Most companies with a data team have a specific kind of problem, and almost none of them name it directly.
It’s not a tooling problem. It’s not a talent problem. It’s not that you need better dashboards or a newer data warehouse or more AI features bolted onto your existing stack.
It’s operating debt.
What operating debt actually is
Operating debt is the accumulated cost of analytics work that was never designed to produce decisions.
It starts small: a dashboard gets built to answer a specific question. The question gets answered. Nobody retires the dashboard. Three months later, a second dashboard gets built for a related question. Now you have two dashboards, two sets of numbers, and nobody agrees on which one is right.
This compounds. A new analyst joins and builds their own version of the truth. A new VP requests a new dashboard. The data team, buried in dashboard assembly, doesn’t have the bandwidth to question whether any of it is producing decisions.
The tell is in the metric review: the meeting that has become a negotiation over which spreadsheet is authoritative, before any decision gets made.
The four signals it’s accumulating
Your analysts are reconciliation layers, not interpreters. They spend most of their time pulling data, reconciling numbers, and distributing reports — not spotting patterns or surfacing anomalies. This is operating debt at the human level.
Every metric review is a negotiation before it’s a decision. Not “here’s what the data shows” — “which version of the data are we using?” The CFO’s office is on a different number than the COO’s office. They don’t find out until the board meeting.
Decision speed hasn’t changed, but your dashboard count has. You have more data, more dashboards, more analytics headcount. Nobody can name a decision that happened faster than it did six months ago. That’s the debt compounding.
The moment you retire an old dashboard, the anxiety starts. Not relief — anxiety. Because now nobody knows which version of the truth they’re supposed to be working from. The dashboard wasn’t tracking a decision. It was tracking a negotiation that never concluded.
Why tooling doesn’t fix it
The standard move when decisions aren’t happening is a new tool: a BI platform migration, an AI pilot, a data mesh rearchitecture. These aren’t wrong. They also don’t fix operating debt. They add a new layer on top of it.
The debt lives in the operating model, not the technology stack. You can migrate to Snowflake and still have three dashboards that contradict each other. You can implement AI-assisted forecasting and still have the CFO’s office using a spreadsheet nobody’s data team has visibility into.
The fix has to start at the operating layer: what decisions do we make, from what data, on what timeline, with what owner?
What resolving it actually looks like
This isn’t a project with a start and an end date. It’s an ongoing discipline: retiring dashboards that aren’t tied to live decisions, locking metric definitions before building new reporting, requiring that every new data initiative be tied to a specific decision it enables.
The companies that have done this well share a common thread: they shifted from thinking “we have a data team” to thinking “we have decision infrastructure.” That reframe changes how every new dashboard request gets evaluated.
If that sounds like something your organization is ready to take seriously — let’s talk about where the drag starts.
Ready to work backward from the decisions your organization actually needs to make?