Private blog networks

By Bill Hartzer · Hartzer Consulting · 25+ years in search, domains and expert witness work

What a private blog network is

A private blog network is a set of websites under one operator's control, presented to search engines and to buyers as independent publishers, existing for the purpose of linking to other sites. The operator either uses it to rank their own properties or sells placements on it to clients who may or may not be told what they are buying.

The word private is doing a lot of work. What makes it a network is common control; what makes it private is that the control is concealed. Remove the concealment and you have a publishing group, an ordinary and legitimate business - large media companies own dozens of titles and link between them constantly. The difference is that a publishing group does not pretend its titles are strangers.

Two construction methods dominate: the expired domain build, acquiring domains that have lapsed but still carry inbound links from their former life; and the aged-domain purchase, buying sites still running and repurposing their outbound links. Both aim at the same thing - a domain that already looks trusted, pointed wherever the operator wants.

Buyers usually encounter networks without the label. Marketplace inventory sold as guest posts, sponsored articles or niche edits is frequently network inventory. If a vendor can guarantee placement on any site in a catalogue of several hundred, within a day, at a fixed price, they are not brokering relationships with independent publishers. They own the sites.

How it works, step by step

Running a network is an operations problem disguised as an SEO tactic. Every site needs six things done differently from every other for the pretense to hold: registration, hosting, template, name, content and publishing rhythm.

1. Acquire domains with residual authority. Usually expired, filtered by an inbound link metric and a topical category, bought at auction or through a drop-catch service.
2. Rebuild a plausible site. Sometimes from archived copies of the original, sometimes with commissioned or generated content on the domain's former subject.
3. Diversify the infrastructure. Different hosts, registrars, nameservers, themes and analytics accounts, to break the technical links between properties.
4. Publish a cadence. Enough posts, frequently enough, that the site does not look like a parked shell.
5. Link outward sparingly. A single followed link to the target inside otherwise unremarkable content, on a small proportion of posts.
6. Monetize or deploy. Point the links at the operator's own sites, or sell placements.

Every one of those steps is a decision that has to be made afresh, correctly, hundreds of times. Nobody sustains that. Operators automate, automation repeats, and a footprint is simply the place where the repetition became visible.

What it costs in time and effort

A self-built network is the most expensive form of link acquisition there is, and the cost structure is routinely misunderstood as cheap.

The capital cost is the domains, and domains with real residual authority are competitively bid - the good ones are not cheap and the cheap ones are not good. Then hosting, fragmented across providers to avoid subnet clustering, forfeiting every volume discount. Then content, on subjects the operator has no interest in, for sites nobody will read, forever.

The operating cost is attention, and it is unbounded. Renewals fall due across different registrars. A platform update breaks something on nine sites at once. A site gets hacked and starts serving spam. This is a portfolio of small businesses, each producing no revenue, maintained purely so that a link on it looks like a link from somewhere.

Rented access to somebody else's network is the opposite: cheap, immediate, and correspondingly worthless, because whatever value the network has is divided among everyone else renting it. The cheapness is itself information - if placement on a supposedly authoritative site costs less than the site would earn from a week of display advertising, it does not have the audience the price implies.

When it works and when it does not

Networks work, briefly, in weak competitive fields, which is why the market persists. A site in a low-competition vertical, pointed at by twenty domains with inherited authority, can move. I have watched it happen and I have watched what came after.

What makes the tactic structurally unsound is not that detection is possible - it is that the economics require it. A network's value depends on nobody knowing it is a network; its revenue depends on selling to as many customers as possible. Every extra customer creates another graph in which its domains co-occur, and co-occurrence across unrelated profiles is the most damning evidence a network can produce. The business model destroys the product.

Then there is correlated failure. Twenty links from twenty independent publishers fail independently: one redesigns, one closes, one goes out of business. Twenty links from one network fail together. If the network is identified and devalued, everything you bought stops working on the same day - alongside the operator's other clients, who may be in verticals you would not want to be adjacent to.

