Broken link building means finding pages that link to a resource which no longer exists, and offering the publisher your working replacement. The pitch is genuinely helpful: their page has an error, you are pointing it out, and you happen to have a fix.
It appears in every tactic list and is described far more often than it is run, largely because the conversion rate is low enough to be discouraging when you first measure it.
The honest numbers
| Stage | Realistic figure |
|---|---|
| Dead resources found per category | 20–80, depending on how mature the field is |
| Referring domains per dead resource | 3–40 |
| Contactable after filtering | ~40% of those |
| Reply rate | 15–25% |
| Reply-to-link conversion | 25–40% |
| Overall attempt-to-link | 5–9% |
Roughly one link per twelve to twenty emails. Compared with cold outreach that is decent, and compared with almost anything else it is slow.
What makes it worth running anyway: the emails cost minutes, the tactic runs continuously in the background, and the resulting links are unusually durable — a publisher who fixed a broken link has actively chosen your URL rather than accepted a pitch.
The procedure
Step 1 — Find the dead resources
Three reliable sources:
- Competitors and former competitors that shut down. When a company in your category folds or is acquired and its content disappears, every link to it breaks at once. This is the richest source in B2B and almost nobody watches for it.
- Retired tools, calculators and reports. Vendors prune content constantly. An annual report from 2021 that fifty publications cited may simply be gone.
- Association and institutional pages. Trade bodies restructure sites regularly and rarely redirect properly.
Practically: run a backlink tool over the domains of defunct or restructured players in your category and filter for 404s with referring domains attached.
Step 2 — Check who linked, and filter hard
Pull the referring domains for each dead URL, then apply the same standards you would to any target: verified organic traffic, topical relevance, a named editor, no footprint clustering.
Expect to discard more than half. A dead resource cited by forty sites is frequently cited by thirty-five sites you would not want.
Step 3 — Have a genuine replacement
This is where the tactic usually fails. If the dead page was a detailed benchmark and your replacement is a product page, the publisher will not swap them — and you have spent your one legitimate contact.
The replacement must do the same job for the reader. Sometimes that means building it, which turns a cheap tactic into a moderately expensive one and is often still worth it if the dead resource had thirty referring domains.
Step 4 — Contact, factually
Short, no pressure, no flattery.
The email
Subject: broken link on [page title]
Hello [name] — the link to [resource] in your piece on [topic] is returning a 404. It looks like [site] took it down some time last year.
We published something covering the same ground: [URL]. Use it if it fits; if not, worth removing the dead one either way.
Three sentences. The final clause matters — it makes the email useful whether or not they take your link, and it is the reason this tactic does not damage relationships when it fails.
Two variants that convert better
Variant one — the acquisition sweep
When a company in your category is acquired, its content frequently gets folded into the acquirer's site with poor redirects, or removed entirely. That produces a large batch of newly broken links, all at once, all topically relevant to you.
Conversion here runs considerably higher — often 15–20% attempt-to-link — because the breakage is recent, the publisher does not yet know, and the resource being replaced is unambiguously gone rather than merely moved.
The operational requirement is watching for it. Set alerts on acquisition news in your category and run the sweep within a few weeks.
Variant two — your own broken inbound links
Not broken link building in the classical sense, and by far the highest-yield version.
Pull every 404 on your own domain that has referring domains pointing at it. These are links you already earned and then broke — usually during a migration, a URL restructure or a content prune.
Fixing them is a redirect, not an outreach campaign. In our audits this routinely recovers between ten and sixty referring domains that the company had already paid for once.
Conversion rate: effectively 100%, because no third party is involved.
Before running broken link building outward, run it inward. Most B2B sites are leaking authority through their own 404s, and recovering it takes an afternoon rather than a quarter.
When it is not worth running
- Your category is young. Few dead resources, because little has been published long enough to die.
- You have no comparable content. Building a replacement for one link is not economic; building it for a resource with thirty referring domains is.
- You need results this quarter. This is background activity, not a lever.
How to run it sustainably
The failure mode is treating it as a campaign. It works as a habit.
- Quarterly sweep of your own 404s with referring domains. Twenty minutes, highest return.
- Alert on category acquisitions and shutdowns. Run variant one when they happen.
- Batch the outward work — twenty emails a month rather than a push. Conversion is the same and it does not consume anyone's quarter.
- Log every contact. Approaching the same publisher twice with different dead links six weeks apart is how a good relationship becomes an ignored sender.
Run this way it produces a steady trickle indefinitely at very low cost — which is exactly what it is for. It is not a tactic that closes a gap; it is one that quietly stops the gap widening.