The Skyscraper Technique Has Been Measured Three Times in Thirteen Years
Eleven percent in 2013, 5.71% in 2020, zero in 2017. Three public campaigns in thirteen years, and not one of them appears on the pages teaching the method today.
On 26 September 2026 I pulled the ten pages currently ranking for "skyscraper technique", stripped the navigation and the newsletter forms out of each one, and read all 19,890 remaining words looking for a single thing: a number. Not a promise, not a screenshot of a traffic graph, a rate. How many emails, how many links.
One page had one.
Six of those ten pages tell you the method has become saturated. One of them tells you what it converts at now, and the figure it gives is a range someone else published. Meanwhile the number that everybody still quotes is the one its author reported about himself in May 2013: seventeen links from a hundred and sixty emails, an eleven percent success rate, sent once, by one person, in one niche.
In thirteen years the technique has been properly measured in public three times, by three different parties, and the three results are 11%, 5.71% and zero. Nobody teaching the method mentions that they disagree. Most of them do not mention any of the three.
That gap is the subject of this post. Not whether the technique works, which is the argument everybody is already having and nobody is settling, but the fact that the most-copied link building method ever published is also the one with the least evidence attached to it.
The plain definition first, because most people arriving here want it. The skyscraper technique is a three-step link building method published by Brian Dean in May 2013: find a page that already has a lot of backlinks, publish something better on the same subject, then email the people linking to the original and ask them to link to yours instead. That is the whole of it. A skyscraper technique example, in its canonical form, is Dean's own: one resource, one hundred and sixty emails, seventeen links. Skyscraper SEO, skyscraper content and the skyscraper method are all the same three steps under different names.
The short version of everything below:
- Three public measurements in thirteen years: 11% (Brian Dean, May 2013), 5.71% (Ahrefs, August 2020), and zero links from 230 emails (Chanty, November 2017). None of the ten pages ranking for the keyword cites the Ahrefs figure or the documented zero.
- 1 of 10 ranking pages states any response rate at all. I counted, by hand, on 26 September 2026.
- 6 of 10 tell you the method is saturated, and none of the six supports that with a measurement.
- The most-cited 2025 decline figure ("1-3%") is attributed to a page that does not contain it.
- Google's own quality rater guidelines reversed on length between 2022 and 2025, which removes the assumption step two rests on.
What the skyscraper technique actually says#
The method was published on 16 May 2013 under a title that did not contain the word skyscraper at all. The earliest archived capture of the page, taken eight days later, is headed Link Building Case Study: How I Increased My Search Traffic by 110% in 14 Days. The name was in the body, not on the billing.
Three steps, quoted exactly as they appear on the live page today:
Step 1: Find link-worthy content Step 2: Make something even better Step 3: Reach out to the right people
Read them again and notice where the work is. Step one is research. Step two is production. Step three is the only step in which a link can actually be created, because a link is a decision made by a person on another website, and the only step that reaches a person is the third one.
The case study underneath the steps is specific in a way most of its descendants are not. The author published a resource, built a prospect list from the sites linking to the incumbent, and states the outcome plainly: "Out of 160 emails I landed 17 links: an 11% success rate." He also states the cost, which almost nobody quotes: "It took 10 gallons of coffee and 20 hours of sitting in front of my laptop." Twenty hours, a hundred and sixty emails, seventeen links.
Those are honest numbers, reported by someone with no obligation to report anything. They are also a sample of one, and they were taken before the method existed as a method.
How to use the skyscraper technique in 2026, in five steps#
Nothing is held back for later. This is the whole method as I would run it now, and every step links down to the section that argues it.
- Pick the incumbent by its linkers, not by its ranking. Export the referring domains of the page you intend to displace and open ten of them. You are not checking authority scores, you are checking whether those sites are still publishing and still editing that page. A dead list cannot be revived by a better article, and this is the check that separates an 11% campaign from a 0% one.
- Write down what you expect before you start. How many prospects, how many emails you can genuinely personalise, what reply rate you would call a success. This takes four minutes and it is the step that makes everything afterwards knowable. The industry has managed three public measurements in thirteen years and mostly quotes none of them.
