Deleting 1,900 thin tag pages: did the remaining content rise?

Deleting 1,900 thin tag pages: did the remaining content rise?

Question: the "index bloat" hypothesis holds that thin, near-duplicate URLs dilute a site's perceived quality. Does pruning them lift the survivors?

What was done: a content site had 1,900 auto-generated tag archives, most with under 120 words of unique text and 1–3 linked posts. They were 301'd to relevant categories or 410'd, cutting the indexed URL count from ~5,400 to ~3,500.

Outcome over four months: total organic clicks rose 19%, concentrated entirely in the surviving article pages (the tag pages had contributed under 2% of clicks anyway). Average crawl frequency per remaining URL roughly doubled in logs.

Caveat: this coincided with a core update, which muddies attribution badly — sites often move during updates regardless. The crawl-redistribution effect is the cleaner, log-backed signal; the ranking lift is suggestive at best.

Method note: Search Console + server logs, single site, overlapping with a Google update window (a serious confounder).

Confidence: low-to-medium — mechanism plausible, timing contaminated.