Every developer has shipped a noindex to production
28 October 2025/1 min read
There is a rite of passage in this job and it goes like this.
You build the site on a staging URL. Sensibly, you add a rule telling search engines to ignore it, because you do not want a half-finished site indexed. Very professional.
You launch. Everybody is pleased.
Six weeks later somebody asks why the site is not showing up for its own business name. Its own name. You could search the exact company and get their Facebook page and a Yelp listing from 2014.
It is the noindex. It went to production with everything else, because it was part of the site, because it was doing its job right up until the moment its job changed.
I have done this. Every honest person who builds sites has done this, or inherited it from somebody who did. It is not a knowledge problem. Everybody knows what noindex does. It is a checklist problem, and specifically it is the problem of a rule that was correct for months and became catastrophic at a single moment nobody marked.
The fix is a launch checklist, and the top item is not "check the noindex." It is: view source on the live homepage and read the head. Out loud, if that helps. Robots meta, canonical, title, description. Thirty seconds.
Do it the day of launch. Do it again a week later, because someone will restore a backup.
I keep the same three lines in every project's handover doc, and I still get a small flinch every time I run them.