We compared our own crawl to a competitor's report, in public.

A native 500-page crawl of gplcoffee.dev against a 531-page WebSite Auditor export from 2026-05-18. Every non-zero delta gets one line explaining it, not a claim that everything matched.

500 / 500

pages fetched, 0 failed

531

page WebSite Auditor export, 2026-05-18

1,376

issues in that export

33 / 33

categories, parser exact

The verdict

The parser was exact on all five summary totals and all 33 categories. Of those 33, 14 matched outright, including all four structural-health checks - robots, sitemap, www, and http/https - plus custom 404 handling. Every remaining category had a non-zero delta, and every one of those traces to exactly one of three causes.

The 500-page cap

A handful of categories differ only because their evidence lives in the part of the site the run did not reach. Empty titles and duplicate-title groups both fall entirely outside the crawled 500 - the counts are floors, not misses.

Genuine site drift since May

The export is from 2026-05-18. Legacy links started 404ing after a domain migration, brand icons picked up alt text, and product titles grew version strings - the site kept moving, and the native crawl caught up to it.

A crawler gap, found and fixed during the pass

Robots-disallowed resources were not checked before this pass, so restricted-from-indexing coverage undercounted WebSite Auditor. Fixed in the same pass that found it, and re-verified.

Four categories, read the way the app reads them

A sample of the full delta table - the same evidence-first format the app itself uses for a finding, not a rounded-off summary.

CategoryWSANativeCauseExplanation
Resources with 5xx03Site driftOne genuine 503 and two domains that went dead since May - real drift. (The raw count was 12 until this same pass also recalibrated away 9 false TIMEOUT matches, which are our checker being throttled, not the site failing.)
Empty alt text1994Site drift188 of the WSA count were brand icons with no alt text in May; they carry it today. The remaining 4 are genuine, one of them correctly classed decorative.
Restricted from indexing190478Gap, fixedResource checks did not consult robots.txt before this pass. Fixed: internal resources are now robots-gated the same way pages are.
Duplicate titles / metas7 / 70 / 4500-page capThe duplicate groups live in the part of the site outside the 500-page cap. A larger run would resolve the rest.

Bring your own WebSite Auditor export - the Site Audit (details) report, not the summary - and we reconcile it against a live crawl of the same site, category by category.

Import your own export

Run the same crawler on your own site

A free scan reads 25 pages with the same catalog and hands back 3 fixes with the prompt to act on them. What the free scan does

25 pages, depth 2, robots.txt honoured. Most scans finish in under a minute, and the report link stays live for 14 days.