Why finance and BI report different numbers

BI dashboards and finance reports usually disagree for one of four reasons. They define the metric differently, cut off the period at different times, pull from different source systems, or apply different logic such as filters, currency conversion or adjustments. Finding which gap applies is faster than rebuilding either report.

Every finance team has sat in a meeting where two revenue numbers were on the screen and the conversation turned into which one was right. Often both were right. They were answering slightly different questions.

The four gaps

1. Definition. One report counts booked revenue, the other recognized revenue. One is gross, the other net of refunds or credits.
2. Timing. BI refreshes from source systems daily. Finance closes the period and posts adjustments afterward. Same month, different moment.
3. Source. BI reads the CRM, billing system or core. Finance reads the general ledger after journal entries and accruals.
4. Logic. Filters, joins, currency conversion and intercompany eliminations get written one way in SQL and another way in a spreadsheet.

How to tell which gap you have

Start with the size and shape of the difference. A constant gap points to definition. A gap that closes after month-end points to timing. Each common symptom below points to a likely gap and a first thing to check.

• The gap is the same every month. Likely a definition gap. First, compare the written definitions: gross vs net, booked vs recognized.
• The gap shrinks or disappears after close. Likely a timing gap. First, compare refresh times and cutoff dates.
• The gap appears only for some products or regions. Likely a source gap. First, check which system each report reads for those segments.
• The gap changes after someone edits a report. Likely a logic gap. First, compare filters, joins and currency handling.

Why fixing reports one at a time doesn't hold

You can reconcile two reports today. Next month a new dashboard appears with its own logic, and the cycle repeats. The fix belongs upstream, in a shared definition that every report reads from, so the question stops being which number is right.

How Preql helps

Preql maps each metric's definition across your existing systems. Business owners approve it, and every change is versioned, so BI and finance can work from the same agreed definition.

FAQ

Which number is right?

The one that matches the agreed definition for the question being asked. If there isn't an agreed definition, that's the real problem.

Should BI tie to the GL to the penny?

For financial reporting figures, yes, after close. Operational dashboards can differ during the month as long as the difference is explained.

Who should own the definition?

Finance should own what a financial metric means. Data teams own how it gets computed.