- Make it better along a dimension you can name in one sentence. Not longer. If you cannot finish the sentence "mine is the only one that ..." then you have not made something better, you have made something bigger, and Google deleted "too short" as a quality failure and added "padded".
- Send fewer emails, each one referencing the specific link that person gave. The original method's whole advantage was that every recipient had already proved they cared, by linking. A template destroys that advantage, which is why the one template everybody copied gets deleted on sight now.
- Record your own number and publish it internally. Emails sent, replies, links. One row per campaign. After three campaigns you own a benchmark that fits your niche, and you can stop borrowing one from a case study whose published rates range from 11% to zero. Then spend the saved hours on the next list.
The steps have not changed since 2013. What has changed is that every one of them now has thousands of people executing it against the same pages, and almost nobody kept score.
The three times the skyscraper technique has been measured#
Eleven percent is a good result. It is also a single observation, and it is worth being precise about what it was an observation of.
In April 2013 the author published a long reference page, then built his prospect list from the sites already linking to the incumbent resource. His own words: "In my case, after cleaning up the list, I had an email list of 160 very solid prospects to reach out to." Then: "I emailed all 160 of them using this template:" And the outcome, with his own framing intact: "Considering that these were cold emails that asked for a link in the first email, an 11% success rate is pretty amazing."
Two other parties have run the method in public and published their arithmetic.
In August 2020, Ahrefs documented a campaign in detail: "In total, we sent 515 emails. 473 were delivered, and 42 bounced." The results: "Based on deliverable emails, that puts our conversion rate at 5.71%, meaning that 27 of the websites we contacted linked to us." The reply rate was much higher than the link rate, and the gap is the interesting part: "The conversion rate for our campaign was 5.71%, but the reply rate was 17.55%. That means 83 people responded to our email, but only 27 linked to us." Eighty-three humans answered and fifty-six of them said no. Eight of them asked for a link swap and six asked for something else in return.
In November 2017, Chanty's CMO published a failure on Search Engine Journal, with the same expectation and the opposite outcome. "Having sent 230 emails, we've expected to see the same 11 percent success rate as it was claimed initially in the Skyscraper technique." What they got: "Backlinks: 0. Zero. None. Nada." Twenty-four replies from more than two hundred emails, three promises that were not kept, and fifteen offers to sell a citation for between forty-five and two hundred dollars. Her diagnosis is the one nobody wants: "If you're just starting out, you probably haven't built up enough authority and brand recognition ... Put simply, nobody knows you."
Three campaigns, thirteen years, and the spread runs from eleven percent to nothing at all. Those are not competing estimates of one underlying rate. They are three different situations that happen to share a method name, and the difference between them has far more to do with who was sending than with what was sent.
Three things about the original campaign have quietly stopped being true.
The first is that the recipients were warm. Every one of the 160 had demonstrated, by linking, that they cared enough about this exact topic to spend a link on it, and in 2013 almost nobody was emailing them about it. The list was warm because it was unworked. That is a property of the moment, not of the method.
The second is that the template was new. There was one template, it had been sent by one person, and the people receiving it had not seen its shape before. Ahrefs, thirteen years later and writing about the same template: "Unfortunately, many SEOs decided to use this exact template word for word ... Judging by my inbox, there are still people using this exact template. And, like everyone else, I delete the email immediately."
The third is that the effort was honestly priced. Twenty hours for seventeen links is roughly seventy minutes per link, and the author said so. Almost every guide that has since reproduced the method reproduces the eleven percent and drops the twenty hours.
A rate without its denominator, its date and its cost is not a benchmark. It is a souvenir.
What thirteen years did to the original skyscraper technique page#
I did not expect this part. I pulled the earliest Wayback capture of the original post, from 24 May 2013, and diffed it against the live page, expecting to find a caveat added somewhere as the method spread. There is no caveat. There never has been. What I found instead was a page that has been edited substantially in every respect except the one that matters.
The three steps are unchanged, character for character. The only difference in that passage is that "There are 3 basic steps to The Skyscraper Technique:" lost the word "basic".
Everything around them moved.
