Compare two
deployments
Scan staging and production at the same moment and see exactly what is different: findings, technologies, published versions and configuration. No score and no winner, because a staging site is supposed to differ. Anything we could not compare is said out loud rather than counted as a match.
What it compares
Security findings on each side and any that differ in severity. The technologies each deployment is built with, and any explicitly published version that differs. The configuration each one serves: canonical, robots, share image, structured data, manifest and more. And a fourth section for anything one side could not be observed well enough to compare, because a difference we could not confirm and a difference we measured are not the same fact.
Questions
What can I compare?
Any two reachable deployments. Staging against production is the common one, but a preview URL against live, or the same site before and after a migration, work the same way. Both are scanned fresh at the moment you press Compare.
Why does it not tell me which one is better?
Because most differences between staging and production are deliberate. Different robots rules, a different canonical, no analytics on staging: all normal, none of it worse. Only the security rows carry a severity, because only a check that actually ran can support one. Everything else is reported as different, which is the honest word.
What does “could not compare” mean?
That one side could not be observed well enough for the comparison to be safe. If a technology was visible in the JavaScript on one deployment and the other's bundles could not be read, saying it is missing there would be a guess dressed as a finding. Those go in their own section instead, named so you can see how much of the comparison is bounded.
Is this safe to run against production?
It is the same read-only scan the free checker runs, on both URLs at once. Nothing is installed, no code is executed, and no data is written. Point it at deployments you own.
Does it store either site?
No. Both scans run for the request and the comparison is returned to you. Nothing about either deployment is kept.