Author: Vladislav Bobrov, founder of Tier1.shop. In SEO and webmastering since 2009.
Imagine: you bought 100 links. The report is beautiful, all links are "in place", the money has been debited. The question that almost no one asks: how many of these 100 pages does the search engine even see? According to independent data from 2024–2026, which we will analyze below, — a significant portion of purchased links do not exist for Google and Yandex at all: the pages containing them did not make it into the index or have already dropped out of it. In this article — exact figures of two waves of link budget losses and a step-by-step instruction on how to check your own links in one evening.
Why a link outside the index is worth almost nothing
For a link to work, the donor page must pass through three stages: the robot has discovered it → crawled it → included it in the index. Only after this does the link enter the link graph and begin to transfer weight and relevance.
Let's be scrupulously honest: John Mueller from Google mentioned that in separate cases, the system may consider a link from a page known to it but not indexed. However, in practice, this changes nothing: it is impossible to verify such "shadow accounting", and non-indexing in the vast majority of cases is a verdict on the quality of the page (you will see the figure below). The working rule is simple: not in the index — consider that the link does not exist.
In 2026, the blow is double. All AI responses — ChatGPT, AI Overviews, Yandex Neuro — are built on top of classic search indexes: an unindexed page does not even enter the raw material for a neural response. We analyzed this in detail in the article about links in the era of AI search: indexing there is called the "entry ticket" — both to the regular search results and to neural network answers.
And a third point that link sellers are silent about: indexing is not a lifelong status. A page can drop out of the index in a month, a quarter, a year — along with your link. How many of them drop out has also been calculated.
First wave: how many pages do not make it into the index
The most extensive public study of the issue is the analysis by the IndexCheckr service on a sample of 16 million pages. The results are sobering:
- Indexed — 37,08% pages. 61.94% did not enter the Google index at all (another ~1% — entire domains).
- Average indexing time — 27.4 days. In the first week, only 14% of pages are indexed; the bulk (50.86%) — on day 8–30.
- «Indexing acceleration services», which are often used to finish off blasts: of the 33 930 pages sent to them, 29,37% were indexed. Seven out of ten rubles spent on «accelerators» yielded nothing.
Why Google is so picky — it doesn't hide it itself. John Mueller's position: «just because we crawled a page doesn't mean we will index it» — Google consciously limits the index with a quality threshold. One hundred percent indexing does not exist as a phenomenon, and Google considers this the norm, not a glitch.
The reason was named by the Indexing Insight study (2025, 1.7 million pages): 88% of non-indexing cases are related to content quality, and not technical errors. That is, a typical non-indexed page is not a «victim of a glitch», but a page that the algorithm evaluated and filtered out.
Add the crawling budget to this. According to Botify, for unoptimized sites, Google crawls only about 40% of significant URLs per month, and on large sites, robots do not visit more than half of the pages at all. A deep internal page of a weak site — the very one where a cheap purchased link is usually placed — can wait for months for a robot that will never reach it.
Second wave: pages enter the index — and die
Let's say you're lucky: the page with your link has been indexed. This is not the end, but the beginning of the second countdown.
According to the same IndexCheckr study, 21.29% of indexed pages later drop out of the index, with about 13.7% doing so in the first three months after indexing. The link "was there and vanished," yet in the seller's report, it still appears as working.
Next is the slow death of the pages themselves, known as link rot. Here are two fundamental studies:
- Ahrefs (a sample of over 2 million domains): over 9 years, 66.5% of links died. The decay rate is approximately 1.3% of the link mass per week; about half is lost over 7 years. The main reason (47.7% of cases) is that the page with the link was simply deleted or the site closed.
- Pew Research Center (2024): 38% of web pages that existed in 2013 had disappeared by 2023; even among pages from a year ago, 8% are already unavailable. For scale: 54% of Wikipedia articles have dead links in their source lists.
Now for the arithmetic of the headline. Let's be conservative: some purchased links are not indexed upon entry, every fifth of the indexed ones later drops out of the index, and simultaneously the profile melts by ~1.3% per week due to page deletions. For a link profile older than a couple of years, the estimate that "up to half of the budget is not working" is not a scare tactic, but the lower bound, which follows directly from these studies. Ahrefs formulates the conclusion even more harshly: for a five-year-old site with 100 backlinks, likely only 30–50 of them are passing value.
Which purchased links are most often not indexed
Anti-rating — by placement types, with reasons:
- Crowd, forums, profiles. Deep pages without internal weight; profiles are often closed with noindex; moderation cleans out messages. The bot may never reach the page — remember the 40% crawling rate at Botify.
- Blast-offs and "pyramids". Generated pages on trash sites — exactly the kind of content that results in 88% indexing refusals based on quality. These are the ones people try to "revive" with accelerators with a success rate of 29%.
- Cheap article sites. An article on the third or fourth level of site nesting, which a robot visits once a quarter. Even if they make it into the index, such pages are the first candidates for de-indexing among those very 21.29%.
- Global rental on internal pages. Pages spammed with outgoing links lose crawling priority and gradually drop out of the index — along with all rented links immediately.
The common denominator of all four types is internal pages of weak sites. They have neither incoming link mass nor internal weight to enter the index and stay in it. For the home page, the situation is mirrored — all the site's weight flows into it, so it is indexed first and stays in the index the longest. Why it works this way, we analyzed in detail in the article about links from home pages.
