The technical audit, and what it found

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

Two-thirds of live link equity arrives through a redirect

Of 656,637 currently-live links, what the target URL responded with:

And 73.5% of all links point at http:// rather than https://.

The 301s are almost certainly the http-to-https migration doing its job, and 301s do pass equity. This is not a fault. But two thirds of the site's accumulated link equity arrives through a redirect, which makes that redirect configuration one of the most load-bearing things on the server — and it is worth proving rather than assuming, because a chain (http, then https, then https-plus-www, then canonical) costs a hop at every step and is completely invisible until somebody checks.

20,768 live links point at targets the crawler could not fetch

9,751 connect failures, 6,576 DNS resolution failures, 4,441 timeouts — on links flagged live, against my own domain. A further 52 are recorded as BannedDomainOrIP and 21 as blocked by robots.txt.

The most likely explanation is not that the pages are down. It is that the host or WAF is intermittently blocking Majestic's crawler.

If that is what is happening, it is worth fixing, because every link tool the industry uses to evaluate this site is being served an incomplete picture of it — including tools a potential client might run before deciding whether I know what I am doing.

Stated as a hypothesis, not a finding. It needs a server-log check to confirm, and until that is done it stays labelled as a guess. Publishing it unconfirmed is the point: this is what the middle of an audit looks like.

Three service pages lost 20,748 links between them

These are commercial service pages, not blog posts. A loss of this shape usually means the URL changed and the redirect was never put in.

Which is the most ordinary failure in this entire study, and the most instructive: the links were earned, the content still exists in some form, and the equity was lost to a housekeeping step nobody owned.

My own properties stopped linking

Among domains with no live link left at all:

Both are mine. hartzer.net was the third-largest referring domain in the entire history of this profile and now sends nothing. Whatever happened — a retirement, a redesign, a parking page — it is entirely within my control, and it is the largest recoverable loss in the dataset.

Separately, two cross-links from billhartzer.com to hartzer.com are flagged deleted: one on “buying brand mentions” last seen 3 June 2026, and one on “seo expert witness” last seen 28 April 2026. I own both ends of both links and did not notice either.

Twenty-five years in this industry and the biggest single finding in my own audit is that I stopped linking to myself. If that can happen here it can happen anywhere, which is the only reason it is worth publishing.

The reclamation pool

14,619 of 22,260 referring domains — 65.7% — have no live link left. Most are junk. Not all:

717 domains that were once good enough to link and no longer do — and, per the evidence on why links die, their pages are almost certainly still live and still maintained.

That is the highest-yield prospect list available to this site, and it required no prospecting at all. It was sitting in an export.

An unexpected asset

/wp-content/uploads/2013/09/not-provided.jpg carries 3,200 live links.

It is a chart I made in 2013 for a post about keyword “(not provided),” which the industry then embedded and credited for a decade. I had no idea.

Worth confirming it still resolves, and worth considering whether it deserves a proper page around it rather than sitting in an uploads directory. It is also a reminder that the linkable asset is frequently not the thing you thought you were publishing.

Recommendations, in priority order

Items 1, 2, 3, 5, 6 and 8 are a single afternoon between them, and they recover or protect the large majority of the exposure. Item 7 is the one that is actually worth real money and it is the one that will take months.

The distribution is typical, and it is the argument for doing an audit at all: most of what turns up is cheap to fix and was invisible until somebody looked.