Devs kept saying 'works on my machine.' I needed proof Googlebot saw something different. This is the screenshot-able workflow that ended the argument in one meeting.
— Search Console → URL Inspection → paste URL → Test Live URL.
— Open 'View Tested Page' → screenshot tab. This is what Googlebot rendered. Half the page was blank.
— 'More Info' → 'Page resources' → I found 3 JS files marked 'Other error' — a CDN with a Googlebot-hostile firewall rule.
— 'View Tested Page' → HTML tab: copy it, diff against the browser DOM.
Turns out a bot-protection rule on their asset CDN was 403-ing Googlebot's fetch of the very bundle that rendered the page.
The fix list: allowlist Googlebot's verified ranges on the asset host, re-test, confirm resources load.
Result: blocked resources went 3 → 0, the render screenshot filled in, and pages stuck in 'Crawled - not indexed' (2,400 of them) cleared to indexed in 5 weeks.
Hydrate Diaries
@HydrateDiaries
The exact Search Console workflow to prove a JS bug to your dev team
This post was published in the Telegram channel Hydrate Diaries. Subscribe here: @HydrateDiaries.