Limits

WHYSE finds the segments that account for a change in observational data. It does not run experiments and it does not claim to prove causation. Proving cause needs a controlled test — WHYSE can tell you which slice to test.

A dimension that wasn't pre-aggregated isn't covered by the fast path. Asking about it triggers a scan, and scans are slow by design — the speed guarantee applies to the dimensions you chose in advance.