I used to think my QA process was solid. I had a checklist. I ran it before every launch. I marked things off. Most of the time, nothing slipped through.

Then one day something did.

A landing page went live with a broken CTA button on mobile. Not a hard bug — just a padding issue that made the button impossible to tap on smaller screens. We'd checked the desktop version. We hadn't checked mobile. The checklist item was there: "Test on mobile." I'd looked at it. I'd assumed someone else had done it.

That's the problem with a checklist that lives in a document. You can read an item, assume it's covered, and move on. There's no real accountability — just a row in a spreadsheet that could mean anything.

After that I started thinking about what a QA process actually needs to be reliable. Not just a list of things to check, but a record of what was actually verified, by whom, on which version of the page, on which date.

The fix wasn't complicated. I needed a checklist that produced a run record — not a document I could edit after the fact, but a timestamped snapshot of what passed and what didn't.

I built that into CheckRun. The extension runs in the Chrome side panel, next to the page you're reviewing. You go through each item marking it Pass, Fail, or N/A. When you're done, you save the run. It's stored with the URL, the date, and the full item-level results. You can't go back and quietly change a Pass to a Fail — the record is the record.

For landing page QA specifically, the pattern I use now:

Open the page in the browser. Open the landing page checklist template in the side panel. Work through it — above the fold, CTA visibility, form functionality, mobile layout, load time, trust signals, analytics tracking. Mark each item. Save the run.

If something fails, it's logged. If I hand the page off to someone else, they can see the previous run and know exactly what was checked and what wasn't. No ambiguity.

The broken CTA on mobile would have been caught. The checklist item would have had to be marked Pass by someone who actually tested it on mobile — not assumed covered because it appeared on a list.

That's the difference between a document and a run. Documents are edited. Runs are records.