ONE WORD, THREE MEANINGS
Every number is correct. They’re just not measuring the same thing.
Three leadership team members pull up their dashboards ahead of a Monday review. Sales reports 340 active customers. Finance reports 290. Support reports 410. Nobody’s system is broken. Nobody’s data is dirty in any technical sense. Each number is entirely accurate, according to its own definition — and that’s exactly the problem, because all three definitions are quietly different, and nobody has ever put them side by side.
Sales counts anyone who’s purchased in the last ninety days. Finance counts anyone with an open invoice. Support counts anyone who’s logged into the product in the last week. Ask any one of these teams whether their number is correct and they’ll say yes, confidently, and they’ll be right — about their own definition, which they’ve never had reason to compare against anyone else’s.
Why this is more common than it should be
Definitions get set locally, by whichever team needs the metric first, for whatever purpose that team has in mind. Sales defines “active” around recent purchase because that’s what matters for a sales conversation. Finance defines it around billing status because that’s what matters for revenue recognition. Neither team is wrong to define it the way they have — the definition genuinely serves their own function well.
The problem only appears the moment the term escapes its original context and starts being used as if it means one universal thing across the business — in a board pack, in a company-wide OKR, in a conversation between two departments who’ve never actually compared notes on what the word means to each of them.
Why this looks like a data quality problem and isn’t one
When numbers don’t reconcile, the instinctive response is to blame the data — assume something’s out of sync, a system integration is broken, or someone’s report is stale. Teams spend real time reconciling numbers that were never supposed to match in the first place, because they’re measuring genuinely different things that happen to share a label.
This is a much harder problem to diagnose than actual dirty data, because there’s no error to find. Every number is correct. The issue is entirely upstream, in the fact that three different, reasonable definitions were allowed to coexist under one shared word, without anyone ever explicitly deciding whether that was acceptable or a problem.
What this costs beyond the confusion
Beyond the wasted time in reconciliation meetings, the deeper cost is a slow erosion of trust in reporting generally. Once a leadership team has been burned once or twice by numbers that “don’t match” for reasons nobody could initially explain, they start treating dashboards with a degree of scepticism that undermines the whole point of having them. Decisions start reverting to gut feel or whichever number happens to support the direction someone already wanted to go, because no number can be fully trusted without a caveat attached.
What actually fixes this
Build a single glossary of the terms that get used across departments, not within them. This doesn’t need to cover every metric in the business — just the handful of terms, like “active customer,” “qualified lead,” or “at-risk account,” that get used in cross-functional conversations, board reporting, or company-wide goals. For those specific terms, one definition, agreed explicitly across the teams that use it.
Make the definition visible wherever the metric appears. A number on a dashboard with no stated definition invites everyone to assume their own. A small, consistent label — “active = purchased within 90 days” — attached to the metric itself removes the ambiguity at the point people actually see the number, rather than relying on everyone remembering a glossary document they read once.
Accept that different teams may genuinely need different internal definitions, and separate those from the shared one explicitly. Sales can keep its own working definition of “active” for its own internal purposes. The fix isn’t forcing every team to use one definition for everything — it’s making sure that whenever the term crosses a departmental boundary, into a shared report or conversation, everyone’s using the same one, and it’s clearly labelled as such.
Revisit the glossary when a new cross-functional report gets built, not just once. New shared metrics get introduced constantly as the business grows, and each one is a fresh opportunity for the same silent fragmentation to happen again, unless checking for definitional alignment becomes a standard step in building any report more than one team will see.
The reframe worth making
Clean data and aligned definitions are two entirely different problems, and most businesses only have systems in place to catch the first one. The businesses that avoid confusing, low-trust reporting aren’t the ones with the most sophisticated data infrastructure — they’re the ones who did the unglamorous work of agreeing what their shared words actually mean, before those words showed up in a room where three different answers looked like a crisis.
If two teams in your business would give you different numbers for the same term without either being wrong, that’s a definitions gap worth closing before it costs you a decision. See how our Technology & Data practice works →
