Crawled but not indexed in Search Console: improve the page or wait it out?

Search Console keeps showing a pile of "Crawled, currently not indexed" URLs on a small service site, and I see two camps. One says Google is telling you the page isn't worth storing, so merge or noindex the thin or overlapping ones and leave it at that. The other says give it time, request indexing, and the page gets picked up once internal links and the sitemap improve.

I lean toward the first camp, at least as a starting point. On a site with a few hundred URLs, crawl budget is rarely the bottleneck, so this status usually points at the page itself. Google fetched it and decided not to keep it. I'd check for overlap and thin content first, then decide what to do. Where do you draw the line between a page worth improving and one worth cutting?

I agree with the diagnosis, but I'd draw the line on intent rather than word count. A 300-word page is worth keeping if nothing else on the site answers its query, and a 1,500-word page is dead weight if a sibling already ranks for the same thing. So before cutting anything, search your own site for the target query and see which URL Google actually shows. If two pages compete for one intent, 301 the weaker into the stronger, which also keeps any links they earned. If the page is the only real answer for a service you sell, noindexing it just removes that service from search. It needs work, and the fix is usually a clearer, more specific page, not a delete.

Are you mostly seeing pages that duplicate something already indexed, or unique pages that just don't seem to matter to Google?

Agreed on intent over word count. Before you decide anything, look at which URL Google is actually showing in the SERP for your target query, because it's often not the page you expected. If the "wrong" URL is the one getting indexed and ranking, the better fix may be consolidating signals onto it rather than trying to make the page you wanted win.

I'd push back on consolidating onto whatever URL Google happens to show. If that page is ranking because it's thin or aimed at the wrong intent, you're just keeping the weak page on top. Pick the URL that actually matches the query, merge the rest into it, then rewrite it.

Marco's right that the URL Google shows isn't automatically the one to keep, but which page you keep depends on how many links and impressions each one has earned. If the ranking URL has real backlinks, those are worth more than a cleaner page on a new URL, so the redirect target should be whichever one holds the signals.

Agreed that the links should decide the redirect target, and the quickest way to sort it out is URL Inspection in Search Console. For each thin page, look at the "Google-selected canonical" line. If Google already names a different URL, it has grouped the two as duplicates, so a 301 is low risk. If the page names itself as canonical and still isn't indexed, the problem is more likely content or intent, and redirecting it into a sibling won't fix that.

After that, check the Links report for external links, since that's where the signal you're trying to keep actually shows up. A thin page with real backlinks is a reason to rewrite it in place rather than delete it, because the links are the part that's hard to replace.

I'm not sure a 301 is low risk just because Google picks another canonical. Google sometimes groups pages that answer slightly different questions, and once you redirect, the second query has nothing left to rank for. So compare both pages against the actual queries they rank for before you redirect anything.

Agreed, and I'd push it a bit further. The Google-selected canonical line only tells you Google grouped the URLs, not that they serve the same intent. A page can sit in a cluster and still pick up impressions for queries the winner never ranks for. So filter the Performance report by each URL and compare queries over the last three months. If the weaker page has real impressions on terms the target doesn't cover, a 301 would quietly throw those away, and in that case it makes more sense to rewrite the page to own that angle.