20-docs-nobody-could-find

The docs nobody could find.

I spent weeks on bilingual documentation for three open-source SDKs. Clean structure, code examples, recipes, glossary. Hosted free on GitHub Pages, custom domain, everything green.

Then I opened Google Search Console.

Barely indexed. Some pages showed up under "discovered, not indexed". Others said "alternate page with proper canonical tag" — when there was no alternate. Search queries for terms that should match the docs returned competitor results, irrelevant Stack Overflow threads, anything but my pages.

Docs that don't rank are a private repo with a domain name. If a developer Googles the problem my library solves and never finds my docs, my library doesn't exist for them.

So I moved the docs off GitHub Pages onto AWS S3 + CloudFront. The reasons were specific, not vague:

→ Control of the sitemap.xml served and linked from robots.txt.
→ Control over response headers — particularly Cache-Control, X-Robots-Tag, and the language alternates.
→ A separate origin per language subpath, so the canonical and hreflang chains were unambiguous.
→ Faster TTFB and stable serving, which Google rewards.

Within four weeks, the indexed page count tripled. Search Console started showing impressions for the terms I was writing for.

Distribution is part of the product. Especially for open source. Especially solo. Especially when you compete against incumbents whose names already own the search namespace.

This closes the 9-part open-source series. The blog continues — at tech.bilouro.com — with longer pieces on the same craft, evergreen, indexable.

Where do your docs live? And have you actually checked Search Console?

Sites: eupago.bilouro.com · vendus.bilouro.com · xmldiffreport.bilouro.com

P.S. Thanks for reading the series. Longer pieces continue on the blog.

#OpenSource #Python #DeveloperExperience