Brian Dean published the original skyscraper technique in 2013. The premise was simple: find content that already earns links, build something better, and reach out to the people linking to the original. It worked well enough that it became one of the most referenced content marketing frameworks of the 2010s. Thousands of agencies ran it. Thousands of blog posts explained it.
By around 2021, it had become so widely practised that it started producing diminishing returns. Everyone was building “better” versions of the same content. Editors were drowning in outreach emails pitching yet another comprehensive guide. The playbook did not break entirely, but it became much harder to run profitably.
In 2026, the landscape has shifted further. AI-generated content has flooded every topic. The bar for what “better” actually means has risen dramatically. And the definition of a link worthy piece has expanded beyond long form completeness to include original perspective, demonstrable expertise, and genuinely novel data.
This piece covers what the skyscraper technique got right that still applies, what has changed, and how to run a modernized version that earns links and, increasingly, AI citations. For the broader context on where this fits in your content operation, the topic cluster strategy guide guide explains how skyscraper-style content relates to pillar and cluster architecture.
By around 2021, it had become so widely practised that it started producing diminishing returns. Everyone was building “better” versions of the same content. Editors were drowning in outreach emails pitching yet another comprehensive guide. The playbook did not break entirely, but it became much harder to run profitably.
In 2026, the landscape has shifted further. AI-generated content has flooded every topic. The bar for what “better” actually means has risen dramatically. And the definition of a link worthy piece has expanded beyond long form completeness to include original perspective, demonstrable expertise, and genuinely novel data.
This piece covers what the skyscraper technique got right that still applies, what has changed, and how to run a modernized version that earns links and, increasingly, AI citations. For the broader context on where this fits in your content operation, the topic cluster strategy guide guide explains how skyscraper-style content relates to pillar and cluster architecture.
What the Original Skyscraper Technique Actually Was?
The three-step original framework:
The theory was sound. Links concentrate on a small number of resources in any given topic area. If you can replace one of those resources with something clearly superior, some percentage of the existing links will migrate to your page.
The insight that made this valuable was that it used existing link profiles as a signal for content that audiences found worth linking to. Instead of guessing what would earn links, you let the existing backlink data tell you. This part remains valid.
What has changed is everything else: what “better” means, how editors evaluate outreach, and what factors determine whether a piece earns links.
The insight that made this valuable was that it used existing link profiles as a signal for content that audiences found worth linking to. Instead of guessing what would earn links, you let the existing backlink data tell you. This part remains valid.
What has changed is everything else: what “better” means, how editors evaluate outreach, and what factors determine whether a piece earns links.
Why the Original Playbook No Longer Works on Its Own?
Problem One: "More Comprehensive" Is Not a Differentiator Anymore
In 2015, finding a 2,000-word blog post and building a 5,000-word version was legitimately better. More depth, more coverage, more usefulness. Today, every topic has thirty 4,000-word blog posts. AI content generation has flooded the market with comprehensive-sounding guides that cover every angle at surface level.
When editors receive skyscraper outreach today, their first question is: “Is this actually different, or is it just longer?” Length and comprehensiveness are table stakes. They are no longer differentiators.
When editors receive skyscraper outreach today, their first question is: “Is this actually different, or is it just longer?” Length and comprehensiveness are table stakes. They are no longer differentiators.
Problem Two: Outreach Inboxes Are Saturated
The editors and webmasters who maintain resource pages and linked articles have been receiving skyscraper outreach emails for a decade. They recognise the pattern immediately: “I noticed you linked to X. We built something better. Would you update your link?” The email structure is so well known that many editors dismiss it on sight.
This does not mean outreach does not work. It means the email needs to be genuinely relevant and specific, not templated, and the content being pitched needs to be visibly and immediately different from what they are already linking to.
This does not mean outreach does not work. It means the email needs to be genuinely relevant and specific, not templated, and the content being pitched needs to be visibly and immediately different from what they are already linking to.
