What twenty years of unbought links actually looks like

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

Why this study exists

Every link building audit you have read is one of three things: anonymized, written about a competitor, or produced by somebody selling the remedy. All three constraints stop the author saying anything uncomfortable, and all three are why the industry keeps repeating claims about links that nobody has actually measured.

This study is about my own site. Nothing is anonymized, nothing is flattered, and the recommendations at the end are a list of things I need to go and fix.

It is also a rare specimen, and that is the part that makes it worth your time. No link to billhartzer.com was ever bought. Every one is natural, with the single exception of my own properties linking to it. That makes it possible to answer a question the industry can normally only speculate about: what does a link profile look like when nobody has manipulated it?

Testing any claim about link decay requires a profile that is old enough to have a history, large enough to be statistically meaningful, and clean enough that what you are watching is natural mortality rather than a purchased campaign falling over. Those three conditions rarely co-occur in a dataset anyone is willing to publish.

The dataset

Trust Flow 37 · Citation Flow 42 · Majestic Million rank 274,226. Links live 656,637, lost 645,202 — 49.6%. Follow 83.6%, nofollow 16.4%, image links 6.4%, 98.3% English-language sources.

Source: Majestic Historic and Fresh exports for billhartzer.com, downloaded 2 August 2026. The Historic export returned 1,301,839 rows against the 1,382,639 links the Summary screen reported — 94.2%. Every figure on this site is computed on what the export actually contained, not on the headline number.

The seven findings

1. A referring domain links for a median of 1,080 days — just under three years. Half the profile is already gone. That converts directly into a replacement rate that belongs in a budget.

2. Durability scales with the strength of the linking domain. A link from a Trust Flow 61+ domain survives roughly four times as long as one from a Trust Flow 0 domain — 9.2 years against 2.4. I have not found this published anywhere, and it means cheap links are cheap twice over.

3. Almost nothing is lost to a dead page. Of 645,202 lost links, 97.4% disappeared while the source page was still perfectly reachable. Four — four links out of 645,202 — were lost to a 404.

4. Concentration is far more extreme than any dashboard suggests. One hundred referring domains out of 22,260 account for 84.3% of every link ever recorded. A single sitewide link is 16.8% of the profile on its own.

5. Seven in ten referring domains are Trust Flow 0 — 15,421 of them, every one unsolicited, on a profile where nothing was ever bought. That is the calibration point for anyone who has been told their junk links are a problem.

6. In twenty years the site earned 446 referring domains at Trust Flow 41 or above — about 22 a year. Useful arithmetic to have in your head the next time somebody offers you fifty of them in a quarter.

7. The same analysis, honestly computed two ways, gives opposite answers. Weight by link instead of by domain and the Trust Flow relationship inverts completely. That one is a methodology lesson, and it is the reason to trust the rest of this.

How this study is organized

Each part below stands on its own. If you only read one, read the first — it is the part with a finding I believe is genuinely new.

The technical audit is the least generalizable section and the most useful one to read if you are about to audit a site of your own, because it is a worked example of what turns up when you actually look rather than run a tool and read the summary.