Your Platform Readiness Score

Insights

Why retail data strategies are failing in the age of AI

The retailers winning the data race aren’t the ones with the most data, the biggest lakes or the most dashboards. In an agentic world are dashboards and reports now dead?

In today’s retail landscape, where AI, data strategy and real-time decision-making are converging, this question is becoming central to how organisations rethink their data foundations.

It’s a topic I explored recently during my tech talk on stage at Retail Technology Show, and it’s the conversation that’s happening in boardrooms across UK retail right now.

It goes something like this. The ambition is clear: unified commerce, AI-driven personalisation, real-time inventory, a single view of the customer across every channel. The vision is agreed. The investment case is approved. The programme begins. And then, somewhere between ambition and execution, the wheels come off.

Not because the vision was wrong. Not because the technology doesn’t exist to deliver it. But because the underlying data strategy and ontology aren’t aligned, leaving data disconnected before you even begin.

Retailers don’t have an AI problem, they have a data problem

Across the market, the pattern is consistent: what looks like an AI problem is, in reality, a data problem.

Data only becomes valuable when it informs decisions in the moment they matter, not when it reports on what has already happened. That shift isn’t about volume or more sophisticated platforms. It forces a more fundamental question: what is your data actually for, and what decisions is it meant to support?

Which raises a harder operation question: does reporting through looking back still serve the business? Does giving your store manager last week’s sales report now feel almost worthless?

AI is forcing organisations to rethink their relationship with data. It is no longer optional; it is a question of competitiveness. Over the last five years the focus was on data platforms, lakes and analytics. The next five will be defined by investment in real-time, accessible, and comparable data design. Focused on exposing data directly from source into agentic systems.

What good retail data architecture must deliver

It requires three things to be true.

Data must be connected across domains, not fragmented across platforms with competing models and definitions.
It must be trusted at source, with consistency enforced upstream rather than reconciled downstream.
It must be available in the flow of operations, with latency measured in seconds, not hours, and structured in a way that reflects how the business actually runs.

Without this, data remains disconnected from execution. It is collected, stored and reported on, but rarely used to drive decisions at the speed and precision required.

Why retail data strategy and transformation programmes fail

On the surface, organisations appear data rich. This is exactly where most transformation efforts begin to break down. It’s a pattern I’ve seen more times than I care to count.

Across the industry, the same failure mode is playing out. Retailers invest heavily in AI-powered platforms and use cases: personalisation, search, forecasting and customer service automation before addressing the underlying data foundations. Early pilots show promise. But when deployed at scale, the cracks appear.

The instinct is to blame the AI. But the issue sits deeper: this isn’t an AI failure, it’s a data failure.

It’s why at PMC we made a decision to stop framing transformation as a technology problem and start talking about it as a data problem that technology can solve, if you approach it in the right order.

UK retail data strategy challenges: what our latest research reveals

We partnered with Retail Economics to survey 100 senior leaders across UK retail and direct-to-consumer brands. The findings were stark but not surprising to anyone who spends time close to the reality of retail operations.

Only 39% of those businesses have fully centralised their analytics, creating a single, real-time view of sales, inventory, and customer behaviour across every system and channel.

Six in ten retailers are navigating one of the most competitive markets in a generation without a complete picture of their own business.

That’s not a technology gap. It’s a strategic vulnerability. And in a market where the distance between the leaders and the laggards is compounding every quarter, it’s one that demands a response.

The business impact of connected retail data platforms

The contrast with businesses that have done this work is not marginal. Among the retailers in our research who have built integrated data layers, 58% report significantly faster decision-making, 48% cite measurable revenue growth and 45% report meaningful cost savings.

These aren’t incremental gains. They’re structural advantages that accumulate over time and become increasingly difficult for competitors to close.

The retailers who invested in connected data infrastructure three years ago are not just ahead – they’re running a different race.

Why AI in retail fails without clean, connected data foundations

There is a dimension of this problem that remains underappreciated across the industry.

AI is rapidly becoming the defining competitive battleground in retail. Whether in personalisation, demand forecasting, dynamic pricing or customer service – every retailer I speak to has AI somewhere on their roadmap. Most have it near the top.

Our research found that 74% of mature digital transformers have fully implemented AI personalisation. Among those still earlier in their transformation journey, the figure is 38%. That gap is not primarily about AI investment or capability. It’s about data readiness.

Many retailers still confuse data centralisation with data readiness. Moving data into a lake or warehouse is part of the journey, but it does not make that data trusted, usable or meaningful, nor does it give AI the context it needs to operate effectively.

For AI to operate effectively, data must carry operational meaning: aligned models across domains, consistent identifiers across channels, and latency measured in seconds, not hours. Without this, AI operates on approximation, not reality.

And that is why pilots fail to scale. Investment in AI without investment in data connectivity is not a transformation strategy, it’s an expensive experiment.

Solving retail data fragmentation and legacy data challenges

These challenges are not unique, and they are solvable.

Every retailer we work with at PMC started somewhere. Most started with fragmented systems, inconsistent data, and a technology estate that had grown organically rather than by design. That’s not failure. It’s the reality of running a retail business through a decade of rapid change.

There is no one-size-fits-all blueprint for AI readiness. Some organisations are working with modern cloud platforms, whilst others are still navigating legacy systems and decentralised ownership. The starting point is always different, but what matters most is aligning data strategy to business outcomes.

What we’ve learned, and what Graphene was built to address, is that the path from fragmentation to connection doesn’t require a big bang transformation. It doesn’t require ripping out everything that works in order to replace it with something untested.

It requires a clear-eyed assessment of where the data breaks down, a disciplined approach to connecting the systems that matter most, and a partner who has done this enough times to know where the complexity will come and how to navigate it.

That is what PMC exists to do. Not to sell technology for its own sake, but to help retailers build the data foundations that make every subsequent investment in AI, in personalisation, in unified commerce, actually deliver what it promises.

The key question retail leaders should ask about data and AI readiness

If you’re a retail leader reading this, I’d invite you to sit with one question: What decisions are we making right now that we would make differently if our data was fully connected?

If the honest answer is uncomfortable, the work is clear. The retailers who get this right in the next eighteen months will have a structural advantage that defines the competitive landscape for years. The ones who wait will find themselves starting from further back than they realise.

Ready to move from fragmented data to decision-ready foundations?

We work with retailers to connect data across their business, enabling real-time decision-making, scalable AI and measurable commercial impact.

Speak to our team to explore what this could mean for your organisation.