The title was rewritten. What was published as "Link Building Case Study: How I Increased My Search Traffic by 110% in 14 Days" is now "Skyscraper Method: +110% Traffic in 14 days (Link Building Case Study)". The method's name was promoted into the billing after the method became famous, which is ordinary SEO housekeeping and worth noticing anyway: the page was retrofitted to the thing it turned into.
The headline outcome number grew. The 2013 capture reads "that single post has driven more than 1900 referral visitors to my site so far." The live page reads "more than 2 milion referral visitors to my site so far" (the misspelling is in the live HTML). Both are defensible statements about their own moment. Together they are a thousand-fold restatement of the same sentence, and anyone citing two million as an outcome of a fourteen-day case study is reading a number that was 1,900 when the case study happened.
Step one's method was deleted. In 2013, step one carried three named sub-sections of actual instruction, headed "OSE Top Pages", "Topsy" and "Google Search", with thresholds as concrete as looking for "content with at least 100 linking root domains". Open Site Explorer and Topsy no longer exist, so the instruction had to go. But it was not replaced with new instruction. It was replaced with an embedded video. On the live page, the sentence "That's easy: find content that's already generated a ton of links. Here's how:" is followed immediately by a video player and then by step two. The most-cited link building method on the internet currently has no written step one.
The proof for the seventeen links was removed. The 2013 version qualified the haul: "There were a lot of gems in that group of 17 links, including a link from a DMOZ listed site on a PR5 page." DMOZ closed in 2017 and PageRank scores stopped being published, so the sentence went. The seventeen links survive on the page. The evidence about what they were does not.
And the page now carries a disclosure that sits oddly against its byline. It still says "Written by Brian Dean". It also says: "Backlinko is owned by Semrush ... Unless otherwise noted, this content was written by either an employee or paid contractor of Semrush Inc." Semrush acquired Backlinko on 19 January 2022, and Adobe completed its acquisition of Semrush on 28 April 2026. The page's own metadata gives a last modification date of 20 June 2024, two years after the author sold the site. Whoever has been editing it since has removed dead tools and dated proof, added a course plug and a tool CTA, and never added a line about how much harder the method has got.
That line exists. It has just never been put where the method is taught.
The author said it out loud on a podcast, Growth Manifesto, dated 2 September 2020. The transcript on that page is auto-generated and reads like it, so treat the wording as approximate rather than quotable prose, but the substance is unambiguous. Asked about the technique, he says the size of content "has just gone out of control", that some pieces are "30 to 40 000 words long", and that "as the technique has become popular yeah the bar just keeps getting higher and higher". He then recommends competing on a dimension other than length.
He has also written it down, on a different page of the same site. Link Building Strategies: The Complete List, last updated 29 December 2025, says: "The one downside of The Skyscraper Technique is that it takes A TON of work. Especially now that people started publishing higher-quality stuff." That is a real concession, and it is current.
So the caveat has existed in public since at least 2020. It is on a podcast, and on a list post, and it is not on the case study that two decades of SEO writing point at. A reader arriving on the original page in 2026 from a search for this keyword gets the 2013 result, the 2013 confidence, and no indication that anything has changed in the world it was measured in.
I counted what the ranking pages actually teach#
I went into this expecting to prove something specific and it did not survive contact with the data, so here is what actually happened.
My hypothesis was that the ranking pages had memorised step two and forgotten step three: pages and pages about how to make something better, a paragraph at the end about emailing people. I counted to prove it. The count came back level.
Across the ten pages, 19,890 words of body copy, 5,200 words go to step two and 4,861 go to step three. That is 1.07 words about making the content for every one word about the outreach. Take the unweighted average of the per-page shares instead and step three is actually ahead, 27.3% to 23.5%. Drop the single longest page in the sample, which carries 27% of all the words in it and is the most step-two-heavy, and the ratio inverts outright to 0.77 to 1. Count strictly, using only the blocks that sit inside each page's own explicitly headed step-two and step-three sections, and it is 0.88 to 1, still in step three's favour.
By volume, the internet writes about the outreach half of this method about as much as it writes about the content half. My hypothesis was wrong and I am not going to bury it.
What the same count found instead is sharper. Ten pages, tens of thousands of words about how to run an outreach campaign, and: four of them print an email template a reader could actually copy; six of them tell you the technique is saturated or overused or "doesn't work like it used to"; one of them states any conversion figure at all, a range of "1% to 5%", on seoClarity's page. None of the ten cites the Ahrefs campaign's 5.71%. None cites the documented zero.
