Reporting Trust Glossary
What Is Tolerable Divergence?
Why it matters
Not every mismatch is a failure.
Sales, finance, operations, and leadership may need different versions of a number because they are making different decisions. A live operational dashboard and a closed finance pack should not always match. A forecast and an invoice report may both be valid while answering different questions.
Tolerable divergence helps teams stop forcing false alignment and start naming the reason numbers differ.
What it looks like in a growing business
Tolerable divergence usually appears around timing, purpose, and authority.
Examples include:
- Daily revenue differs from month-end recognised revenue.
- Sales pipeline differs from finance-approved forecast.
- Operational customer counts differ from contracted customer counts.
- A live dashboard differs from a closed board pack.
- A local spreadsheet includes judgement that is not yet in the official report.
The difference is tolerable only when it is named, explained, and safe for the decision.
How to spot it
When two numbers disagree, ask:
- Are these reports meant to answer the same question?
- Do they use different timing rules or cut-offs?
- Is one report final while the other is still moving?
- Which decision does each number support?
- Has the acceptable difference been explained to users?
If the difference is expected and visible, it may be tolerable. If it surprises people, it is still a trust problem.
What to do next
Define which differences are acceptable and which are not.
For priority metrics, write the timing rule, decision context, authoritative report, and caveats into the metric definition. Make the distinction visible near the report, not only in a separate document.
For a practical example, read finance vs sales numbers don’t match. For the broader framework, get the free opening chapter.