How to check your purchased links: an audit in one evening
Important framework: the donor page is someone else's; you cannot add it to your own Search Console. Therefore, we check using indirect but reliable methods.
Step 1. Collect a list. Export the URLs of all pages where your purchased links are placed — from seller reports, emails, tables. Specifically, the URLs of the donor pages, not your acceptors.
Step 2. Check indexing in Google.
- The operator
site:URL-страницы— is a fast way, but with a caveat: Google saves resources on operator queries and sometimes shows "nothing found" for pages that are actually indexed. An empty result is a reason to double-check, not a verdict. - More reliable is searching for a unique phrase in quotes: copy a characteristic sentence (not a template one) from the donor page and search for it in quotes. A page in the search results means indexing is confirmed.
- Cross out the operator
cache:from old guides — Google disabled it in 2024; it no longer exists for verification.
Step 3. Check in Yandex. The operator url:адрес-страницы provides a definitive answer. Plus, check Yandex.Webmaster → "Links" → "External links": there you can see which donors Yandex has actually credited to your site.
Step 4. Indirect verification via Search Console. The "Links" report in GSC shows the domains from which Google sees links to you. The appearance of a donor in the report is a good sign; its absence is worrying, but not final (the report updates with a lag and does not show everything).
Step 5. Mass verification. When there are hundreds of links, manual operators do not scale — use batch index checkers. Additional insurance: set up regular monitoring, because (to repeat) 21% of those indexed later drop out of the index, and 1.3% of links die weekly.
Step 6. Calculate the real price of the link. The final line of the audit, for which everything was started: paid for 100 links, 55 are in the index? This means each working link cost you almost twice the stated price. This is the figure that should be compared between providers — not the price list.
What to do with non-indexed links
Give fresh ones time. The peak of indexing occurs on day 8–30 (half of everything that will ever be indexed is indexed then). It is too early to panic on the third day.
Older than 60–90 days outside the index — almost a sentence. If a page has not entered the index in three months, the probability that it will enter later is small. This is a legal basis to demand a replacement or refund from the seller.
About "indexing accelerators" — soberly. Their measured effectiveness is 29.37%. The reason is simple: 88% of non-indexing is due to page quality, and no ping can cure that. A separate story is the IndexNow protocol (Bing and Yandex): it really speeds up the inclusion of pages in these indexes, and this has become more important with the growth of ChatGPT, which takes sources from Bing. But keep in mind: Google does not support IndexNow — services promising "instant indexing in Google via IndexNow" are selling something that doesn't exist.
Three questions for the seller before payment: where physically the link will be placed (home page or internal page?); does he guarantee indexing and within what timeframe; what happens if the page drops out of the index. From the answers, it is immediately clear who sells working links and who sells lines in a report.
Re-audit — once a quarter. Link profile — not a purchase, but an asset with depreciation: 1.3% per week. A quarterly check following the steps above takes one evening and shows the real state of what you paid for.
How this issue is solved with us
At Tier1.shop we place links only on the home pages of donor sites — and the home page is almost always in the index: it is the page most visited by robots and the most stable page of any site, receiving all the internal and external weight of the domain. Therefore, we can afford something that a crowd or blast seller cannot: a guarantee of indexing for every placement — the donor's home page enters the index on average within 1–14 days, and if it drops out of the index — we replace it for free. Plus, you receive a report (URL + screenshot) for every placement, so the audit steps from this article are directly applicable to our links — check for yourself.
Frequently Asked Questions
Does a link count if the donor page is not in the index?
Practically no. Formally, Google allows the counting of links from known but non-indexed pages, however, it is impossible to verify this, and 88% of non-indexing is a verdict on the quality of the page. Working rule: no index — no link. For AI answers, the rule is even stricter: only indexed sources get in.
How long to wait for link indexing?
Based on data from 16 million pages — on average 27.4 days; half are indexed on day 8–30. Not indexed within 60–90 days — demand a replacement. Tier1.shop placements from home pages are indexed within 1–14 days (source: Tier1.shop placement statistics), and this is a guarantee, not a forecast.
site: shows "not found", but the page is in the index — does this happen?
Yes, regularly: Google responds to operator queries on a residual basis. Double-check by searching for a unique phrase from the page in quotes — this is more reliable.
Do indexing acceleration services work?
Measured success rate — 29.37%. The reason for non-indexing in 88% of cases is the quality of the page, which an "accelerator" does not change. The exception is IndexNow for Bing and Yandex (actually speeds it up), but Google does not support this protocol.
The link was in the index and disappeared — is this normal?
This is a statistical norm of the web: 21.29% of indexed pages are later de-indexed, and ~1.3% of links per week die physically (pages are deleted). This is exactly why a quarterly re-audit is needed and a seller who is responsible for indexing after placement, not just at the moment of the report.
How to check indexing in Yandex?
The operator url:адрес-страницы in Yandex search gives a definitive answer. For the overall picture, look in Yandex.Webmaster: "Links" section → "External links".
Briefly about the main points
The link budget leaks in two waves: at the entry (62% of web pages do not enter the Google index at all, and 88% of rejections are due to quality) and over time (21% of indexed content is deindexed, 1.3% of links die weekly). Checking your links takes an evening: site: and a phrase in quotes for Google, url: for Yandex, GSC and Webmaster reports, batch checkers — and a recalculation of the real price of a working link. To prevent the problem from recurring, buy placements where indexing is guaranteed by design, — on the main pages.
Questions — to the support bot @tier1shopsupport_bot, we will answer.