So the problem is not that the industry stopped talking about step three. It talks about step three the way you talk about weather. Six pages agree that conditions have changed. None of them brought a thermometer.
There is a reason word volume flatters the outreach half here, and it cuts against my own figure, so it belongs in the open. A large part of those 4,861 words is tool walkthroughs rather than craft: how to filter a referring-domains export, how to find an email address, how to warm a sending domain. The pitch itself, the part where you have to say something a human will respond to, is usually a bullet list. Volume is not instruction.
The eleventh page, and the number that is not where it says it is
One page in my sample would not load. Search Engine Land's guide, The skyscraper technique's surprising transformation in the AI era, published 17 November 2025, blocks automated fetching, so I read it through an archived copy and excluded it from the count rather than treat it differently from the rest.
It is the most authoritative-looking recent page on this keyword, and it carries the statistic that everything else in 2026 seems to be quoting:
According to reporting by SerpReach, 61% of SEOs still believe the technique works, yet actual success rates have plummeted from the original 10-20% benchmarks down to a dismal 1-3% for most practitioners.
I went to the SERPreach page it credits. The strings "1-3%", "1–3%", "10-20%" and "dismal" appear on it zero times. What that page actually says about the poll runs the other way: "In a recent poll by Ahrefs, 61% of SEOs confirmed that skyscraper link building still works when executed correctly." The Ahrefs poll in question is from January 2023 and has 1,242 votes, so "recent" is doing work too.
The single most-quoted 2025 figure for this technique's decline is attributed to a source that does not contain it. I am not alleging bad faith and I have no way to know how it got there. I am saying that the number you have probably seen, the one that makes the drop feel measured, has no measurement under it, and that this is exactly what happens to a topic where nobody has run the experiment for six years.
The same thing has happened in the other direction. Ross Hudgens' When the Skyscraper Technique Comes Up Short is routinely cited as an argument against the method, including by publishers who should read to the end. Its actual verdict, in its own words, is "Skyscraper Technique: Yay or Nay? The answer is a resounding yay. The technique works, and it's the foundation of our content marketing services." The piece is a critique of misapplication written by a believer. Its original URL now returns a 404, which is presumably why so few of the people citing it have read it.
Skyscraper content and length: why "better" became "longer", and what Google says#
Step two says "make something even better". It does not say make something longer. The original page, however, does say this, and the sentence is still live:
In some cases, publishing an article that's simply longer or includes more items will do the trick. If you find a link magnet with a title like "50 Healthy Snack Ideas", publish a list of 150 (or even 500).
Five hundred snack ideas. That is the seed of thirteen years of arms race, and Google has spent most of those years saying the opposite in increasingly plain language.
Its SEO Starter Guide carries a section titled "Things we believe you shouldn't focus on", and inside it, a sub-heading reading "Minimum or maximum content length":
The length of the content alone doesn't matter for ranking purposes (there's no magical word count target, minimum or maximum, though you probably want to have at least one word).
The helpful content documentation puts the same thing as a diagnostic question you are supposed to fail: "Are you writing to a particular word count because you've heard or read that Google has a preferred word count? (No, we don't.)" John Mueller, on Reddit in August 2019, in a comment that is seven words long: "Word count is not a ranking factor. Save yourself the trouble." Google's public Search Liaison account, June 2023: "The best word count needed to succeed in Google Search is ... not a thing! It doesn't exist."
None of that is new, and quoting it is not the interesting part. The interesting part is that Google's own quality rater instructions have reversed direction on this, and you can read the reversal.
The Search Quality Rater Guidelines are the document Google gives to the people who grade search results. In the version dated 28 July 2022, section 6.3 was headed "Unsatisfying Amount of Main Content", and having too little was a reason to rate a page Low:
Some Low quality pages are unsatisfying because they have a small amount of MC for the purpose of the page.
In the version dated 11 September 2025, that section does not exist. In its place, at 5.2.2, is a section headed "Filler as a Poor User Experience":
Sometimes, MC includes "filler" - low-effort content that adds little value and doesn't directly support the purpose of the page. Filler can artificially inflate content, creating a page that appears rich but lacks content website visitors find valuable.
