Add a data quality check that explains a revenue swing

Finance says revenue is wrong. No idea why.

IntermediateRead 4 minDoing: a few hours6 steps

What you'll have at the end

A three-part check, wired into the pipeline, that flags a real revenue swing, ranks which segment carries most of it, and names the exact rows behind that segment's drop when they fail a plain revenue integrity rule.

Before you start

You need

A fact table holding revenue at the order or transaction level, with at least one column you could segment by, such as product category or region, plus the ability to schedule a query against your warehouse.

Not covered

Diagnosing a total that came out too high instead of too low; a doubled number usually traces to a different mechanism, a join producing duplicate rows.

Leans on

6 steps

Where to go next

Fix a warehouse total that's doubled

the number went up instead of down, usually a join duplicating rows, a different mechanism than a value silently zeroing out.

Checked 25 Aug 2026

Part of the Data Warehousing cookbook