities, and increasingly they aren't simply looking to experiment with a few use cases.

Some of the UK's largest retailers want to become genuine AI powerhouses, using the technology to transform customer experience, supply chains, personalisation, operations and ultimately the way the business makes decisions. 

That enthusiasm feels particularly noticeable when compared with more heavily regulated sectors such as financial services and insurance. These industries are also making significant investments in AI, but their adoption often sits alongside much more complex regulatory, risk and compliance considerations. Retail has a different set of pressures. Customer expectations move quickly, competition is intense and AI can have a very direct impact on revenue, efficiency and customer loyalty. For many retailers, there is a real sense that they need to move quickly or risk being left behind. 

But there is a fairly significant catch: you can't build an AI powerhouse on top of poor, fragmented or insecure data. 

The retail data maturity gap 

One of the most interesting things about the current market is just how different retailers' starting points are. 

Some of the UK's largest and most established businesses have spent years investing in data, cloud, digital and technology capabilities and are now in a position to scale their AI ambitions. Tesco, Sainsbury's, and Kingfisher are all examples of retailers applying AI across areas such as personalisation, forecasting, supply chain optimisation, customer service and operational efficiency. 

The conversation for these organisations is increasingly about scale: how do we take AI beyond experimentation and embed it across the business? 

At the other end of the market, there are organisations still investing heavily in the foundations. Primark, for example, continues to develop its data and analytics capabilities, including its data warehouse, cloud-first approach, governance, data quality and architecture. B&M has similarly identified smarter data and customer insight as an area with significant opportunity as it continues to modernise its technology and processes. 

That doesn't mean these businesses aren't interested in AI. In many cases, they are just approaching it from a different starting point. Before you can build sophisticated AI capabilities, you need reliable data, modern platforms, strong architecture and people who understand how to bring those things together. 

And this is where the market gets particularly interesting. Some retailers are asking, “How do we scale our AI strategy?”, while others are still asking, “How do we get our data into a state where we can actually do AI properly?” 

Both are part of the same journey. 

Governance might be boring, but it matters 

Let's be honest: data governance isn't the most exciting part of the AI conversation. It is difficult to get people enthusiastic about data ownership, lineage, access controls, quality and policies when there is a shiny new AI tool sitting in front of them. 

But the more valuable and connected a retailer's data becomes, the more important those fundamentals are. And this week's news is another timely reminder of why.

On 6 October, ASOS confirmed it was investigating unauthorised activity involving third-party platforms used to communicate with customers after an unauthorised notification was sent to shoppers. The company said basic personal information, including names and contact details, may have been accessed, although it does not believe payment-card information or account passwords were affected. The National Cyber Security Centre has also confirmed it is aware of the incident. 

The incident came after ASOS customers received an alarming notification through the retailer's own app, with the attackers apparently attempting to draw attention to the compromise and extort the business. ASOS has restricted access to the affected notification platforms and is investigating with specialist advisers and the relevant authorities. 

And ASOS isn't an isolated example. 

The cyber incidents affecting Marks & Spencer and Co-op in 2025 demonstrated just how significant the operational and financial consequences can be when major retailers experience security incidents. 

These events aren't simply “data governance problems”, and stronger governance alone cannot prevent every cyber-attack. But they reinforce something the retail industry cannot afford to overlook: as businesses collect, connect and use more data, they need to understand where that data sits, who can access it, how it is protected and how resilient the underlying technology is. 

That becomes even more important as AI increases both the volume and the value of the data being used. 

The AI race is becoming a data maturity race 

This is perhaps the biggest shift we're seeing across the sector. The question used to be “Who is using AI?” Increasingly, the more useful question is “Who has the foundations to use AI effectively at scale?” 

The retailers furthest along their AI journey aren't necessarily those with the biggest AI announcements. They are often the organisations that have already invested in understanding their data, modernising their platforms and building the teams needed to turn technology into something commercially useful. 

That also explains why the people side of this transformation remains so important. 

AI isn't removing the need for Data Engineers, Architects, Governance specialists, Data Product leaders, Analysts, AI professionals and Transformation specialists. If anything, the skills required are becoming more sophisticated. Retailers need people who can bridge the gap between technical capability and commercial outcomes: understanding the data platform, but also why the merchandising team needs it; understanding AI, but also how it can improve customer behaviour and business performance; and building governance that enables innovation rather than simply creating another layer of bureaucracy. 

That combination of technology, data and people is becoming increasingly valuable.

So, where does retail go next? 

The next phase of AI in retail is unlikely to be about who can announce the most impressive AI initiative. It will be about who can operationalise AI and make it work consistently across the organisation. 

There will continue to be a huge difference between retailers that are already building sophisticated AI ecosystems and those still modernising their data estates. That isn't necessarily a problem. Every organisation has a different starting point, and the important thing is understanding what needs to happen next. 

What is clear is that the ambition is already there. Retail doesn't just want to experiment with AI; some of the biggest players want to make it a genuine competitive advantage and become AI powerhouses in their own right. 

The challenge is making sure the foundations can support that ambition. 

AI might be the exciting part of the transformation, but data, governance, technology and people are what will ultimately make it work.