Check whether the outcome improved
Exercise the comparison contract
Render matched current and baseline fixtures with known filters, units, aggregation, and coverage. Compare both values before checking the displayed difference. Include an intentional benchmark difference and confirm its context is explicit, rather than forcing every report into a previous-period comparison that may not be the reader’s intended task.
Check context changes and missing references
Change a relevant filter or measure definition and observe how each side responds under the documented rule. Exercise a missing or zero baseline where meaningful. Confirm the report avoids presenting an ordinary percentage change when the reference cannot support that calculation or when its context no longer matches the declared comparison.
Inspect shared and exported context
Open a saved or shared view and a supported export using the same fixture. Confirm the baseline definition and displayed result remain understandable outside the original editing session. Check whether later changes to a mutable benchmark alter an existing comparison in a way the report fails to reveal or explain to its readers.
What to carry forward
Accept when population selection, value calculation, change arithmetic, and context explanation agree with the declared comparison. Record unavailable saved or export paths separately. Keep verification open if only the difference formula was checked; mathematically correct arithmetic does not establish that the current and baseline values answer comparable questions.
Technical background: Metabase documentation.
Keep the decision with the work.
Use a Work Item in Aglet to record the problem, the evidence you have, and the next decision. Add an owner and priority, then keep updates in the discussion so the next person can follow the reasoning.
Create an account See the product workflow