Why aren’t more retailers AI-ready?

For many retailers, artificial intelligence adoption has become a race. 

Competitive pressure, board-level expectations and a steady stream of success stories have created urgency to act quickly. But in many cases, speed has replaced clarity.

Many retailers would argue they are already “AI-ready”. Retail is, after all, one of the most data-rich industries in the world. Yet AI initiatives continue to stall, underperform, or quietly get reset. The reasons for this have far less to do with algorithms than most would like to admit.

Too often, AI initiatives are launched before organizations have agreed on what problems they are trying to solve, which decisions they want to improve, or what data those decisions actually depend on. Tools are selected, dashboards built and models deployed, all before the foundations are in place. In some cases, AI is applied to problems that did not require it in the first place, simply because it is seen as the thing to do.

Retailers are not failing to adopt AI. The enthusiasm is there. The investment is there. What is missing is the foundation that makes AI worth adopting at all.

The overlooked foundation: data governance

AI readiness does not start with technology. It starts with a question: where can we use intelligence to remove barriers that are holding us back from achieving our business goals? Once that is clear, the next step is ensuring the data that underpins those decisions is usable, consistent and trusted. That is the role of master data governance.

For many retailers, governance still carries the wrong connotations. It is often viewed as a compliance exercise, owned by a central team, focused on controls, policies and box-ticking. Necessary, but slow. Important, but disconnected from commercial outcomes. That perception is outdated.

Modern data governance, supported by automated master data management approaches, is practical and operational. Core retail data domains such as products, pricing, suppliers and locations follow repeatable patterns. Governance frameworks no longer need to be invented from scratch for each initiative or business unit.

With the right structure and expertise, getting data into a governed, usable state is a straightforward, repeatable process. Rules for how data is created, validated and changed can be embedded directly into day-to-day workflows rather than enforced retrospectively. Quality improves as part of normal operations, not through constant clean-up exercises. 

In other words, governance is no longer the complex, high-risk undertaking many retailers still assume it to be. The technical capability exists, and the path to implementation is well understood.

So, if governance is no longer the barrier, where is the real challenge?

The non-technical reality of AI readiness

The feeling of “we need to implement some form of AI, but we don’t know what”. The lack of a foundation. Misconceptions about the role of data governance. All of these are symptoms of a bigger problem. Organizational mindset. 

Retailers often skip the foundational steps. Instead of starting with governance, then harmonising data through agreed standards that reflect both industry norms and their own operating model, they jump straight to operationalizing data. They try to optimize routes, personalize offers or automate decisions without first agreeing how the underlying data should be owned, defined and improved.

This happens not through bad intent, but through lack of understanding. Governance is still seen as overhead rather than enablement. Leadership teams may not fully appreciate the role it plays in accelerating outcomes, not slowing them down.

There is also the organizational hangover of legacy thinking. “What we have works, so why change it?” The problem is that processes lose value over time. Data volumes increase, complexity grows, and quality becomes the deciding factor in whether AI delivers value or multiplies problems. Governance models that worked ten years ago are no longer fit for purpose today.

The reality is that modern master data management approaches remove much of the historical pain. They save time, improve quality and create the conditions AI depends on. The question is no longer why governance, but why not.

Why mindset matters more than tools

The harsh truth is that AI initiatives struggle to scale, not because governance is too difficult, but because adopting it properly requires a shift in mindset. Data stops being something teams work around and becomes something they actively manage.

Retailers already have the technology they need to become AI-ready. What will determine success is whether they are willing to change how they think about data, governance and responsibility.

Until that happens, AI will remain an ambition rather than a capability.

Richard Henry is commercial director of Bluestonex.