CASE STUDY
Daft.ai: Technical SEO and CMS Architecture
Technical SEO · Payload CMS · GitHub · Structured Data · Oct 2025 – Jan 2026
Indexing restored within 48 hours after robots.txt and GSC fixes.
Results snapshot
- Resolved robots.txt blocking across all blog URLs
- Restored accurate GSC tracking with proper domain verification
- Implemented canonical and schema infrastructure through GitHub
The challenge
Daft.ai is an AI infrastructure company building tools for distributed and multimodal workloads. A Screaming Frog crawl revealed that every blog post URL on the site was blocked by robots.txt, so none of their content was eligible to be indexed by Google. Separately, their Search Console property was configured as daft.ai rather than www.daft.ai, skewing performance data and making organic visibility impossible to track accurately.
What I achieved
Indexing & Search Console
Fixed the robots.txt configuration to allow crawling of the blog. Posts began indexing within 24–48 hours. Added www.daft.ai as a verified GSC property to establish accurate baseline tracking going forward.
Schema markup & canonical URLs
Working directly in GitHub, I implemented missing canonical URLs across blog content and added structured data for Organization, Blog, and BlogPost schema, which had been entirely absent from the site. Changes were pushed and verified through Vercel’s deployment pipeline.
CMS architecture
Inside Payload CMS, I audited the heading architecture and corrected widespread H1 duplication across blog posts. I also evaluated the related posts field, confirmed it was neither rendered nor included in structured data, and implemented it into the JSON-LD Article markup to strengthen crawl depth and entity relationships.