The challenge
Technically the website was in good shape, but it got little out of its reach: many impressions, few clicks, almost only searches for the brand name. For AI search, the basics that systems like ChatGPT need in order to cite a source were missing.
The audit found concrete causes: not a single FAQ section, incorrect company data in the schema, a self-description that differed on every channel, and a strategically important page that had never been indexed by Google because of a canonical error.
My approach
Check first, then act precisely: I analysed the website via crawl, the Webflow API and Search Console, agreed the measures with the team and built them directly into Webflow without a redesign.
- GEO audit with live crawl, full Webflow audit and Search Console analysis
- Ten FAQs on four key pages with FAQPage markup, text and markup word for word
- Canonical error fixed, the AI page is in Google's index for the first time
- Company data in the schema corrected and merged into one entity
- One entity text for website, schema and Google Business Profile
- New titles and snippets for the articles with the most reach
- Heading structure, llms.txt and an update playbook for the team
Results
Modulr now describes itself the same way everywhere, and AI systems find the answers in machine-readable form. The AI page is indexed by Google for the first time, and the revised articles get more clicks from the same reach. With the update playbook the team keeps the measures up to date itself.