Diagnose why your pages rank but don't convert clicks
You ship content. You see impressions in Search Console. The clicks don't follow. The metadata layer is broken — but generic SEO advice doesn't tell you which titles to fix, in what order, or how to prove the lift.
This is a complete 6-phase agent playbook. Paste it into any frontier AI assistant and it will run the audit end-to-end as your senior SEO partner: lock a baseline, diagnose the pattern, score against an explicit rubric, rewrite the worst offenders, plan a staged deployment, and quantify the lift at Day 7, Day 14, and Day 30.
Free .md download + weekly newsletter for technologists. Unsubscribe anytime.
Already have it? View your other resources ->
6 sequential phases. One agent conversation.
The order matters. You cannot prove lift without a baseline. You cannot rewrite without a diagnosis. You cannot deploy without a measurement plan. Here are the first three phases:
- •11-question structured interview about your site, audience, and constraints
- •Pull top-10 pages by impressions, clicks, CTR, and average position from GSC
- •Site-wide aggregates locked verbatim — the proof-of-lift reference for Day 30
- •Topic vs. Outcome, Specificity Signals, Value Cues, Boilerplate Metas
- •Trust Signals, Action Verbs, Length Compliance — each scored PASS / PARTIAL / FAIL
- •Diagnostic Summary with the 1–3 highest-leverage changes ranked by CTR impact
- •5-point Title Rubric: keyword position, specificity, value cue, length, no anti-patterns
- •5-point Meta Rubric: deliverable named, trust signal, action verb, length, no AI tics
- •Prioritization Table tagging each page HIGH / MED / LOW — focus the rewrites
Phases 4 (Rewrites), 5 (Deployment Plan), and 6 (Day 7 / 14 / 30 Measurement) are inside the FREE downloadable .md file. Enter your email above to unlock them, then paste the file into your favourite agent and run the audit.
Need the broader system this audit fits inside? The SEO Strategy Playbook covers technical health, structured data, programmatic SEO, and AI discoverability across 9 phases.
Not a checklist. A diagnostic conversation that ends in measured lift.
Baseline-locked from minute one
Most title rewrites fail silently because no one captured the “before” state. This workflow refuses to advance past Phase 1 until impressions, clicks, CTR, and average position are locked verbatim. Phase 6 quotes them back unchanged at Day 30.
Explicit rubric, not opinion
Every rewrite is scored against a 10-point rubric — keyword position, specificity signal, value cue, trust language, action verb, length compliance, no AI-content tics. A senior practitioner does not flatter; they sharpen. The agent will tell you when a rewrite still reads like a file name.
Staged deploy with revert gates
Rewrites ship in batches sorted by expected lift. If Batch 1 shows > 25% REGRESS at Day 7, the agent blocks Batch 2 deploy. If a page over-promises and CTR rises but bounce spikes, the workflow recommends REVERT within 48 hours. You do not ship-and-pray.
One Markdown file. Six phases. A senior SEO partner on demand.
The total focused work is 60–90 minutes spread across the 30-day window. You end with rewritten metadata that ships, a deploy date tracking table, and a Day 30 Final Assessment Report that quantifies the lift.
A complete agent playbook in one .md file
Paste it into Claude, ChatGPT, Gemini, or Cursor. The agent runs the audit end-to-end as your senior SEO partner, asking one question at a time and respecting phase order.
A locked Baseline Snapshot mechanism
The single most important artifact: page-level and site-level metrics frozen at audit start. Without it, Phase 6 cannot prove lift. The brief enforces capture before any rewriting begins.
A 10-point combined rubric
Five points for the title (keyword, specificity, value cue, length, anti-patterns), five for the meta (keyword, deliverable, trust signal, action verb, length). Score every page, prioritize the rewrites that matter.
Day 7 / 14 / 30 measurement framework
WIN / WATCH / REGRESS scoring with explicit gating logic between batches. If Batch 1 has > 25% REGRESS, the agent tells you not to ship Batch 2 — diagnose first. Quantified lift, not vibes.
Three steps, instant download
No account. No course to complete first. Download the file, open Google Search Console alongside your AI assistant, and run the audit phase by phase.
Enter your email
You get instant access to the Title & Meta Audit Workflow Markdown file. You are also subscribed to a weekly newsletter for technologists — unsubscribe anytime.
Download the .md file
A single Markdown file designed to be pasted into a fresh AI session. The agent reads the entire brief and runs the 6 phases with you over the 30-day window.
Run the audit phase by phase
Phase 1 takes 15–20 minutes (interview + baseline capture). Phase 4 (rewrites) takes 20–30 minutes. Phase 6 runs three times, 10–15 minutes each, at Day 7, 14, and 30 post-deploy.
Built for technologists running content sites who can read a Search Console report
Scenario 1: “My pages rank but nobody clicks”
Classic CTR problem. Run Phase 1 to lock the baseline, then Phase 2 will surface the pattern — usually topic-titles, boilerplate metas, or missing specificity signals. The rewrite output in Phase 4 fixes it page by page.
Scenario 2: “I have 50+ pages and no idea where to start”
Phase 3 produces the Prioritization Table. It tags each page HIGH / MED / LOW based on combined rubric score and monthly impressions. You stop guessing and rewrite the pages with the highest expected click lift first.
Scenario 3: “I rewrote my titles last quarter and nothing changed”
Likely missing the measurement framework. Without a Baseline Snapshot and Day 30 comparison, you cannot tell whether the rewrites worked, whether the search engine A/B tested old-vs-new, or whether you regressed silently. Phase 6 closes that loop.
Prerequisite: you have at least 14 days of impressions and clicks data in Google Search Console (or equivalent). Without a baseline, the audit cannot prove lift — the agent will redirect you to install GSC first and come back when data has accumulated.
Stop ship-and-pray rewrites
Six phases. A locked baseline. A 10-point rubric. WIN / WATCH / REGRESS scoring at Day 7, 14, and 30.
Free .md download + weekly newsletter for technologists. Unsubscribe anytime.
Already have it? View your other resources ->