Broken Link Building: How to Qualify an Opportunity Before You Pitch

Broken link building is a link acquisition tactic: you find a page that links to a URL that no longer works, you confirm you have already published something that does the same job the dead page did, and you tell the publisher about the defect while suggesting yours as the replacement. The tactic itself is sound. The prospect lists are the problem, because a link checker reports HTTP responses and cannot tell you whether a failing response is an opportunity. This guide is written for founders, growth leads and SEO leads at B2B software companies running outreach with a small team and no link budget, and it works through the qualification step first and the outreach second.

The short version

• A dead URL is not automatically an opportunity. It is one only when the target genuinely returns a 404, the linking page is indexed, and the link would pass signal.
• A URL that redirects is not broken. Google's documentation says a redirecting URL will not be indexed, and there is nothing on the publisher's page for them to fix.
• A soft 404 returns a 200 status code with an error page sitting on it, so a link checker reading status codes alone records it as working.
• Relevance cannot be judged from a dead URL's slug. Read the archived version of the page before deciding whether yours is a replacement or merely a page on the same subject.
• Qualify first. Every prospect you remove before writing an email is the cheapest work in the campaign.

What is broken link building, and how is it different from the skyscraper technique?

Broken link building replaces a dead page. The skyscraper technique replaces a live one, and that single difference changes every step that follows.

In broken link building the publisher has a defect on their page: a link that resolves to nothing. You are reporting a problem and offering a fix in the same message. In the skyscraper play the publisher's page works exactly as intended, and you are asking them to prefer your version of a resource over one they already chose and linked to deliberately. Those are different conversations, and they fail for different reasons.

The two get filed together as outreach tactics because the outreach step looks similar from the outside. Our guide to the skyscraper technique covers whether that second play is worth a month of a small team's time. This piece stays on the dead-page half.

Why is a broken link not automatically an opportunity?

Because a link checker reports HTTP responses, not opportunities. Three situations that show up on a typical broken-link report are not prospects at all, and each has a specific check that catches it before it costs you an email.

False positive 1. The target redirects rather than 404s: the URL in the link is old, the destination moved, and the server now sends visitors somewhere else. Google's Search Console documentation on page indexing statuses describes this case plainly: "This is a non-canonical URL that redirects to another page. As such, this URL will not be indexed." Nothing on the publisher's page is broken for a reader, because a click still lands somewhere. There is no defect to report and no replacement to offer, and a pitch that claims otherwise tells the publisher you never checked.

False positive 2. The target returns 200 with error content: this is a soft 404, which Google's documentation defines as a page where "The page request returns what we think is a soft 404 response. This means that it returns a user-friendly "not found" message but not a 404 HTTP response code." Google's recommended fix for the site that owns it is explicit: "We recommend returning a 404 response code for truly "not found" pages and adding more information on the page to let us know that it is not a soft 404." This one can be a real opportunity, but it is a different message. The publisher's link goes to a page that loads, so nothing in their own tooling has told them anything is wrong.

False positive 3. The linking page is worthless to you: the target really is dead and the link really is broken, but the page carrying it is not in Google's index, or the link carries rel="nofollow" or rel="ugc". A publisher fixing a broken link swaps the destination and keeps the attribute, so a nofollowed dead link becomes a nofollowed live link. Our guide to link attributes like nofollow and ugc covers what each one signals. A fix here is still a favor to the publisher; it is just not a link acquisition.

From my experience auditing these lists, redirects and soft 404s turn up on them routinely, and both look identical to a genuine 404 in a report until you check the response yourself.

What the report shows	What it actually is	What to do
404 or 410 on the target, indexed linking page, followed link	A genuine prospect	Qualify relevance, then pitch
Any 3xx on the target	A redirect, not a broken link	Drop it
200 on the target, error content on the page	A soft 404	Pitch, but lead with the diagnosis
Nofollowed or ugc link, or an unindexed linking page	Broken, and worth nothing to you	Drop it unless you want the referral traffic

How do you run the three checks before you pitch?

Manually, on every prospect, in about two minutes each. The checks are not sophisticated. They are easy to skip, because a report that already says "broken" reads as though the work is done.

Step 1. Read the response code, not the report: request the target URL yourself and look at the status line rather than the tool's label. A 404 or 410 is a genuine dead page. Any 3xx is a redirect and disqualifies the prospect immediately. A 200 means you are not finished, which is what step 2 is for.

Step 2. Open every 200 in a browser: a page that returns 200 and shows a "page not found" message is the soft 404 case, and it only becomes visible when a person looks at it. This is the single check most worth automating last and doing first, because the tool cannot see it by definition.

Step 3. Check the linking page, not just the link: view the source of the page carrying the link, find the <a> tag, and read its rel attribute. Then confirm the page is actually indexed. You cannot run URL Inspection against a site you do not own, so search Google for a distinctive full sentence from the page in quotation marks. If the page comes back, it is indexed. If nothing comes back, treat it as unindexed: a link from it is unlikely to be a discovery path for anything.

Before any email: the target returns 404 or 410, or returns 200 with error content you have seen with your own eyes. The linking page appears in Google for an exact-phrase search of its own text. The link is not rel="nofollow" or rel="ugc". And you have read the archived version of the dead page, which gets its own section below, because it is the check that decides relevance rather than eligibility.

How do you find dead links worth pitching?

