Measurement, risk and recovery

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

Measurement is where most link programs go wrong

By this stage in the course you can find prospects, write to them, and run a campaign long enough for it to produce something. This module is about the layer that decides whether anyone believes it: what you count, how you report it, how you tell a genuinely damaged profile from a merely messy one, and what you do in the rare case where something has actually gone wrong.

I spend a lot of my professional time reading link profiles for people who are frightened of them, often in litigation or after a traffic drop. The single most consistent finding is that the profile is fine and the measurement is not. Numbers get quoted that describe the index rather than the site. Junk gets treated as a symptom when it is the default condition of publishing on the internet. Disavow files get uploaded that remove real links to solve an imaginary problem.

What this module covers

• What to measure and what to ignore — why total backlinks is the least useful number in link building, and the short list that is worth tracking.
• Building a report that survives scrutiny — distinct versus duplicate links, stating your source and date, and showing loss as well as gain.
• Spotting an unnatural profile — separating the things that look alarming and are normal from the patterns that genuinely indicate manipulation.
• Manual actions and disavow — the narrow situation where a disavow file is the right tool, and why most of them do harm.
• Running an ongoing program — link reclamation as the highest-yield recurring work, and the monitoring method that actually finds lost links.

The calibration to carry through the module

One number should sit behind everything you read here. On a twenty-year profile of 22,260 referring domains where not one link was ever bought, 69.3% of the referring domains are Trust Flow 0. That is 15,421 domains a toxicity tool would flag, on a profile that has never done anything at all.

Junk is not a symptom. It is what happens when you publish on the internet for a long time: scrapers copy you, aggregators list you, abandoned directories index you, and low-quality sites reproduce your material. If your profile looks like that, it looks like a normal profile. The lessons in this module are largely about not panicking, and about knowing precisely what a real problem looks like so that you can recognize the rare occasion when you have one.

The lessons

1. 8.1What to measure and what to ignoreWhy you count referring domains, and the short list of metrics worth tracking.
2. 8.2Building a link report that survives scrutinyDistinct versus duplicate links, and the definitions a report needs to hold up.
3. 8.3Spotting an unnatural profileWhat manipulation really looks like, and what merely looks alarming.
4. 8.4Manual actions and disavowThe narrow case where disavow applies, and why most disavow files do harm.
5. 8.5Running an ongoing programLink reclamation as the highest-yield ongoing work, and how to monitor properly.