Broken Link Building: A Step-by-Step Guide That Works

Broken link building is a simple trade. You find a page that has died but still has links pointing at it, you publish a live page covering the same ground, and you tell the people whose links are now broken. They fix a fault on their own site. You get the link.

It works for a reason that has nothing to do with clever tactics. Every other outreach email asks an editor for a favour. This one reports a defect. A dead link is a small embarrassment — it makes a page look unmaintained and it sends readers to a 404 — and the person responsible usually wants to know. That changes the conversation from selling to helping, which is why the tactic survives while others decay.

It also has the worst reputation for wasted effort in outreach, and the reputation is earned. Most dead URLs have one or two links from pages nobody maintains, and most emails go to addresses nobody reads. The difference between broken link building that produces links and broken link building that produces nothing sits almost entirely in the qualification step, before a single email is written.

Sources are not equal. In rough order of yield:

• Competitor 404s. Export a competitor’s backlink profile and filter for target URLs returning 404 or 410. Every one is a page in your niche that earned links and no longer exists. This is the best source because the links are already topically matched to what you sell.
• Dead sites and sunset products. Tools that shut down, blogs that lapsed, companies acquired and folded. An entire domain’s worth of inbound links pointing at nothing, and an obvious replacement need.
• Resource pages. Find the roundup pages in your market, crawl their outbound links, and keep the ones that fail. You get a broken link plus a page that already exists to list resources like yours.
• Your own 404s. URLs on your site that once had links, from an old migration or a deleted post. Reclaiming these needs no outreach at all, only a redirect.

The last one gets skipped constantly and is usually the fastest win available. Pull your crawl errors report, cross-reference it against your own backlink data, and redirect anything with links to the closest live equivalent. Do that before you spend a day building prospect lists for other people’s dead pages.

This is where the tactic is won or lost. For each broken URL, check:

• How many distinct domains link to it. One or two rarely justifies building a page. Five or more and the effort spreads across a batch of emails.
• Whether it is really dead. A 301 or a temporary 503 is not an opportunity. Confirm a 404 or 410, and check twice, a few days apart.
• Whether the linking pages are alive. Indexed, updated in the last year or two, reachable through internal links. A broken link on an abandoned page stays broken forever.
• Who maintains them. A named author or a section editor with a current contact route. No contact, no opportunity, however good the link looks.
• The context of the link. Read the sentence around it. If it cites a specific figure or a specific tool, your replacement has to cover that, not merely the broad topic.
• Whether you would build the page anyway. If the answer is no, the link is not worth the content it would take.

Ten qualified dead URLs are worth more than three hundred unqualified ones. Ruthless filtering here is what separates a week of work that ends in links from a week of work that ends in an empty inbox and a theory about why the niche is difficult.

Open the dead URL in a web archive and read what was actually there. You are not cloning it — copying archived content is both a copyright problem and a bad page — you are matching the intent that earned those links in the first place.

Three practical rules:

1. Match the promise, beat the execution. If the dead page was a checklist from years ago, yours is the same checklist, current, with the gaps the original left. If it was a data page, the data has to be yours and it has to be newer.
2. Keep the anchor text true. Existing links say things like “guide to warehouse safety signage”. If your replacement is a product page with one paragraph on signage, the editor will decline, and correctly. The page has to be a fair substitute for what the anchor promises.
3. Build once, pitch many. Only commit to a page when several qualified dead URLs on the same topic can point at it. Building a bespoke page per link is how this tactic becomes uneconomic.

Keep the page free of gates, interstitials and pop-ups. An editor checking your replacement gives it about ten seconds and will not fill in a form. A page that demands an email address before it shows anything is an easy no, and you will never hear why.

The email is short, factual, and does the editor’s work for them. What it needs:

Leave out the compliment about their excellent blog, the paragraph about your company, and any mention of SEO. Subject lines that work are literal: the page name and the words broken link. Send from a named person’s address with a real signature, and send one follow-up after about a week, then stop.

The offer to remove the link matters more than it looks. It signals that you are reporting a fault rather than trading, and editors who take that option often remember you for the next roundup.

A yes in your inbox is not a link. Editors forget, edits sit in a draft, a link goes up carrying a nofollow attribute, or it goes live and disappears at the next redesign. Broken link building generates a lot of promises, which makes it the tactic where untracked outreach leaks the most value.

Record every prospect with a state and a date: requested when you sent the email, promised when they agreed, live when you have seen the link on the page yourself, lost when it goes. Verify the live state by hand — open the page, find the anchor, check the rel attribute — because a promise filed as a link quietly inflates every number you report afterwards.

Past roughly fifty placements, re-checking by hand stops happening. That is the case for automated live and lost checking of the kind in Growmindr’s backlink tracker: each placement is re-crawled on a schedule and stamped with its current state, so a link that vanishes in month seven arrives as an alert rather than as an unexplained ranking dip six months later. A short follow-up referencing the original fix recovers a fair share of them.

The honest limitations, so you can decide whether it deserves a slot in your month:

• Reply rates are low. This is a volume tactic dressed as a precision one. Plan for dozens of qualified prospects per replacement page, not five.
• Editors sometimes just delete. Removing the dead link is less work than evaluating your page. That is a legitimate outcome, and you offered it anyway.
• Contact routes are the bottleneck. Plenty of strong prospects have no byline, no address, and a contact form that goes nowhere.
• Scrapers pollute the counts. A dead URL with forty links can turn out to be forty copies of one syndicated article.
• Some sites are abandoned. If nothing has changed there in three years, the link will not be fixed no matter how good your email is.

Used well it is a reliable mid-tier tactic: relevant links, no payment, and a defensible reason to make contact with people you want to know anyway. Used as the whole programme it stalls, because the supply of good dead URLs in any one niche is finite. Run it alongside work that creates something worth citing, not instead of it.