The short answer: because a report is a snapshot, not live monitoring. It documents the state at the moment of scanning - and deliberately doesn't change, even when reality already has.
Why the report doesn't "see" your fix
If the report updated itself in the background, it would lose its most important property: it would become an unverifiable moving target. A snapshot gives you three things a live view can't:
What to do: rescan
After implementing the fixes, run a rescan (the button in the report or next to the project on the home screen). A new audit is created: fixed checkpoints change status, the Quality score updates, and the old report stays in the history as a reference point.
Three traps worth knowing about
1. 30-day data reacts with a delay.
Checkpoints analyzing a 30-day window (e.g. event continuity, channel conversions) may keep showing the problem for a while after the fix - because the analysis window is still dominated by pre-fix data. Configuration checkpoints (settings, attribution, integrations) react immediately; behavioral ones - gradually. If you fixed something "data-related", give the window time and rescan again after 1-2 weeks.
2. The fix may not have worked.
Not malice - the most common scenario in analytics. A change deployed to staging but not production; a tag fixed in GTM but the container unpublished; a fix working on desktop but not mobile. If the rescan still shows the error, check the evidence on the checkpoint - it points to what the automation still sees.
3. Checking off in the Action plan ≠ verification.
The plan's checkbox is your declaration for organizing the work. Only a repeat audit changes statuses in the report.
Related articles: