The noindex that the developer swore he removed

The noindex that the developer swore he removed

A fintech blog published 60 high-effort articles and not one got indexed. The team checked the source: no noindex anywhere. They escalated, blamed Google, waited.

The data was brutal — 0 of 60 indexed after a month. But the raw HTML really was clean. So where was the tag?

The reveal came from comparing raw response to rendered DOM. A global meta component set noindex as a safe default, then a client-side effect was supposed to switch it to index once the article loaded. The catch: Google often takes the meta robots value from the rendered snapshot, and on these pages the switch ran after a slow data call — sometimes after the renderer had already read 'noindex.'

Turns out 'default to noindex, fix it in JavaScript' is a coin flip you lose at scale. The first value Google sees can stick.

The fix: default the meta robots to index in the server response and only inject noindex server-side for pages that genuinely need it. Never let indexability depend on a race.

Result: 0 indexed became 58 of 60 within three weeks, organic sessions climbed from 200 to 9,400/month.