Your company doesn't have a cost problem. It has a leak nobody ever put in the budget.
Leadership teams can squeeze every zloty out of the fixed cost line. They renegotiate leases, renegotiate vendor contracts, trim marketing budgets by 10%. And at the same time, they lose far more every single day — in a place no financial report shows, because it isn't a cost. It's a leak.
This article answers one question: where does your margin physically disappear when nobody's watching — and why don't you see it in Excel, in the P&L, or in the boardroom.
Data Debt Isn't a Technical Term. It's a Bill That Arrives With Interest
Data Debt is the sum of every "temporary" workaround, manual fix, and inconsistency in a company's data that nobody ever fixed, because "it worked." Like any debt, it doesn't hurt until it comes due — at the worst possible moment: during an audit, during a funding round, during a decision that needed to be fast and accurate.
EBITDA Leak is the direct financial consequence of Data Debt — a measurable erosion of operating margin that doesn't come from one decision or one mistake, but from thousands of small frictions in everyday data work: repetition, corrections, waiting, guessing.
The distinction between a "cost" and a "leak" matters. A cost is visible — it has a budget line, an owner, a limit. A leak is diffuse — nobody is "responsible" for it, because it's made up of dozens of small decisions made by dozens of people, each of which seems reasonable in isolation.
A company doesn't go bankrupt from one mistake. It bleeds margin through a thousand invisible frictions — and calls it "a normal day at work."
This is exactly why EBITDA leakage rarely makes it onto a board agenda — it has no natural home in the standard reporting structure, even though its financial consequences often outweigh many "real" costs.
Why Most Leadership Teams Don't See It
Standard financial reporting is built to catch costs — not friction. A P&L will show you salaries, licenses, rent. It won't show you the three hours an analyst spent today manually merging two spreadsheets because System A and System B "don't talk" to each other.
There are a few reasons this pattern is so persistent:
- The leak has no single culprit. Nobody decided "we will lose 40 hours a month manually wrangling Excel." It's simply the sum of individually sensible actions.
- The cost of friction is spread across many people, so it looks small to each of them. Two hours a day for one person is a rounding error. Two hours a day for twenty people is half a full-time role — every single day.
- Effects show up late, and in a different place than the cause. A data error in last week's sales figures surfaces a month later as a mispriced inventory position. Nobody connects the two.
This makes EBITDA Leak one of the few business problems that grows proportionally with company size, rather than shrinking. The bigger the organization, the more systems, the more hands touching the same data — the more places friction compounds.
[LINK: DataMinq homepage / "Data Debt" philosophy section]
The Three Places Where EBITDA Leaks Most Often
Based on experience working with service businesses, e-commerce companies, and organizations preparing for a transaction, EBITDA leakage almost always concentrates in three areas. They aren't exotic — they're boring, which is exactly why they get overlooked.
1. Manual Data Work — the "Time Tax" Nobody Books
This is the most widespread and least visible leak. Someone in the company — usually the most competent person on the team — spends several hours a week copying, cleaning, and merging data from different sources just to produce one usable report.
That time never shows up in any budget as "the cost of bad data." It shows up as part of the analyst's salary, so it looks like normal work. In reality, it's work that shouldn't exist at all — it exists only because the company's systems don't speak a common language.
The scale of this problem in mid-sized companies is often larger than leadership expects — realistically, we're talking about one to two full working days per week, per person handling reporting, spent purely on "stitching" data together, not analyzing it. This isn't analytical work. It's manually patching a hole that reopens every week.
The cost doesn't end with the salary. It ends with the most valuable person on the team having no time left for what they were actually hired to do — thinking, not copy-pasting.
2. Decisions Made on Late or Inconsistent Data
The second leak is more expensive, but harder to prove — because it doesn't show up in any cost report. These are business decisions made on data that was outdated, incomplete, or simply different across two systems.
Single Source of Truth means a situation where the organization has one authoritative place where a given number lives — one version of revenue, one version of inventory, one version of customer acquisition cost. Without it, two departments can walk into the same meeting with two different numbers, each fully convinced they're right.
The consequences of this leak are very concrete:
- A pricing decision made on outdated product margin data.
- A marketing budget shifted toward a channel that "looks" effective, because the report didn't account for returns.
- Inventory ordered based on a sales forecast that didn't reflect the latest shift in demand.
Each of these decisions seems rational in isolation, at the moment it's made. The problem is that the foundation it stood on was already outdated by the time the decision was reached. This isn't human error — it's a structural feature of an organization that never invested in keeping data current and consistent across systems.
3. Knowledge Trapped in One Head and One Spreadsheet
The third leak is quiet — until the moment it becomes a crisis. The company has one person who "knows how it works." Their own spreadsheet, their own macro, their own way of merging data — and only they understand why a given number looks the way it does.
This works fine until that person takes a vacation, changes jobs, or simply gets sick during a critical month-end close. That's when the company discovers that its key reporting process was never actually a process — it was a dependency on one person.
This is especially painful in three situations: during scaling (new hires have no way to learn the process, because it's never written down anywhere, only "in someone's head"), during team turnover (one person leaving stalls reporting for weeks), and during capital transactions — because for an investor or a buyer, a critical financial process depending on one person and one spreadsheet is a red flag that directly lowers the company's valuation during Due Diligence.
What This Means for EBITDA If Nothing Changes
None of these three leaks will close on their own. They grow with company scale, because more people, more systems, and more transactions simply create more points where data drifts apart.
The consequences compound in three places on the balance sheet that leadership rarely connects to "a data problem":
- Fixed costs grow faster than revenue — because the company hires more people to do manual work a system should be doing.
- Strategic decisions get made late or on flawed assumptions — which translates directly into lost margin, misallocated budgets, and delayed reactions to the market.
- Company valuation drops in an investor's eyes — because Data Debt is one of the first risk signals due diligence is especially sensitive to.
The most expensive part of this mechanism is psychological, not financial: a leadership team that isn't confident in its own numbers makes decisions more slowly and more defensively. Fear of a wrong number is more expensive than the wrong number itself.
What to Do Differently
You don't need to fix everything at once. Three steps that actually close the leak — in the order they make sense:
- Map where someone in the company is actually manually merging data — don't ask "do we have a problem," ask "who spent more than 2 hours last week copying data between systems."
- Identify one number that has two different versions inside the company — revenue, margin, acquisition cost. That's your first candidate for a Single Source of Truth.
- Check whether any critical reporting process depends on one person — if it does, this is more urgent than it looks, regardless of whether you're planning a capital transaction or not.
This isn't a quarterly project. It's a shift in how you look at data — from "it works, so don't touch it" to "let's calculate what it's actually costing us to keep it working."
Code is a cost, a solution is an asset. Manual data work feels like "just work." In reality, it's a cost the company pays every week for never having invested once in fixing the problem at its source.