blog-decay

Export page-level performance data for two comparable periods from Google Search Console as JSON, then run the content_decay.py script with both files. It flags pages with declines of 20% or more and assigns severity levels and recommended actions.

Run python3 scripts/content_decay.py current.json previous.json with two GSC page exports. You can adjust the decline threshold with --threshold, switch metrics with --metric impressions, and output a markdown report with --format markdown.

The script accepts JSON lists of page rows containing page or url, clicks, and impressions fields. It also accepts the object shape returned by the blog-google command where rows sit under a top-level rows key.

Refresh when the page still has demand and likely needs freshness, title, or internal link updates. Prune when the page had very low prior demand and does not justify rewrite effort, and consolidate or redirect when the loss is severe enough that merging into a stronger URL recovers value faster.

A page is only marked dropped_out after confirming identical filters, sufficient row limits, matching dimensions, and URL inspection between exports. Without those checks, missing pages are marked needs_validation because export differences can mimic a true drop.