A page now earns a Low rating if it "contains a large amount of low quality and unhelpful filler".
Between 2022 and 2025 Google deleted "too short" as a quality failure and added "padded" as one. The skyscraper technique's most-copied instruction is the practice that moved from safe to penalised while nobody was re-reading the page.
The practitioners got there first, for what it is worth. Ahrefs, on the arms race: "Someone creates a list with 15 tools. The next person ups it to 30. Another 'skyscrapers' it to 50, and the next increases it to 100. Not only is it a never-ending arms race, there's also no value for the reader ... No one wants to skim through 5,000 words or hundreds of items just to find what they need. Curation is where the value is." Ross Hudgens, under a heading that reads "Better Does Not Equal Bigger": "You could list every tool known to man, but you'd actually become less and less valuable with every tool added."
And Semrush, which owns the site the method was published on, says it plainly on its own page about the technique: "Many SEOs misinterpreted 'better' to simply mean 'longer.' They started creating overly lengthy articles that don't provide additional value to readers."
One honest complication, because it is the argument the other side actually has. The same helpful-content document that denies a word-count target also asks: "Does the content provide a substantial, complete, or comprehensive description of the topic?" Comprehensive is about coverage, not length, but Google has never said so in those words, and "comprehensive" is the hook the whole longer-is-better industry has hung itself on for a decade. The instruction is not as unambiguous as the quote-mining on either side suggests.
Why skyscraper outreach stopped converting at eleven percent#
Here is the mechanism, and it has nothing to do with content quality.
The original campaign's advantage was never the article. It was the list. A hundred and sixty people who had each already spent a link on this topic is a qualified audience in the strictest sense: they had proved their interest with an action, not a click. Emailing them was not cold outreach. It was warm outreach to strangers, which is a rare and very effective thing.
That property is not durable, because it is not a property of the list. It is a property of how many other people have the list.
A method that works because its recipients are unworked will stop working in proportion to its own popularity. This is what makes the skyscraper technique different from a tactic that simply gets patched by an algorithm update. Nothing at Google changed to break it. It was consumed by its own readership. Every guide that reproduced it recruited more senders into the same finite set of inboxes, and the thing that made the original conversion rate possible was spent.
The method did not degrade. It was used up.
You can watch the author arrive at the same diagnosis from a different direction. On Reverse Outreach he writes: "Outreach-based strategies like The Skyscraper Technique still work GREAT. (When done right.) But there's one big problem with most link building strategies: They don't scale." Scale is exactly what saturation looks like from the sender's side: the more of it you attempt, the worse each unit performs, because the cost of being specific rises with volume while the value of being generic falls to zero. BuzzStream measured that curve across 31 million emails and 86,000 campaigns in 2025 and found reply rates falling "from 1.15% (1–50 emails) to 0.37% (1001–5000) — a 68% drop", while being careful to add that "most importantly, this is all correlative."
Which brings us to the number you actually want, and to why nobody can honestly hand it to you.
There is no shortage of published outreach benchmarks. There is a shortage of comparable ones. BuzzStream, analysing "over 51 million emails" from its link-building customers in 2025, reports "the overall reply rate for link builders is 3.59%, meaning almost 1 in 28 emails gets a reply". Hunter, from 31 million emails in its 2026 report, gives 4.5% overall and, specifically, "Building links for SEO? Digital PR gets a 13% average response rate". BuzzStream, from a separate dataset, puts digital PR at "less than 1%". Backlinko and Pitchbox's 12-million-email study, still the most-cited figure in the industry, reported 8.5% in 2019 and has not been rerun. Belkins, across 7.5 million emails in 2025, reports 0.45%.
Those numbers span more than an order of magnitude, and it would be easy and wrong to draw a declining line through them. Belkins explains why, in the single most useful sentence anyone has published on this subject:
A 5% reply rate against openers and a 0.45% reply rate against total sends can describe the same campaign; they're just measuring different things.
