docs-seolisted

# docs-seo — SEO Analysis for Documentation An SEO audit that reads a page and guesses what it *ought* to rank for is worth very little. Google already knows what each page ranks for, at what position, and how many people saw it and did not click. **Establish that first.** Then every title and description judgement below is made against a real query at a real position, and the report becomes a ranked work order instead of a list of opinions. ## Before you start Read `metrics/metric-dictionary.json` for the metrics in `metadata.measures`. Two things there will otherwise make you confidently wrong: - **Traffic is context, never a verdict.** Crawlers can be the overwhelming majority of raw pageviews, and behavioural metrics exclude bots while pageview counts do not. Use traffic to weight the fix queue, not to decide whether a page is healthy. - **On-site search CTR and Google's click-through are different numbers.** The first is readers using the search box inside your docs; the second is the search engine's result page. Do not quote one as the other — the fixes differ. ## Workflow 1. **Establish the real positions — before reading a single page.** Get the actual search performance for the docs: average position, impressions, clicks, and the queries each page already ranks for. What you need out of this step is a table of *page → the query it ranks for → position → impressions → clicks*, and specifically the **"worth improving" band — positions 5-2