Junk removal company on a headless stack: SEO through the repository

A local junk removal company had a modern website: content in Sanity, hosting on Vercel, code in GitHub. It also had every classic problem of a client-rendered site, and with no SEO plugin to configure, every fix had to go through code.

What I found

• Client-side rendering left crawlers with thin initial HTML.
• Many pages shared the same title and meta description.
• Heading levels were skipped or repeated.
• Missing pages answered with a 200 status, producing soft 404s.
• Canonical tags and structured data were missing or inconsistent.

Division of labor

The developer handled the infrastructure: prerendered HTML and deployment changes. I set the order of work, specified what each page type needed, reviewed the developer's changes, and checked the production site independently. For metadata, I made the code changes myself in a pull request that the developer reviewed and merged.

How the result was checked

• Crawls before and after prerendering, comparing extracted text.
• Raw HTML fetched from production to confirm content was present without scripts.
• Search Console's live URL test on representative pages.
• A sitewide review of titles, descriptions, canonicals, and status codes.

Why this way of working helps

On headless sites, SEO lives in templates, routes, and build settings. Working through the team's pull requests means every SEO change has a commit, a reviewer, and an easy rollback, and nobody has to wonder who changed what.

Status

Core fixes are live and verified, with a final audit under way. See JavaScript and headless SEO.

Updated October 6, 2026