Guides
We Checked 7,017 Backlinks We Delivered — Here's How Many Were Actually Dofollow
Every backlink seller talks about dofollow links. Almost none will tell you what share of a real delivery actually is one. Across 7,017 links we placed: 82.5%.
Backlinkcart Team
4 Aug 2026 · 5 min read
Every backlink seller talks about dofollow links. Almost none of them will tell you what share of a real delivery actually is one.
We can, because we keep a verified record of every link we place. This is what 7,017 of them look like — every link across six completed orders, checked and counted rather than estimated.
Dofollow vs nofollow backlinks: the real split
Of 7,017 delivered links:
| links | share | |
|---|---|---|
| Dofollow | 5,786 | 82.5% |
| Nofollow | 1,231 | 17.5% |
So roughly one link in six passes no direct authority. That number is not a failure and it is not padding — but you should know it before you buy, because almost nobody quotes it.
Nearly all the nofollow comes from three link types
This is the part worth understanding. The nofollow links are not scattered randomly through the delivery. They cluster almost entirely in three categories:
| Link type | Delivered | Nofollow | Share of that type |
|---|---|---|---|
| Wiki | 572 | 562 | 98.3% |
| Social Bookmark | 265 | 265 | 100% |
| Directory / Citation | 233 | 233 | 100% |
Those three types account for 86.1% of every nofollow link in the dataset.
Meanwhile the two biggest categories go almost entirely the other way:
| Link type | Delivered | Dofollow |
|---|---|---|
| Web 2.0 | 4,268 | 99.5% |
| Contextual / editorial placements | 1,524 | 96.3% |
The reason is structural, not commercial. Wikis, bookmarking sites and business directories mark outbound links nofollow by platform policy. No supplier can change that, and any seller promising you dofollow citations is either mistaken or lying about what they are placing.
That matters for how you read a bulk offer. When someone sells you "5,000 backlinks" heavy on directories and bookmarks, a large share of that number is nofollow by construction. It still has a job — brand footprint, crawl discovery, a natural-looking profile — but it is not the job people assume they are paying for.
Do bought backlinks stay live?
We re-check placements after delivery. Across the 500 links we have verified so far, 423 were still live — 84.6%.
Be careful with that number, including when you see it from us. It covers 7.1% of the corpus, and the sample leans heavily on two categories (441 of the 500 are Web 2.0 or contextual placements). It is a real measurement of the links we checked, not a projection across all 7,017.
By type, among what we verified:
| Link type | Checked | Live |
|---|---|---|
| Contextual / editorial | 310 | 81.6% |
| Web 2.0 | 131 | 87.8% |
| Wiki | 13 | 76.9% |
The honest takeaway is not "84.6% of backlinks survive." It is that a meaningful share of any link order stops working, and nobody tells you which ones unless they re-check. That is why we do, and why replacement is free for twelve months.
91% of the links came from different domains
6,389 distinct domains across 7,017 links — 91.1% unique.
That ratio matters more than most buyers realise. Fifty links from fifty domains is worth far more than fifty links from five, because authority is counted per referring domain, not per link. A pack that quietly places two hundred links across a dozen sites is a much smaller campaign than the headline number suggests.
If a seller will not tell you the domain count, ask again. It is the single fastest way to compare two offers that both claim the same number of links.
What to do with this if you're buying backlinks
Ask for the dofollow split, not the link count. Any supplier with real records can produce it in a minute. One who cannot is not tracking their own deliveries.
Expect nofollow in the foundation tier, and refuse to pay editorial prices for it. Directories, bookmarks and wikis are legitimately part of a natural profile. They are also the cheapest links in the market, and they are nofollow essentially always.
Ask how many unique domains. Then divide. The number you get is closer to the size of your campaign than the number on the invoice.
Ask whether anyone re-checks. A link that disappears three weeks after delivery was never really delivered. If the answer is no, the live rate on your order is unknown to you and to them.
Method, and its limits
The dataset is every link across six completed Backlinkcart orders — 7,017 rows, each recorded at delivery with its host, link type and follow status. Dofollow status is known for 100% of rows. Live status is known for 500.
Three limits worth stating. This is one vendor's delivery record, not a survey of the Indian market. The order mix is our mix — heavy on foundation packs, which pushes the Web 2.0 share up and would look different for a buyer purchasing only guest posts. And the live-rate sample is both small and unevenly distributed across types, as described above.
We have published the figures we can stand behind and labelled the one that is a sample. If that seems like a lot of hedging for a marketing post — that is rather the point. The numbers in this industry are usually presented with none.