From four kinds of page, all of which share one property: they carry outbound links that nobody has a standing reason to revisit. The SERP already covers this layer well, so keep the sourcing step short and spend the time you save on qualification.

• Resource pages: curated link lists in your subject area, which link out heavily by design.
• Competitor backlink profiles: pages that once linked to a competitor URL that has since been removed or restructured.
• Shut-down tools and products: when a tool dies, every listicle and tutorial that recommended it is left carrying a dead link.
• Deprecated documentation: vendors reorganize docs, and tutorials that deep-linked into the old structure break in bulk.

The last two are worth more than they look, because a single shutdown produces a cluster of broken links that all share the same replacement need.

How do you check what the dead page actually said?

Read it in the archive before you decide anything about relevance. Open the dead URL in the Internet Archive's Wayback Machine, find the last capture from before the page broke, and read the page as the publisher's reader would have found it.

Four things are worth reading closely:

• What the page was: a tutorial, a dataset, a free calculator, a template, a directory, a research write-up. This is the job it did, and it is not recoverable from the slug.
• How deep it went: a 400-word definition and a 4,000-word reference are not substitutes for each other even when their titles match.
• What format it was in: a downloadable file, an interactive tool and an article are different offers to the same reader.
• What the linking page promised: the anchor text and the sentence around it tell you what the publisher thought they were sending their reader to, which is the standard your page has to meet.

In our implementation work, the comparison that decides this is purpose to purpose rather than topic to topic. A dead free calculator is not replaced by an article explaining the formula, however good the article is. A dead tutorial is not replaced by a product page about the same subject. When the honest answer is "mine is adjacent", the prospect is finished, and finishing it there costs nothing.

Sometimes the archive has nothing useful: no capture exists, or the crawl caught only the navigation. Say so to yourself rather than proceeding on the slug. In that case the anchor text and its surrounding sentence are the only evidence you have about what the link was for, and that is thinner evidence. Treat those prospects as lower confidence rather than as equivalent.

What does a pitch look like when the page is genuinely dead?

Short, because it is a bug report with a suggestion attached rather than a persuasion exercise. Three parts, in this order.

Part 1. The specific defect: the exact broken URL, and where it sits on their page, so they can find it without hunting. If it is a soft 404 rather than a hard one, say that the page loads but shows an error, because otherwise their own check will contradict you.

Part 2. What the dead page was: one sentence describing what their reader was supposed to get. This is where the archive reading pays off, and it is also the part that proves you looked.

Part 3. What you have: one sentence on your page and why it does the same job, not why it is good. The replacement is an offer, not the price of the information.

Where does that sit against Google's spam policies? The test the policy applies is one of purpose: it catches links made "primarily for the purpose of manipulating search rankings". The practices it enumerates run wider than paid placements: alongside buying and selling links for ranking purposes and excessive link exchanges, the list covers automated link generation, links distributed through footers and widgets, and optimized links dropped into forum posts.

Reporting a defect does not itself create a link. It hands someone the information needed to repair their own page, and whether your URL becomes that repair is their editorial call, made after you have stopped talking. What would move this into the policy is paying for the placement or trading something for it, and that is a decision you make rather than a property of the tactic. Send the report even when your page is a weak fit and you expect nothing back.

What should you expect from a broken link building campaign?

Nothing you can safely borrow from someone else's campaign. The number worth holding yourself to is the one your own funnel produces. Ahrefs' guide to the tactic, published in May 2022, states no success rate, response rate or conversion figure for it anywhere on the page. That is a statement about the one page we read end to end, not about the category.

The useful response is to stop looking for an external rate and measure your own funnel instead. Three numbers, tracked from the first campaign.

Number	What it tells you
Prospects qualified, out of prospects found	Whether your sourcing is aimed at the right kind of page
Pitches sent, out of prospects qualified	Whether the relevance test is being applied or quietly skipped
Links live, out of pitches sent	The only outcome number, and the slowest of the three to move

The first number is the one that moves first and the one you fully control, which makes it the right thing to optimize in month one. If the qualification rate climbs while the outcome number sits still, the sourcing improved before the outreach did. That is the correct order rather than a problem to fix.

Do not import a rate from someone else's write-up. A figure produced in a different subject area, with a different page and a different sender, tells you nothing about your own list.

When is broken link building the wrong tactic?

Whenever one of four conditions holds, and three of them are about you rather than the SERP.

• You have no genuinely matching page: and the tactic is not a reason to write one. A page built to replace one specific dead URL is a page built for one publisher, and it reads that way.
• Nobody owns the outreach: every step above is manual, and the qualification pass is the part that gets dropped first when it is nobody's job.
• Your niche has no resource pages: some subject areas simply do not maintain link lists, and the prospect pool never reaches a useful size.
• You need links this quarter: the qualification pass exists to remove prospects rather than to keep them, and whatever survives it depends on other people's editorial timelines.

If two or more of those hold, the honest move is a different play rather than a harder push on this one. Our guide to white-hat link building sets out the alternatives and where each one fits.

The habit worth building is the cheapest one in the whole tactic: qualify before you write. A prospect removed in two minutes of checking costs nothing. A prospect emailed without checking spends an email, and sometimes a first impression, on a defect that was never there. Broken link building is one tactic inside a wider authority programme, which is how we frame digital PR.

This week, take the last broken-link report you ran, pick ten prospects off the top of it, and run the three checks on each. Count how many survive. That number, rather than a benchmark from someone else's campaign, is your actual starting point.