The Skyscraper Technique in SEO: Does It Still Earn Links?

The skyscraper technique is a link building method with three moves: find a page in your niche that has already earned links, publish something better on the same topic, then email the people linking to the original and ask them to link to yours instead. If someone has handed you that as this quarter's link plan, the question you need answered is not how it works. It is whether it deserves a month of your team's time. This piece is a verdict rather than a tutorial, written for marketing and growth leads at B2B software companies who have to make that call. It puts two published campaign write-ups side by side, one on Backlinko's own page about the technique and one from a team that ran it and earned nothing, then gives you a four-question test for working out in advance which of those two outcomes you are heading for.

The short version

• Two first-hand campaign write-ups land at opposite ends: 17 links from 160 emails, and zero links from 230. Both are single campaigns, and neither is a rate to plan against.
• The three steps are not where it fails. Step three is where the cost lands, but step one, choosing the target, is what decides the outcome.
• It still earns links under conditions you can check before you start. The check is four questions long and takes an afternoon.
• "Better" does not mean "longer". If your version does not change the answer, you have no pitch to send.
• Nothing in the technique breaches Google's spam policies. Whether it works is an entirely different question.

What does the skyscraper technique involve in SEO?

Three steps, and Brian Dean's page at Backlinko names them as "Find link-worthy content", "Make something even better", and "Reach out to the right people". You identify a page that has already collected backlinks, build a version you judge to be stronger, and pitch it to the sites already linking to the original.

The logic is straightforward: if a page has links, demand for that subject is already proven, so you are not gambling on whether anyone cares. You are gambling on whether your version is enough of an upgrade to be worth someone editing a page they already published.

That is the whole method. It is genuinely simple, and the simplicity is also what makes it easy to run badly. Everything that follows here is about the gap between the three steps as written and what happens when a team actually executes them.

What do the technique's own published numbers say?

One campaign, reported at an 11% success rate. Backlinko's page states it verbatim: "Out of 160 emails I landed 17 links: an 11% success rate." The same page carries the headline claim "+110% Traffic in 14 days" and reports that "organic search traffic to my entire site...doubled in just 14 days".

Read those numbers with their conditions attached, because the conditions are doing most of the work. This is one campaign, run by the author of the page, promoting a piece he published in April 2013 about Google's ranking factors. The write-up reports an email list of 160 prospects on that one topic, gathered from the links already pointing at the page being outscaled. That is itself a measure of how much link demand the subject carried before any outreach started. Treat it as a favorable starting position rather than a neutral one.

What that figure is not: a reply rate you can forecast against. An 11% success rate on one campaign, published by an advocate of the method, is the best documented case rather than the expected case. Anyone quoting "around 11%" as what skyscraper outreach returns has turned a single data point into a benchmark it was never built to support.

What happened when a different team published its results?

Zero links, from 230 emails. In 2017, Olga Mykhoparkina, then CMO at Chanty, published a first-hand account of a failed skyscraper campaign in Search Engine Journal. The write-up reports 230 emails sent, 24 answers, three promises to link, 15 replies involving a paid link instead, and a final tally the article states plainly: "Backlinks: 0. Zero. None. Nada."

The asset was not thin. The account describes a 40-hour effort producing an article of over 4,000 words against competitors averaging around 2,000. By the technique's own instructions, they did step two correctly.

Reported	Backlinko campaign	Chanty campaign (SEJ)
Emails sent	160	230
Links earned	17	0
Stated outcome	"an 11% success rate"	"Backlinks: 0. Zero. None. Nada."
Topic	Google's ranking factors	Team communication
Published	Campaign 2013; page last updated 2024	2017

The account names three reasons, and the two a team can act on both sit in step one rather than step three. The first is topic choice: the subject did not motivate anyone to link, unlike a ranking-factors resource. The second is that the site had no established authority behind the pitch, which the article puts bluntly as nobody knowing who you are.

The two campaigns above ran the same three steps and landed at opposite ends of the range. That spread is the actual finding. It tells you the technique's result is set by the situation you point it at, not by how well you follow the instructions.

Why does page one for this technique look the way it does?

Because on page one for this query, the ranking pages are overwhelmingly pages teaching the tactic rather than judging it. We pulled the US SERP for skyscraper technique through Ahrefs on 16 September 2026. Of the nine editorial results at positions 2 through 10, eight teach or define the technique rather than judging it. The one result that questions the technique outright is a Reddit thread sitting at position 10, titled "SEO Tip #27. Ditch the Skyscraper Technique Already".

The authority numbers on that page are worth seeing, because they undercut the advice being given on it:

• Referring domains: the eight non-forum results range from 0 to 13. Four of them have none at all, and Ahrefs returns no referring-domain figure for the Reddit thread.
• URL Rating: 0 to 7 across the set.
• Measured organic traffic: roughly 145 visits combined across all nine, with the strongest single page at 99 and five pages at zero.

Put plainly: on this SERP, most of the pages teaching how to earn links with this technique have earned very few themselves.