Some of these divide by sends, some by delivered, some by sequence, some count follow-up replies and some do not, and the populations range from link builders to B2B sales teams. The same vendor reports a 12% median in 2018 and 3.59% pooled in 2025, which is a fourfold divergence inside one company's own data. You cannot benchmark yourself against any of this. You can only benchmark yourself against yourself, which is the entire practical argument of this post.
If the step that never gets done is the one that produces the links, that is the step worth handing over: GenLink runs it and gives you your own number instead of 2013's.
Does the skyscraper technique still work in 2026? The case for yes#
The strongest version of the opposite argument comes from the source, it is current, and it deserves to be quoted at full strength rather than straw-manned.
On Link Building Strategies: The Complete List, last updated 29 December 2025, the position is unambiguous: "I first wrote about The Skyscraper Technique in 2013. And it still works GREAT." The same passage concedes the cost, which is what makes it credible rather than promotional: "The one downside of The Skyscraper Technique is that it takes A TON of work. Especially now that people started publishing higher-quality stuff. But if you're willing to put in the work, this is still one of the best link building strategies out there."
Ahrefs, in the same article where it calls the method oversaturated and deletes the emails personally, still concludes: "Fundamentally, it makes sense that the Skyscraper Technique still works. After all, the principles are the same behind (almost) any link building strategy." And the Ahrefs poll everyone quotes, 1,242 votes in January 2023, had 61% of respondents saying it still works.
Two honest caveats about weight. Those Backlinko pages have been modified under Semrush ownership, so a present-tense claim on them in 2026 cannot be attributed to the author as a 2026 statement. And they sit on the property the method made famous, which is not a disqualification but is a fact a reader should have.
Set the provenance aside, though, because the argument is correct in an important range of cases. In a niche nobody has run this method on, it works exactly as written, for exactly the reason it worked in 2013. If you operate somewhere the SEO industry does not read, and the sites linking to your incumbent have never received an email like this, the list in front of you is as warm as the 2013 list was. Saturation is not a global constant, it is a per-niche one, and the check in step one above is how you find out which kind of niche you are in.
What I cannot give you is the number. My count establishes that the pages ranking for this keyword do not publish one; it cannot conjure one. Three measured campaigns in thirteen years is not a dataset, my own sample of ten pages is small and one page in it carries 27% of the words, and a different counter drawing the line between step two and step three would land on different totals. Anyone who tells you the skyscraper technique now converts at some specific rate is quoting 2013, quoting a figure with nothing under it, or quoting themselves. The third of those is the only one worth having.
Skyscraper technique alternatives: what to do instead of building taller#
The practical translation of everything above is short: stop competing on the axis that is free to copy, and start keeping the number that nobody else has.
Pick the incumbent by the liveness of its linkers, not by its position. A page ranking first with four hundred referring domains that have not published since 2021 is a worse target than a page ranking fourth with sixty that are still editing weekly. The prospecting question is who will act, not who ranks, and that question is answerable before you write a word.
Make the improvement nameable. One sentence, finished out loud: mine is the only one that has the actual policy documents, or the only one with a number counted this month, or the only one written by someone who has run it. If the sentence needs a comma, you are describing length.
Then write to the person, not to the list. The reason the original worked is that each recipient had done something specific and the email could reference it. That is still available to you and it is still the entire advantage; it is just laborious now in a way it was not when the list was 160 people and nobody else had it. Most of this post's argument reduces to a single operational fact: the step that produces links is the step that does not scale, which is why it is the step everyone quietly drops.
That is where the hours go, and it is worth knowing what could replace them.
Step two is the part you can finish. Step three is the part that never gets finished, because it is the same three hours of tab-switching repeated once per prospect, and there is always a more urgent hour available. That is the asymmetry this whole method collapsed on. GenLink takes step three off the list: it replaces the referring-domains export you paste into a spreadsheet, the email-finder tab you never close, the sending domain with no history that gets you filtered before a word is read, and the follow-up you were always going to send tomorrow. It finds the sites, finds the person, writes a pitch that references the specific link they already gave, sends it from a mailbox with real sending history, and triages what comes back. The free tier runs one agent, so you can point it at a single page and watch the whole loop before you decide anything.
