Compare two audits

Every audit run saves to the website's history, and differences between runs surface in three places: the project's Insights feed diffs the two most recent audits automatically, the deep-audit workbooks keep permanent rows so two workbooks line up row by row, and the audit history shows the grade for every run so the trend is visible at a glance.

Compare after you ship fixes — the point of a re-audit is proving the fix moved the number. Compare before a client review so the story is "what changed since last time", not a re-reading of the same report.

Compare your latest runs

To compare two audits:

1. Run a fresh audit on the website, or let a schedule → run one.
2. Open the project's Insights tab. It lists what improved, what got worse, and what appeared or disappeared between the latest two runs, alongside the report's recommendations.
3. For deep audits, open the two workbooks side by side — because rows are permanent, row three in March is row three in June.

Read the differences

• Improved: Checks that failed before and pass now — your shipped work, confirmed.
• Regressed: Checks that passed before and fail now. Treat these as the urgent list: something recently deployed broke them.
• New: Usually new pages or new checks, not sudden decay — check the crawled page count before assuming the site got worse.

Grade movement without a matching insight usually traces to category weighting. Open both reports' category tables and compare sub-scores.

Limits and usage

• Insights diffs the 2 most recent audits. To compare runs further apart, open the two reports from history directly.
• Comparisons are only as like-for-like as their crawls: a 200-page run against a 500-page run of the same site will differ partly because of scope. Keep the audit scope → constant for runs you intend to compare.

If something isn't working

Insights shows nothing. The website has fewer than two completed audits, or the second run failed — check the audit history first.

Everything looks changed. The two runs crawled meaningfully different page sets (the scope changed, or the site's sitemap changed). Compare the crawled counts in each report's health strip.

A regression you can't reproduce. Old stored reports keep the rules that existed when they ran. A check added since then shows as "new" against an old report, not as your site regressing.

Related guides

• Schedule automatic re-audits
• Read your audit report and grade
• Run the Technical SEO audit — permanent rows exist for exactly this comparison.