Score history and changing criteria
Each captured assessment records evidence, assessment time and the active criteria used. Historical scores retain their original identity and meaning when Flowpane adopts new criteria.
Compare like with like
Compare numeric scores under the same criteria. A Site at 100 today may receive a lower score under newly adopted criteria even if the website is unchanged. Criteria changes, rule dependencies and capability changes are distinguished from Site regressions; they do not manufacture a comparable score-drop alert.
A real public change, source/public inconsistency or genuinely unhealthy connector can cause a comparable score fall. Monitoring frequency and healthy inactivity do not reduce quality scores. A quiet Site can retain assurance when current heartbeat, source generation and measured public consistency remain valid.
Reports retain captured results. A Site trend needs multiple authoritative assessments under the same criteria. Unavailable does not imply that a Site has remained unchanged, and missing historical criteria does not establish comparability.
Finding lifetime
Findings preserve their first observation and original criteria alongside current assessment identity. Current, resolved, not applicable and criteria retired describe different outcomes. Resolution requires positive current evidence where the rule remains active. Reopening preserves earlier history.
Retiring an optional surface can remove its applicability where no explicit obligation remains, while retaining its earlier history. A criterion becoming not applicable or retired is not credited as a customer fix.
Verified impact over time
Flowpane Impact separates verified Flowpane-assisted actions from external resolutions. Recovered points require observed before and after scores under the same criteria. Longer periods prioritise verified fix counts; recovered points remain limited to current criteria and are separated when Sites use different criteria.
Proposed scores are estimates. Editing, Save and downloads do not create observed recovery. After authorised publication, inspect the resulting public verification.