The limit on that observation, and it matters: this describes one page of results for one query on one day. It is not a description of what has been written about the technique. Substantial critiques and documented failure accounts do exist, including the Search Engine Journal piece cited above, which has been public since 2017. They simply do not rank for the technique's name, which is why the reader who searches it meets a wall of instruction and no verdict.

When does replication plus improvement still earn links?

When four conditions hold at once. This is the test worth running before you commit the month, and it takes an afternoon rather than a sprint.

Question	A yes looks like	A no means
1. Does a page on this topic already attract links?	Existing pages on the subject carry a real spread of distinct linking domains, and some of those links are recent.	Too little link demand to redirect. Stop here.
2. Does your version change the answer, or only the length?	You have data, a method, or a correction the existing pages do not contain.	You have a longer article and no pitch. Stop here.
3. Does the recipient gain something by editing their page?	Their link is broken, outdated, or points at something your version demonstrably supersedes.	You are asking a stranger for unpaid work.
4. Can you sustain the outreach volume the yield implies?	Someone owns outreach for the full campaign and has the hours in their week.	The asset ships and the pitching quietly stops.

Question one decides the other three. That is what I've seen in practice: where nobody has ever linked to a page on the topic, no amount of improvement manufactures the demand, and the campaign quietly becomes cold outreach with extra steps. Questions three and four then decide whether a genuine yes on the first two actually converts.

How to read your score: four yeses means the technique is a reasonable use of the month. A no on question one or two means do not start, because those two are structural and no effort in step three repairs them. A no on three or four is a resourcing problem, which is fixable, but not by writing more words.

What does "better" have to mean now?

Different, not longer. Length is the easiest reading of "make something even better", and it is the reading with no pitch inside it. A page that says the same things at greater length gives an editor nothing to tell their reader.

The improvement that still works is information gain: something in your version that cannot be found in the page you are replacing. Original data you collected. A method written out so it can be followed. A claim in the existing page that is now wrong, and the evidence that it is wrong. A decision framework where everyone else has a list.

The test is a single sentence. If you cannot write the one-line reason your version is worth linking to, without using the words comprehensive, complete, or updated, you do not have an improvement yet. Go back to step two rather than starting step three.

How long does a skyscraper campaign take, and what does it cost?

Longer than the asset, because outreach is the real line item. Neither of the two accounts above is a benchmark for your team, but both give a sense of the unit. The Chanty write-up describes a 40-hour effort for the article alone, then 230 emails on top. The Backlinko campaign reports 160 emails. Outreach at that volume is a job someone holds for weeks, not an afternoon of sending.

Budget against the two things you control rather than a timeline someone else hit. The first is the asset build, which is a known quantity once you have scoped it. The second is outreach capacity, which is where campaigns die: a well-built asset with nobody pitching it earns whatever links wander in on their own.

The honest planning assumption: treat the link count as unknown and the hours as known. You can commit to sending 200 pitches. You cannot commit to earning 20 links, and any plan that does is quoting someone else's campaign as your forecast.

Is the skyscraper technique black hat?

No. Google's spam policies define link spam as "the practice of creating links to or from a site primarily for the purpose of manipulating search rankings", and the behaviors it documents include paid links, link trading at scale, automated link generation, and paid articles carrying ranking credit. Emailing someone to tell them a better resource exists appears nowhere in that list.

One bullet in that policy is worth holding onto, though: "Creating low-value content primarily for the purposes of manipulating linking and ranking signals". A skyscraper asset built only to attract links, adding nothing a reader would value, is the version that drifts toward the line. The technique is not the problem. Building filler in order to run it is. Our guide to where Google draws the line on links sets out the full boundary.

What should you run instead when the test says no?

Work backward from which question failed. A no on question one means the topic has no link demand, so the fix is choosing a subject that does, or building an asset that creates demand rather than borrowing it. A no on question two means your differentiator is missing, which is an asset problem before it is an outreach problem.

• No link demand on the topic: stop replicating and start building. Our breakdown of which asset formats actually earn links covers what each format costs and who can realistically pull it off.
• No differentiator: the asset needs original input, not more sections. Data you gathered, a method you can document, or a position you will defend.
• No outreach capacity: pick a tactic with a lower per-link labor cost. Our guide to which tactic fits your stage maps the options against team size and authority.
• No authority behind the pitch: this is the constraint the Chanty account named, and outreach volume does not solve it. Earning coverage through a story worth writing about builds the recognition that makes later pitches land.

From my experience, the assets that earn links reliably are the ones answering a question the reader could not answer anywhere else, not the ones answering a familiar question at greater length. That is the same standard question two applies, which is why teams that fail the test rarely need a different tactic so much as a better asset. If the work needs a partner rather than a playbook, our digital PR practice exists to build exactly that kind of authority.

One small discipline outlasts any single campaign: before committing to a link tactic, write two lines. What the target page's existing links prove about demand, and what your version adds that cannot be found anywhere else. Teams that cannot fill in the second line stop shipping campaigns that were never going to earn anything.

This week, take the one page you were planning to outscale and run questions one and two against it. Pull its referring domains, count the distinct ones, and write the single sentence explaining why your version deserves the link. If that sentence needs the word comprehensive to work, you have your answer before you have spent the month.