Your Data Catalog is a Ghost Town

The playbook I used to fix this for Mars, Inc. (and how you can use it too)

Most data catalogs are ghost towns. Despite the investment in implementation and integration, adoption flatlines. The search bar sees no traffic, and meticulously documented tables gather digital dust.

This is a common outcome of data catalog projects. The root cause is rarely the technology itself. The problem isn’t the tool. The problem is the strategy. We treat the catalog as a project to be completed, a box to be checked, rather than a living, breathing resource to be integrated into the daily workflow.

The result is a vicious cycle: stale data erodes trust, which hinders adoption, which in turn ensures the data remains stale. The tool meant to create instead becomes a symbol of wasted effort.

Diagnosing the Ghost Town: Why Did Everyone Leave?

Before we can fix it, we need to understand what went wrong. Ghost towns don’t happen by accident. They are created by a failure to provide sustained value.

1. The ‘Build It and They Will Come’ Fallacy: This is the most common pitfall. You can’t just drop a new tool into an organization and expect a behavioral revolution. Adoption is not a default state; it’s an active pursuit.

2. It’s a Repository, Not a Resource: If the primary function of your catalog is to be a passive dictionary that data people update for other data people, it’s doomed. For everyone else (marketers, finance analysts, product managers), it needs to solve a daily, painful problem. Can they find the data they need in under 30 seconds? Does it indicate whether it’s trustworthy?

3. The Curse of Stale Data: The fastest way to erode trust is with outdated information. A catalog that presents stale datasets or obsolete points of contact is effectively publishing its own irrelevance. Sustainability is a core requirement for a trusted catalog. A manually maintained system, by definition, is prone to becoming obsolete.

4. It’s Too Damn Hard: Clunky UX, slow research, complex permission structures. If it’s easier for an analyst to Slack a data engineer and wait two hours for a response than it is to use the catalog, the catalog will lose every time.

This cycle leads to wasted budget, wasted effort, and a reinforced cynicism that ‘data is a mess.’ The very tool meant to create clarity becomes a monument to chaos.

From Ghost Town to Thriving Metropolis: A Lesson from Mars

So, how do you turn it around? I faced a similar challenge on a large-scale program with Mars, Inc. The business problem was different (optimizing a multi-million-dollar retail promotion program), but the core data adoption challenge remains the same. We had a wealth of data, but it was underutilized, inconsistent, and not trusted to drive decisions. It was a ghost town in a different zip code.

Our success on that project, which led to a 12% increase in promotion ROI ($4.8 million in annual savings) and scaling across divisions, wasn’t just about better analysis. It was about applying a rigorous framework to fix the data ecosystem itself. We conducted a full data landscape audit, mapped source systems, reconciled master data, and, most critically, designed a new, trusted framework for data definition and reporting that people actually used.

This success demonstrated that the solution is never a one-time fix. It’s a continuous program that demands a clear strategy, measurable goals, and consistent communication. Fundamentally, it’s about shifting from being an implementer to being a community organizer.

I’ve developed this methodology into a step-by-step playbook, complete with the exact tools we used to drive adoption.