Up to this point in the series, we’ve been talking at a higher altitude.

We’ve explored why product data matters, how bad data quietly taxes the business, where ownership tends to break down, and why tools and org charts alone don’t solve the problem. Those conversations were intentional. Before you can build anything meaningful, you have to see the problem clearly and agree that it’s real.

However, at some point, the conversation has to move out of theory and into structure. Not structure as in “another initiative,” but structure in the sense of how the business is actually going to manage product data as part of day-to-day operations. This is where the series turns from perspective to practice.

What follows is the foundation for that shift.

When Jay and I first started talking through this work, we kept circling the same explanation without trying to. Every time we described why things were harder than they should be, or why progress kept stalling, we ended up talking about fuel.

The problem was not the car. The problem was not the engine. The problem was the fuel.

Every business understands fuel in a very practical way. Fuel isn’t the machine. It isn’t the destination. And most of the time, when it’s working, you don’t think about it at all. But when fuel quality is inconsistent, everything else starts working harder than it should. Performance drops. Maintenance increases. People spend more time compensating than improving.

Product data behaves the same way.

One of the reasons product data is so hard to manage is that we tend to collapse different things into the same word.

Content is raw material. Supplier spreadsheets, PDFs, spec sheets, emails, shared drives, legacy files, and a fair amount of information that lives only in people’s heads. Most organizations have no shortage of content. In fact, they’re usually drowning in it.

Product data is what content becomes after it has been interpreted, structured, standardized, and made repeatable. It’s content that has been refined to a point where the business can actually rely on it.

And the systems we depend on — ERP, PIM, eCommerce platforms, pricing tools, warehouse systems — are engines. Engines don’t refine fuel. They assume refinement has already happened.

A lot of organizations think they have a data problem when what they really have is a refinement problem. Raw material is being poured straight into engines that were never designed to clean it up, and the organization quietly absorbs the consequences.

Everyone wants better performance. Better search. Better merchandising. Better digital experiences. Better analytics. These are visible, exciting, and easy to rally around.

They’re also downstream.

No one would try to tune an engine or push it harder while feeding it unstable fuel. You don’t start by polishing the dashboard if the fuel line is full of sediment.

Refinement exists to protect the system.

When refinement doesn’t happen centrally, it happens locally. Sales cleans things up just enough to quote. Operations fixes what breaks fulfillment. Customer service patches what customers notice. Marketing works around gaps with copy and disclaimers.

None of this is incompetence. It’s adaptation. It’s how businesses survive when inputs aren’t stable.

The problem is that now you don’t have one fuel grade moving through the organization. You have many. And they’re all being consumed by systems that assume consistency.

At that point, performance problems stop being solvable with tools or projects. You can add horsepower, but you’re still feeding the engine something different every day.

When refinement is inconsistent, the business doesn’t stop. It compensates.

You see manual checks where automation should exist. Spreadsheets that “just help this one team.” Fields that technically exist everywhere but mean something slightly different depending on who you ask. Simple questions turning into long email threads because everyone is trying to reconcile their version of the truth.

This isn’t failure. It’s the organization protecting itself from bad fuel.

But that protection has a cost. It consumes time and attention. It slows improvement. And it makes every change feel risky, because no one is quite sure what else it might break.

Leadership doesn’t need to manage individual data fields any more than it needs to manage individual barrels of fuel. What leadership actually manages is the refining operation.

That means deciding where refinement happens, in what order, with what standards, and with what accountability. It means treating product data as a managed commodity, not a byproduct of other work.

As fuel is refined, it becomes safe for more parts of the business to depend on it. Each step of refinement increases the number of business functions that can consume the data without correcting it, reinterpreting it, or working around it.

That’s the operating mindset behind the Product Data Framework.

To make this concrete, the framework follows a deliberate sequence of refinement stages. Each stage represents a higher grade of fuel and brings additional business functions into scope.

At a high level, those stages are:

• SKU Definitions – establishing clear, stable product identity
• Regulatory Data – ensuring the product can legally and safely be sold and handled
• Operational Data – enabling storage, movement, and planning without surprises
• Fulfillment & Pricing – supporting accurate ordering, shipping, and margin control
• Marketing & Merchandising – powering discovery and presentation once the fuel is stable
• Cross-Reference & Substitution(The area every company wants to perfect first!) – enabling intelligent comparison and replacement only after trust exists

Each stage builds on the one before it. You’re not just “adding more data.” You’re refining it to a point where more of the business can safely rely on it.

Why Marketing and Search Are So Tempting — and So Often Mis-Sequenced

Cross reference, Marketing data and search improvement are powerful. They’re visible. They show up in demos. They feel like progress.

That’s why so many organizations want to start there.

But starting with marketing data without refining the fuel first is like installing a high-performance engine and hoping it somehow cleans up whatever you pour into it. It won’t. It will amplify inconsistency, not resolve it.

The refining process isn’t about slowing progress. It’s about making sure that when you get to those stages, the data can actually support them without forcing humans to absorb the damage.

Before moving on, think about one operational issue that shows up regularly in your organization. Ask where the information behind it originates, where it is refined today, and who ends up compensating when it isn’t clear.

If the answer feels scattered, you’re not looking at a performance problem. You’re looking at a refinement problem.

And refinement is something you can manage.

Now that we understand the analogy, it should help us follow the entire process of refining.  We will start next issue with the 1st Refinery: SKU definitions. Not because it’s exciting, but because everything else assumes you can answer one basic question consistently — what this product is, and whether the rest of the business can trust that answer without reinterpretation.

That’s where the real work begins.