Check whether the outcome improved
Test representative clicks
Exercise a typical value, a boundary value, an empty value, and a value sharing a label with another group. Record parent filters and expected children for each case. Include a date boundary so an off-by-one period cannot hide behind a normal result.
Reconcile populations
For each click, compare the parent contribution with the child population or aggregate after applying the documented relationship. Check that hidden constraints, grouping changes, and labels are visible enough to explain discrepancies instead of treating every different count as failure.
Check return and export paths
Navigate back to the parent and inspect any table or export reached from the child. Confirm filters and period context remain explicit, and record any surface where the same drilldown uses a different scope or omits the clicked value.
What to carry forward
Accept the drilldown when representative clicks preserve the documented parent-to-child relationship, explain intentional narrowing, and keep context visible through return and export paths. Fail verification for any unexplained population change, then attach the smallest fixture that reproduces it.
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