Anatomy of a 130,000-Page YMYL Recovery

Large-site recovery starts with segmentation

A 130,000-page health platform cannot be diagnosed through a handful of example URLs. Recovery work begins by segmenting performance, crawl behavior, indexation, templates, directories, intent, and change dates so the loss can be located instead of averaged away.

The recovery sequence

1. Overlay traffic and visibility changes with algorithm, release, hosting, and template events.
2. Inventory URLs by directory, template, status, canonical target, clicks, impressions, and last crawl.
3. Separate valuable historical URLs from duplicates, thin variants, and dead inventory.
4. Repair taxonomy and consolidation rules before expanding content.
5. Test changes on bounded cohorts, then monitor crawl, indexation, queries, and unintended loss.

A decision table before redirects

url,history,intent,target,action,verification
/old-question,valuable,duplicate,/topic-hub,301,target indexed
/thin-variant,none,redundant,,410,removed from index
/core-guide,growing,unique,,keep,queries stable

Protect the site while simplifying it

Pruning is not a volume target. Each merge, redirect, removal, and template rule needs a reason, a destination where relevant, and a post-release check. In YMYL, editorial trust and medical accuracy remain separate workstreams from technical consolidation.