How to Spot a Private Blog Network
Before You Buy.

What a Network Actually Is

A private blog network is a set of websites under one operator, built to sell links rather than to be read. The domains are usually bought expired, so they arrive with a history of real links pointing at them. The operator puts a theme on each, fills it with articles that exist only so a paid link has somewhere to sit, and sells placements on the strength of authority scores the domains earned under previous owners.

Google's spam policies name this pattern directly, under link spam: buying or selling links for ranking purposes, and "creating links that weren't editorially placed or vouched for". A link from a network is exactly that. The risk is not theoretical — a manual action against the buyer removes rankings the links were bought to earn, and cleaning it up costs more than the placements did.

The difficulty is that a well-run network looks, at a glance, like a set of small independent blogs. Each has a name, a logo, a theme, categories, an about page. The tells are structural, not cosmetic, and they show up in the places an operator maintains by script rather than by hand. What follows is where to look.

The Sitemap That Names Its Siblings

The single most conclusive check takes thirty seconds and needs nothing but a URL bar. Fetch /robots.txt on the domain and read the Sitemap: lines. The Robots Exclusion Protocol lets a site declare where its sitemaps live, and an independent site declares one or two, both on its own hostname.

The pattern to look for is a robots.txt that declares sitemaps on other hostnames: dozens of them, for domains that have nothing to do with each other on the surface, all written into one file. That is what a shared WordPress installation does when one plugin generates the robots file for every site it manages and lists every sitemap it knows about.

An operator who writes a robots.txt by hand would never publish a list of every domain in the network. One who lets a plugin do it publishes it on every site, every day, without ever reading the output. When you find a robots.txt naming sitemaps on foreign hostnames, you have found the network and, usually, most of its members. Note the domains — and note that a listing on any of them is a listing on all of them.

The same file tells you something quieter. If it names a sitemap that returns a 404, nobody has looked at the site's own configuration in a while. Our ingest treats a dead declared sitemap as a reason to look elsewhere rather than to give up — the editorial guidelines describe how — but on a site you are about to pay for, it is worth knowing that the operator has not.

Authority That Clusters

Independently built sites have authority scores scattered across the range, because they were built at different times by different people for different reasons. A network's domains were bought to a specification — a floor on Domain Authority, a ceiling on price — and their scores cluster around that specification.

Clustering on its own does not prove shared ownership. Plenty of unrelated sites land on the same score, and any marketplace sorted by DA will group them. What it tells you is that a set of sites may have been selected on that number, which is worth knowing before you read anything else about them.

The more telling version is a cluster on one metric and a scatter on another. Moz DA is the score sellers price against, so it is the one a network buys for. Ahrefs DR is computed from a different index and is harder to move by the same purchase. When a group of sites sits in a tight band on DA and spreads across forty points on DR, the DA is what was bought and the DR is what is true. The difference between the two scores is worth a post of its own; here the point is simply that agreement between them is a mark of an ordinary site, and a systematic gap in one direction across a whole set is a mark of a purchased one.

The registration date is a third reading, and anyone can look it up. Network domains are usually bought expired, and a name that lapses, is deleted and is registered again starts over with a new creation date in its RDAP record, while the links other sites point at it stay where they were. A domain registered last year that already outscores ones registered a decade ago is carrying its previous owner's links, links bought since, or links earned unusually fast.

Links Per Referring Domain

Divide a site's total backlinks by the number of domains those links come from. A natural profile sits low — most sites link to you once, a few link two or three times — so the ratio lands somewhere between one and a handful. A site with thousands of links from a few hundred domains has been linked to by the same small group over and over, which is what sitewide footer links and reciprocal network links look like in aggregate.

Ahrefs explains its own Domain Rating as a function of the linking domains and their strength, not the raw link count, which is why a high ratio inflates backlink totals without moving DR much. If a seller leads with a backlink count and goes quiet on referring domains, do the division yourself.

