Trace Cohen open-sources AI SEO toolkit behind 4.6M impressions
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Trace Cohen published an open-source "AI SEO Playbook" repo at TraceCohenTech/ai-seo-playbook containing 14 diagnostic scripts, 9 battle-tested configs, structured data schemas, and CI automation, distilling the system behind ValueAddVC.com's jump from 604K to 4.62M monthly impressions.
- The content engine uses multi-model AI agent orchestration — Opus/Fable for planning, Sonnet for writing, Haiku for grunt work — combined with voice training, a 5-format content rotation, and anti-AI fingerprint detection to pass 9 publish gates.
- The GSC feedback loop runs weekly automated reports that triage pages into title rewrites, cannibalization merges, stale-page refreshes, or kills, feeding improvements back into Google Search Console data every week.
- Growth metrics are skewed by GEO traffic: overall CTR is just 0.4% because impressions grew roughly 8x largely from AI-overview citations that don't click through, though human-intent CTR improved to 6.8% on ranked lists and 3.2% on question-led posts.
- A single-day spike on Aug 13 hit 127K impressions and 854 clicks, described in the source as part of a "near-vertical" growth curve.
- Safety and cost controls include repo locks, rebase guards, Vercel [nobuild] tags, a deploy-tick pattern, self-healing heartbeats, and isolation of content writers from git to prevent runaway build costs.
- The companion guide "The AI SEO Playbook: How I Used AI to Build a Content Engine That Hit 4.6M Impressions in 3 Months" covers the methodology behind the scripts, including the GSC reckoning, keyword anticipation, and living-page refreshes.
- The repo ships with samples for every script (weekly-report.json, rewrite-candidates.json, cannibal-clusters.json, etc.) plus a GitHub Action that runs the weekly report every Sunday and can auto-create a summary issue.
Why it matters: By releasing the actual scripts, configs, and quality gates rather than just theory, Cohen turns a documented 4.62M-impression result into a replicable system for operators willing to wire up Google Search Console API access — but the 0.4% CTR caveat is the load-bearing detail: the traffic spike is largely AI-overview citation impressions, not click-throughs, meaning anyone copying the playbook should expect visibility gains without a proportional lift in human visits.
Ask SkimNews
