Reporting Trust Glossary
What Is a Logic Leak?
Why it matters
A logic leak turns reporting into memory work.
The official dashboard may look clean, but the real business rule might live in a spreadsheet adjustment, a manual exclusion, a copied formula, or a caveat that only one person remembers. When that logic is not visible, the number becomes hard to explain and harder to trust.
The leak is not always technical. It is often operational context that never made it back into the reporting chain.
What it looks like in a growing business
Logic leaks often appear around exceptions.
Common signs include:
- A spreadsheet adjusts values before a meeting.
- A dashboard excludes records without explaining why.
- Finance applies a month-end correction outside the BI layer.
- One person knows which customers, orders, or dates should be excluded.
- Reports match only after someone applies a manual step.
The issue is not that judgement exists. The issue is that the judgement is hidden.
How to spot it
When a report disagrees with another report, ask:
- What manual steps happen before the number is used?
- Are there exclusions or adjustments outside the main reporting path?
- Which caveats do people explain verbally but not document?
- Does a spreadsheet contain business rules missing from the dashboard?
- Would the number still be explainable if one person was unavailable?
If the answer depends on undocumented memory, you may have a logic leak.
What to do next
Trace one disputed number from source to decision.
Write down where manual judgement enters, which rules are legitimate, and which steps should be documented, automated, or exposed as caveats. Keep the first pass small; the goal is to make hidden logic visible.
For related inspection points, read source-to-report lineage explained. For the broader framework, get the free opening chapter.