The Shape of the Traffic Line

Open the domain in Semrush, or whichever traffic tool you use, and look at the two-year organic traffic chart before you look at the number. A real audience grows the way audiences grow: unevenly, over months, with a slope. A network domain is typically flat at zero for the entire history, then steps vertically upward in the last few weeks — a square-edged block that begins the day the operator started pushing traffic at it and ends the day they stop.

Two shapes are worth naming. A single spike that returns to zero is an event, not an audience — a viral post, a bought burst, a bot — and the current figure is whatever the spike left behind. A plateau that appears from nowhere and holds is bought traffic that has not been switched off yet; the tell is the vertical left edge, since nothing organic goes from nothing to its full level in a week.

The number itself can be tested without opening a tool. Semrush builds its organic estimate from the rankings it tracks, so divide the estimate by the count of keywords behind it: the listings here carry both figures, and the guide sets out how to read one against the other. An estimate that rests on a handful of terms can be moved by a handful of terms, which is exactly what an operator aims at when the point is to make a domain look visited. A seller who quotes the visits and not the keywords has left out the half that tests them.

The extreme case is an estimate that rests on a single ranking keyword. Whatever the visit count, it is one lost ranking away from zero, and none of it will reach an article placed somewhere else on the site.

Twins in the Template

Networks are built by script, and scripts leave fingerprints of their own. Two sites that share a theme, a category structure, a post count per category, a headline formula and a footer layout were built by the same process, whatever their about pages say. The check is to open two suspected members side by side and count.

The headline formula is the quickest of these to spot. Scripted content tends to reuse a handful of openings ("Beyond the…", "Navigating the…", "Decoding the…") across every site it fills, and a category the domain name promises is sometimes missing altogether.

Three smaller checks in the same family. Look at where the images are served from — a shared image CDN across supposedly unrelated sites is a shared owner. Look for the theme's demo content still present on the homepage; an operator managing forty sites does not notice "Welcome to Catch Magazine" on one of them. And look at the authors: a network's posts either have no byline or have the same invented one on every site.

Why One Signal Is Never Enough

Every signal above has an innocent explanation. Small independent blogs run on the same popular theme. A site can genuinely spike after a feature. A hobbyist can buy a Moz score by accident, by linking from a large forum they run. A new owner can register a lapsed name in good faith and inherit links nobody bought. A charity's robots.txt can name a partner's sitemap for a reason that made sense to someone. If you delist a site on one fingerprint, you will be wrong often.

A network shows several at once, because they all come from the same cause. Shared infrastructure produces the robots.txt tell and the template tell, shared image CDN included. Buying expired domains to a specification produces the authority cluster and the DA-versus-DR gap and the young registration under an old score. Linking the members to one another produces the heavy link ratio. Dressing the domains up to look visited produces the traffic tell: a vertical edge on the chart, and a visit count resting on a handful of keywords. When three or more of the seven line up on one domain, the innocent explanations have to be true simultaneously, and they are not.

This is also why this directory publishes the raw numbers instead of a verdict. A listing that carries both authority scores, the registration date, the Semrush traffic estimate beside its keyword count and the backlink total beside the referring domains holds the figures for four of the seven on one page, and the full table, sorted by DA, shows a fifth. What it does not do is decide for you — that is the editorial position in the about page, and it is the reason the numbers are there.

The Twenty-Minute Method

For any site you are about to pay for, in this order. Fetch /robots.txt and read every Sitemap: line; a foreign hostname ends the exercise. Fetch the sitemap it names and count the URLs; a "publication" with forty pages is a placement page. Open the domain in whichever traffic tool you have and look at the shape of the chart before the number, then divide the number by the keywords behind it. Note both authority scores, the gap between them and the year the domain was registered. Divide backlinks by referring domains. Open the homepage and two articles and read them as an editor would: is there a subject, a voice, a reason these pages exist? Then open a second site the same seller offers and hold the two side by side.

If you are choosing from the listings here, the sitemap count, the traffic against its keywords, the two scores with the registration date and the link ratio are already on the entity pages, taken on 29 Sep 2026 and dated. The robots file, the chart, the reading and the second site take a browser tab each. Twenty minutes on a site you are about to spend money on is not diligence; it is the minimum.