Research
Why most guest post sites fail vetting: data from 94,224 sites

This report sets out what happened to every site that went through our checks in one month. It is the source for the BacklinkURLs figures quoted in our guides.
The headline number, 1.8%, needs care, and the report explains what it does and does not mean. Individual sites are never named. Every figure is a count or a share, and groups under 20 sites are left out.
Summary of findings
| Finding | Figure | Base |
|---|---|---|
| Passed every free check | 1.8% | 94,224 sites checked |
| Did not load, or redirected to another domain | 18.2% | 94,224 sites checked |
| Had gray-topic content | 25.7% | 8,778 sites scanned |
| No posts in the past 12 months | 24.0% | 1,173 sites read |
| Fewer than 5 of 10 recent posts on topic | 43.3% | 1,157 sites with enough posts |
| Failed on traffic | 63.8% | 318 sites with traffic pulled |
| Below 30% of their own traffic peak | 39.5% | 387 sites with 24-month history |
| Sellers who replied | 20.3% | 548 delivered emails |


What happened to all 94,224 sites
How to read the top rows. Our candidates come from a broad sweep of the web, filtered for home, garden, DIY and lifestyle blogs. 52.9% were turned away for being outside that niche and 21.7% for not being blogs. Those are sorting results. They do not mean three quarters of the web is poor.
The rows that generalise. 17.1% of candidate sites did not load at all, and another 1.1% sent the visitor to a different domain. Lists of sites go stale quickly.
Gray topics
Gray-topic contamination explains why one post is enough for us to remove a site and how to run the search yourself.
Posting rate and subject
A site with no posts in a year can still rank and still sell links. It is also the kind most likely to be sold on and turned into a link farm.
Traffic
This is the finding we think matters most. Sites that load, publish on topic, have no gray content and answer email still fail on traffic about two times in three. Nothing visible on the site shows it. How to read traffic history gives the method.
Seller outreach


Method
- Sample. 94,224 candidate sites gathered by searching for blogs in home, garden, DIY, renovation and lifestyle topics, plus sites seen cited in AI answers on those topics.
- Dates. Checks dated October 20 to November 20, 2026. Seller emails from October 19.
- Order. Free checks first, traffic last, as set out on the methodology page. A site leaves at its first fail.
- Tools. Direct fetches of each site's homepage, sitemap and recent posts. Third-party estimates of Google traffic and ranking keywords over 24 months. Public registration records and the public web archive.
- Counting. One row per site. Shares are of the base shown beside each figure.
Limitations
- One niche. The off-topic and not-a-blog shares describe our filter. They would differ for any other subject.
- First fail only. Later checks are undercounted in the main chart.
- Estimates. Traffic figures are modelled from search rankings and can differ from a site's own analytics.
- Word matching. The gray-topic scan can miss content that avoids the obvious terms, and can flag an innocent use of a word.
- A snapshot. One month, in late 2026. The pipeline is still running and the totals have moved since.
- Survivors. The traffic sample is small because few sites reach that stage.
Download the aggregate data (CSV)
How to cite this
BacklinkURLs Research. "Why most guest post sites fail vetting: data from 94,224 sites." BacklinkURLs, Nov 20, 2026. https://backlinkurls.com/research/why-most-guest-post-sites-fail-vetting/
Related guides
Sources
Last updated Methodology