Crawled but not indexed in Search Console: fix it or let it go?

Search Console keeps flagging a chunk of pages as "Crawled, currently not indexed", and people split hard on what to do about it. One camp rewrites the content and requests indexing every time. For most sites I'm in the other camp. If Google crawled the page and still chose not to index it, the page usually isn't adding anything the index needs. Thin tag archives, filter URLs and near-duplicate product variants are the typical cases, and forcing them in tends to spread crawl attention thinner across the pages that matter.

The exception is when the flagged URLs are your money pages. Then it's a quality problem, and the real test is whether the page answers the query better than what's already ranking. Where do others draw that line, especially on sites with a few thousand URLs?

Start with whether the page has one query it's meant to own. If it doesn't, it's really a tag or filter page, and the fix is to noindex or consolidate it, not rewrite it. If it does have a target, check whether indexed pages link to it internally. A page with a clear target and real internal links that sits in that status for months is losing to something that already ranks, and that's when a rewrite is worth the effort. In general, most flagged URLs fail that first test.

Kai, I don't think crawl budget is the right lever here. On a site with a few thousand URLs it's rarely what's holding pages back. Google's own documentation says it mostly matters for much larger sites, or ones that churn through pages constantly. And requesting indexing on a handful of URLs isn't pulling attention away from your money pages.

The better reason to keep thin tag and filter pages out of the index is quality. They dilute how the site reads overall, and noindexing them is the cleaner fix either way.

I don't think noindex is the cleaner fix either way, Rowan. It works for a tag page nobody links to. But a filter or variant URL that has picked up backlinks, or still gets traffic from old links, loses that value when you just hide it. A 301 or rel=canonical to the main page passes those signals along and takes the duplicate out of competition. Noindex leaves the URL live and still crawled, so it cleans up less than it looks like it does.

I'd sort the flagged set by inbound links first. No links and no real query means noindex or let it go. Links but no unique purpose means consolidate. Only the pages with a genuine target are worth a rewrite.