Scaling a consumer software knowledge base across languages (anonymised)

Background

A founder‑led consumer software publisher needed thousands of help and landing pages localised with solid SEO hygiene: stable slugs, title/description fields, canonical links and hreflang. The CMS mixed custom templates with content blocks.

Stakeholders: the founder and editorial lead, web/SEO, a localisation PM, and translators/reviewers for fr‑FR, de‑DE, it‑IT, es‑419, pt‑BR and zh‑CN. Cadence was weekly, with frequent updates to evergreen articles.

What I built

A connector that exported the right fields (including SEO metadata), protected shortcodes and HTML fragments, and returned translated content with slugs and taxonomies intact. A small checker flagged broken anchors, orphaned screenshots and duplicated titles before publish.

How we ran it together

Editors tagged batches, translators worked in tidy packages with context screenshots, and returns landed in the CMS as drafts for a quick visual skim. SEO kept hreflang and canonicals stable across languages without manual patching.

Outcome

Publishing got faster, fewer broken links reached production, and traffic stabilised because structure stayed consistent while copy improved.

What I’d keep today

The same connector shape and the pre‑publish checks. They keep editors focused on content, not URL surgery.