The platform that became its own case study
How we built a recommendation engine that learns what each visitor needs — and why we published the playbook.
Context
Wellness & AI publishes daily case studies, blog posts, resources, and done-for-you bundles across multiple domains and audiences. Every visitor lands somewhere different. The content library grew past the point where a static menu or category grid could surface the right next read. Practitioners missed individual resources. Individuals missed practitioner bundles that solved the same problem from a different angle.
The shift
Instead of hiring a recommendation vendor or bolting on a third-party widget, we built the engine ourselves using the same 3-Layer methodology we teach. Research: mapped the signal taxonomy. Ledger: created a lightweight signal table. Protocol: wired a scoring algorithm that weights affinity, penalises dismissed items, and caps type diversity.
Approach (in shape, not in recipe)
The system tracks browse signals (weight 1), purchases (weight 5), and dismissals (weight -3). It builds a per-session affinity profile across domains, audiences, and layers. For logged-in users the profile persists. For anonymous visitors it still works with session-scoped signals. A daily cron aggregates impressions, clicks, and dismissals into analytics and emails the admin team.
What an honest observer would notice
Within the first day: the Suggested for you strip appeared on every major page. Each card has a dismiss button that teaches the algorithm. Admin analytics show domain affinity heatmaps, most-dismissed content, and click-through rates. The engine surfaces practitioner bundles to visitors who read practitioner case studies, and individual sleep resources to visitors who read sleep-related blog posts.
Done-for-you
Want this shipped to you?
Recommended next