It also fails on relevance. An expired domain that was once a regional newspaper, now publishing generic articles and linking to a plumbing company, is not an endorsement of a plumbing company. It is a domain wearing a costume, and a link's value comes from the judgment behind it.

Common mistakes

• Believing your network is different. Every operator I have spoken to believes theirs is the clean one. The footprints below are found on clean ones.
• Fixing one footprint at a time. Operators diversify hosting, then get caught on templates, then on the metric signature - which they cannot fix at all, because it is a consequence of how the domains were linked before they bought them.
• Assuming inherited metrics survive repurposing. A domain's authority came from links pointing at content that no longer exists. Rebuilding a different site on the same URL does not renew that endorsement, and the links that produced the metric are themselves aging out - on my own profile the median referring domain stops linking after about 1,080 days.
• Running one content pipeline. Same writer, same length, same publishing hour. The rhythm is a fingerprint even when nothing else is.
• Buying network links without knowing it. The most common mistake in the category. Marketplace inventory at guaranteed availability is network inventory. Run the checks before you buy.
• Concentrating the program in one network. If a network is your link strategy, an operator you have never met controls your rankings.

A worked example

This is the clearest case I have on file, and it came out of my own link graph rather than an investigation. Two hops out from billhartzer.com - a profile where no link has ever been bought - sits this run of domains:

Seven domains. Trust Flow spans four points, Citation Flow spans two, and the gap between them is 13 to 15 in every row. The names are nonsense five and six letter brandables of the kind a script generates in batches - and two of them, zlutag and ylutag, differ by one character.

Why is that conclusive? Flow metrics are computed from a domain's inbound links. For seven domains to land in the same narrow band, they must have been linked in the same way, at the same time, from the same places. Independent sites with different histories do not converge on the same two numbers. Identical metrics mean a shared link source, and a shared link source means one operation.

Then the finding that closes it: the same domains appear in a second, entirely unrelated site's link graph - a manufactured profile I was analyzing for other reasons. One network, two unconnected graphs, same seven names. A network sells to everybody, so it ends up in everybody's graph.

Note what this does not prove about me. I did not place those domains and cannot remove them. They sit at tier 2, linking to sites that link to me. A dirty second tier says nothing about the site at the center; it says something about how the sites in between were assembled.

How to measure it

There is no meaningful way to measure a network's performance, because the failure mode is a cliff rather than a slope. Measure exposure instead: whether network links are in your profile, and how much of your visibility depends on them.

Run the following on your own first tier, and on any prospect before money moves:

1. The metric signature scan. Sort referring domains by Trust Flow and look for runs where the Trust Flow and Citation Flow pair repeats within a few points across supposedly unrelated sites. This is the cheapest test and the strongest.
2. The naming scan. Nonsense brandables, keyword-stuffed strings, clusters of cut-price extensions registered in a batch.
3. Links per referring domain, by tier. A tier where every referring domain sends exactly one link is what buying one at a time looks like. A deep tier with several links per domain is a set of sites wired to each other.
4. Cross-profile co-occurrence. Referring domains appearing in two unrelated clients' profiles. Nothing legitimate produces that.
5. Topical coherence. A network's inventory mixes unrelated verticals, because it takes any customer.

Then quantify what share of your referring domains come from that set. That number is your correlated risk, and it is the only figure in this tactic worth reporting.

The verdict

A private blog network is the most operationally demanding thing in link building, sold on a secrecy its own business model erodes, producing an asset whose value can vanish in a single event. Self-built, it costs more than earning the links would have. Rented, it is shared with your competitors, and its metrics were manufactured by exactly the process you are paying to benefit from.

The evidence above is the part a buyer should remember. Seven domains, metrics identical to the point, nonsense names, showing up in two unrelated link graphs - found in ten minutes with a public tool by someone who was not looking for them. If that is what a network looks like from outside to me, consider what it looks like to an organization with crawl data on the whole web.

Learn to detect networks. Do not build one, and do not buy from one without knowing that is what you are doing.