What Is the Skyscraper Technique?

The Skyscraper Technique is a link-building method popularised by Backlinko's Brian Dean in the mid-2010s. It has three steps: find a page that has already earned a lot of backlinks, publish something clearly better on the same topic, then email everyone linking to the original and ask them to link to yours instead. It still works occasionally — but its reply rate collapsed once the whole industry started running it.

The three steps, precisely

Find proven link-worthy content. Not a topic you like — a specific URL that has demonstrably attracted backlinks. The evidence that people link to this subject already exists.
Make something better. The original framing offered four levers: more thorough, more current, better designed, or more comprehensive.
Reach out to the right people. Pull the backlink profile of the original page, and email each site that links to it, pointing out your version.

The logic is elegant. You've removed the two biggest gambles in link building — will anyone link to this topic? and who do I email? — because both questions have already been answered by someone else's page.

Why it worked so well at first

In its first few years the technique had a genuine structural advantage. Almost nobody was doing it, so a publisher receiving one of these emails got a novel, well-researched note rather than the fourth one that week. "Better" was easy to achieve, because a lot of the pages earning links in 2012 were 600 words with no images. And content was simply scarcer, so a superior resource stood out.

None of those three conditions holds now. That's the whole story of the technique's decline — not that the logic broke, but that the environment it exploited disappeared.

Why it works far less now

Everyone runs it, so the inbox is saturated. Any page with a healthy backlink profile has been "skyscrapered" repeatedly. Editors recognise the email format on sight.

"Better" got redefined as "longer." The most common implementation is to count the competitor's words and write more of them. Length is not an improvement, and a 6,000-word version of a 2,000-word article that answers the same question worse is a downgrade wearing a bigger suit — the same failure behind thin and padded content generally.

There is no reason for a publisher to swap a working link. This is the flaw the technique never solved. The link in their article isn't broken. It isn't wrong. Editing a live post to point at a stranger's page is real work with no benefit to them. Compare that to broken link building, where the pitch begins with a genuine defect on their site — that one has an actual hook.

The link data is noisy. Backlink tools list every domain pointing at the original, including scrapers, dead pages, aggregators, and syndicated copies. A raw export of 400 "linking sites" often contains a few dozen real, reachable publishers.

AI made mass-produced "comprehensive" content free. When anyone can generate a longer version of anything in a minute, being longer stopped being a differentiator entirely.

When it still works

Two situations, and both share a trait: the recipient gains something by making the edit.

The original is genuinely obsolete or gone. If the top-linked resource cites 2019 data, references a product that shut down, or now 404s, then the outreach isn't "mine is nicer" — it's "the thing you're sending readers to is out of date." That's a favour, and the reply rate reflects it. This is where the Skyscraper Technique and broken link building overlap, and it's the version worth running.

You added something the original genuinely lacks. Not more words — a different kind of asset: original data the original didn't have, a free calculator, an interactive tool, a dataset, current pricing, or first-hand test results. When your page contains a fact that exists nowhere else, a publisher has a real editorial reason to cite it.

Everything else — nicer design, more subheadings, a longer word count — is not a reason for a stranger to edit their article.

How to run it properly

If you're going to do it, do the version with the higher floor:

Choose the target page by link freshness, not link count. A page that earned 300 links in 2016 and none since is a dead vein. One earning links this quarter has an active audience of people who cite the topic.
Filter the link list by hand. Remove scrapers, syndicated copies, dead domains, and anything with no author. What's left is your real list, and it's usually small.
Segment by why they linked. Someone citing a statistic wants a better statistic. Someone linking from a resource list wants a resource. Two different emails.
Lead with what's new, not what's better. "You cite the 2019 figure for X — we re-ran it this year and it's moved to Y" beats "I wrote a more comprehensive guide."
Make the ask small. Suggest adding your page as an additional reference rather than replacing theirs.
Cap the follow-ups at one. The tactic's reputation was destroyed by three-email sequences.

The honest arithmetic

Here's the cost the original framing skips. You spend days producing a genuinely superior resource, hours cleaning a link list, and more hours writing tailored emails — for a small number of links, from publishers you don't control, on a timeline you can't predict. That's worth it when the asset has standalone value and keeps earning links passively. It's rarely worth it when the asset exists only to justify the outreach: if the page wouldn't have been worth publishing without the email campaign attached, the campaign is subsidising a page nobody needed.

The structural problem — you do all the work, then wait on a stranger who owes you nothing — is the same one behind why link building is hard in general. A predictable baseline works differently: with Backlinkster you're matched 1-for-1 with a site owner in a related niche who is already looking to trade, each side places the other's link in their own content, and the system verifies both are live and dofollow.

Frequently asked questions

Does the Skyscraper Technique still work in 2026? In a narrow form, yes. It works when the page you're targeting is outdated or dead, or when your version contains genuinely new information — original data, a tool, current numbers. The generic "I wrote a longer guide" version has a reply rate close to zero.

Who invented the Skyscraper Technique? Brian Dean of Backlinko named and popularised it in the mid-2010s. The underlying idea — find what already earns links and do it better — predates the name, but the three-step framing and the outreach template are his.

How long does a Skyscraper campaign take? Realistically two to four weeks: several days to research and build the asset, a day to clean the link list, then one to two weeks of outreach and replies. Any ranking effect from the links follows the usual delay — see how long backlinks take to work.

Is the Skyscraper Technique against Google's guidelines? No. Publishing a better resource and telling relevant publishers it exists is normal editorial outreach, not a link scheme under Google's link spam policies. It only crosses the line if you pay for the placements.

What's the difference between Skyscraper and broken link building? Skyscraper targets pages that link to a working competitor page and argues yours is better. Broken link building targets pages linking to a dead URL and offers a replacement. The second has a much better reply rate because it starts with a problem the recipient actually has.

The bottom line

The Skyscraper Technique's insight is still correct: build things that have already proven they attract links, and tell the people who already link to that topic. What broke is the execution most people copied — write something longer, mail the whole backlink export, hope. Run the version where your page contains something the original genuinely doesn't, target pages that have gone stale, hand-filter your list, and make a small ask. That's a real tactic with a modest, honest yield. The rest is a template that stopped working a decade ago.