A last structural note. The skyscraper technique is one pitch shape among several, and its shape happens to be the hardest kind to land, because you are asking someone to replace a link that currently works. Broken link building asks them to replace one that does not, resource page link building asks them to add to a list they maintain on purpose, and link reclamation asks them to finish something they already started. If your skyscraper campaign is converting badly, the fault may be the pitch shape rather than the article, and the complete link building method is the same argument laid out end to end.
The shape of the whole thing#
Strip out thirteen years of reproduction and the sequence is short.
A method was published in May 2013 with three steps and one honest measurement: 160 emails, 20 hours, 17 links, 11%. The steps were correct and remain correct. The measurement was a sample of one, taken at a moment when nobody else was running the method, which is precisely the condition that made the number achievable.
Since then it has been measured in public twice more, at 5.71% and at zero, and neither result appears on any of the ten pages currently ranking for the keyword. The page that originated it has had its own step-one instruction deleted and its own evidence removed. Its author has conceded, on a podcast and on another page of the same site, that the bar keeps rising; that concession has never reached the case study. The site has changed hands twice. Google has quietly inverted its own guidance on length, from penalising pages for being too short to penalising them for being padded. And the most-quoted figure for the technique's decline turns out to be attributed to a page that does not contain it.
The skyscraper technique is not broken. It is barely measured, and it has been barely measured for so long that the industry mistook an anecdote for a benchmark and then invented a second one to describe its decline.
The fix is not a taller building. It is a spreadsheet with three columns, filled in by you, about your niche, this quarter.
The method has always been three steps, and the third one is the only one that produces a link. Everything above is an argument for spending your hours there instead of on the height of the building. Hand that step to GenLink and what leaves your stack is the prospect spreadsheet, the contact-finding subscription, the cold domain, and the second email nobody sent. What stays with you is the judgement nobody can automate: which page of yours actually deserves to displace the one that is ranking. Start free with one agent and watch it run a single page end to end.
Frequently asked questions#
Does the skyscraper technique still work in 2026? It works wherever step three is still novel to the people receiving it, and that is a property of your niche, not of the method. The test is cheap and you can run it this afternoon: pull the backlinks of the page you intend to beat, open five of the linking pages, and check whether those sites have published anything on your topic in the last two years. Dormant sites cannot be revived by a better article. Sites still publishing weekly are the 2013 condition, and for you the method is intact.
How long should skyscraper content be? Long enough that a person who already linked to the original would feel silly linking to the original again. That is the only length target that means anything, and it is usually shorter than the word counts people quote. Length is a proxy people reach for because it is measurable, and it stopped being a useful proxy the moment it became the thing everyone optimised.
How many people should I email? Fewer than the tutorials suggest, and each one individually. The arithmetic that matters is not how many you send but how many distinct reasons you can write. If you can only write forty genuinely specific opening lines before you start recycling, your list is forty people long, whatever the export says.
What is a realistic response rate? Before you can answer that you have to decide what you are dividing by, and almost nobody says. Replies divided by emails sent, by emails delivered, by people who opened, and by campaigns rather than messages are four different fractions, and published benchmarks quietly use all four. Pick one definition, write it at the top of your sheet, and never change it. Your first three campaigns are your benchmark; everybody else's is a different fraction of a different population.
Is the skyscraper technique against Google's guidelines? No. Asking someone to link to a better resource is not on Google's list of link spam examples, and outreach is not a scheme. What can cross the line is what people bolt onto the method when the emails stop working: paying for the placement, swapping links, or offering something of value in exchange. That pressure is real and it arrives in your inbox unprompted, which is worth knowing before it does. The technique is clean. The panic response to it failing is what gets people in trouble.
What is the difference between skyscraper and broken link building? The pitch. Skyscraper asks someone to replace a working link with a better one, which requires them to agree that yours is better. Broken link building asks someone to replace a dead link, which requires them only to agree that dead links are bad. The second is a much easier thing to get someone to agree with, which is why its reply rates hold up better under saturation.
Why did my skyscraper post get no links? Almost always because the outreach never happened, or happened once. The step-three half of the method is the half that gets cut when a project runs late, and a published page with no emails behind it is not the skyscraper technique at all. It is just a long post. The second most common reason is that the emails went out to a list scraped from the wrong page, so the recipients never linked to the original and had no reason to care.
