Does Google count unindexed backlinks

What is documented

Google describes its process in three stages: crawling, indexing and serving. Links are part of how Google finds pages; its guide says most new pages are discovered "when Google extracts a link from a known page to a new page." That sentence establishes one fact clearly: links are read when the page containing them is crawled.

Google does not publish how it weights links from pages it chose not to index. Anything stated beyond this is inference, and should be read that way.

Three situations

Never crawled

This case is not controversial. If Googlebot has never fetched the page, the link is unknown to Google. A large share of cheap links sit here, on new or deep pages with no path from anywhere Google visits. Getting such a page crawled is exactly what a backlink indexer is for, and the mechanism is in how backlink indexers work.

Crawled, not indexed

This is the contested case. Search Console's page indexing report shows "Crawled - currently not indexed" for pages Google fetched and did not store. Google has therefore read the HTML at least once.

The following is inference rather than documented fact. A page Google declined to store is one it judged low value or redundant, and the reasoning shared widely among SEO practitioners is that its outbound links inherit that judgement. Google may still use such a link for discovery of the target URL, which is a different thing from crediting it as an endorsement. The practical consensus is to assume these links count for little and to treat indexation as the goal.

Indexed

Once the page is in the index, its links are part of the data Google's systems work with. Whether a given link then helps rankings depends on many things the index status says nothing about: the relevance and reputation of the page, the anchor, whether the link carries nofollow, sponsored or ugc, and whether Google treats it as an attempt at manipulation. Google's spam policies class links created "primarily for the purpose of manipulating search rankings" as link spam.

Why the question matters commercially

Link building is bought per placement. If a share of placements sits on pages that never enter the index, the effective cost per useful link rises. Measuring the share of indexed linking pages is therefore one of the few objective quality checks a buyer has. The methods are in verifying backlink indexation, and reasons for failure in why backlinks don't get indexed.

A reasonable policy

Practitioners who cannot resolve the theoretical question tend to follow a simple policy:

1. Check every new linking page for indexation after a sensible wait; how long backlink indexing takes gives typical windows.
2. Submit clean, unindexed pages for crawling.
3. Re-check, then separate pages that need a fix from the site owner from pages that are simply not going to be indexed.
4. Count only indexed placements when judging a vendor or a campaign.

This policy avoids depending on an unknown. It treats an indexed linking page as the condition for a link to have a chance, without claiming more.

Relationship to crawling and indexing

The underlying distinction is between being fetched and being stored. The crawling vs indexing entry explains it from Google's documentation. An indexer influences the first; the second remains Google's decision. IndexChex, publisher of this wiki, frames its own service the same way: it guarantees Googlebot crawls and measures indexation as an outcome, which is why it pairs submission with a bulk index checker.

Common misconceptions

• "Google counts every link it can find." Google crawls and indexes selectively, and has said it does not guarantee either.
• "An indexer makes a link count." It makes a crawl happen. The page still has to earn its place in the index, and the link still has to be one Google values.
• "If the target page ranks, the links must be indexed." Rankings have many causes; only a direct check shows which linking pages are indexed.