Weekly striking distance SEO briefs in Linear

Every Tuesday at 8am in my local timezone, run a striking distance SEO analysis and file content briefs as Linear issues. This is an agent-based workflow because the clustering and brief drafting is judgment work, not a fixed pipeline.

Step 1. Pull the last 28 days of search performance from Google Search Console using Search Analytics Query, grouped by both query and page. Use the most recent complete 28 day window. Search Console data lags 2 to 3 days, so end the window a few days before today.

Step 2. Filter the rows to where average position is between 8 and 20 and impressions are at least 200. These are the striking distance opportunities.

Step 3. Cluster the surviving queries by topical intent. Group each cluster under the best matching landing page, which is the page on my site that already collects most of the impressions for that cluster.

Step 4. Score each cluster by impressions multiplied by the gap between expected CTR for position 5 (use roughly 14 percent) and the cluster's current actual CTR. This is the impact heuristic.

Step 5. Pick the top 5 clusters by impact score.

Step 6. For each top cluster, call Inspect URL on the target page in Google Search Console. If the page is not in an indexed status, skip that cluster and substitute the next best one.

Step 7. For each surviving cluster, draft a concise content brief in Markdown with these sections: target query cluster (primary query plus 3 to 8 supporting queries); target page URL and its current average position; why it is stuck (pick the most likely reason among intent mismatch, thin content, missing sections, or weak title and meta); and 2 to 3 concrete optimization moves covering on-page edits, internal link sources from specific other pages on my site, and FAQ or schema additions.

Step 8. For each brief, call Create Issue in Linear in my SEO team. Title is the primary query for the cluster. Description is the full brief in Markdown. Priority is set by impact score: highest impact cluster = Urgent, next two = High, remaining = Normal.

If no clusters survive the filters and the indexing check, file a single Linear issue summarizing what was scanned and why nothing qualified this week, so the cadence stays visible to the team.