Problem Three: AI Mode Changes What Gets Rewarded
Today, a meaningful portion of search queries are answered by AI Mode without users clicking through to any specific page. The pages that get cited in those AI answers are not necessarily the most comprehensive. They are the pages that AI systems treat as authoritative sources:
pages with original data, clear authorship, demonstrated expertise, and content that makes it easy for an AI to extract and cite specific facts.
A 6,000-word guide that is a better version of a competitor’s 4,000-word guide is not necessarily a more citable source. An 800-word page with original research findings that no other page contains is often more AI-citable than either.
pages with original data, clear authorship, demonstrated expertise, and content that makes it easy for an AI to extract and cite specific facts.
A 6,000-word guide that is a better version of a competitor’s 4,000-word guide is not necessarily a more citable source. An 800-word page with original research findings that no other page contains is often more AI-citable than either.
What the Modernized Skyscraper Technique Looks Like?
The modernized version keeps the core insight (use existing link profiles to identify what earns links) but rebuilds the execution around what actually differentiates content.
Step 1: Find the Right Target Content
The selection criteria have changed. In the original technique, you looked for highly linked content. In the modernized version, you look for highly linked content that has exploitable weaknesses.
The best targets are pages that:
The best targets are pages that:
The weakest targets for a modernized skyscraper are pages that are current, well-researched, and have genuine original insights. You will not out-compete those by adding 1,500 more words. Find pages with real weaknesses you can address.
Practical identification process:
Practical identification process:
Step 2: Build Something Genuinely Better, Not Just More
This is where the modernized technique diverges most sharply from the original. “Genuinely better” requires one or more of the following:
The content gap analysis is the core tool here. Before writing a word, conduct a gap analysis of the top 5 to 10 ranking pages for your target keyword and create a structured inventory of what each page covers and what each one misses. Your piece should cover what everything else covers (so it is complete) and fill the gaps that nothing else fills (so it is genuinely better).
Step 3: The Outreach Campaign
The outreach process works like this in the modernized version:
The realistic expectation for conversion rate is 5 to 15 percent of contacted sites. This means a page with 200 backlinks, with 80 viable outreach targets, should yield 4 to 12 new links from a single campaign. Combined with the organic links the piece earns naturally over time, this is a meaningful lift. For a deeper dive into what makes outreach succeed, the companion piece on building high quality backlinks covers the full outreach system.
What Has Changed About Skyscraper in the AI Mode Era?
AI Mode introduces a new objective alongside traditional link earning: AI citation. For some queries, appearing in AI Mode answers is more valuable than ranking in position 1 of traditional search results. The modernized skyscraper technique should be built to achieve both.
What AI Mode systems look for in a citable source:
What AI Mode systems look for in a citable source:
A skyscraper piece designed for the AI era should be written with these AI citation signals in mind from the start, not added as an afterthought.
The Gap Analysis Template
The gap analysis is the most important prep step and the one most commonly skipped. Here is the process in detail:
| Step | Action | Output |
|---|---|---|
| Identify target keyword | Choose the primary keyword and check current top 5 ranking pages | List of 5 competitor URLs to analyse |
| Audit each page | For each competitor page, record: H2 sections covered, data cited, examples used, tools referenced, date of last update, word count, author credibility | Competitor coverage map |
| Map the gaps | Identify sections, angles, or data points that appear in zero or fewer than 3 of the 5 competitor pages | Gap list |
| Check SERP features | Note which SERP features appear for this keyword: featured snippet, People Also Ask, AI Mode panel, video. Each feature is a formatting target. | Format targets list |
| Identify data opportunities | Is there original data you can generate (survey, analysis, experiment) that no existing page includes? | Research brief if applicable |
| Outline your piece | Build the outline including all high-coverage sections (so you are complete) plus all gap sections (so you are better). Assign format targets. | Content outline |
Common Skyscraper Mistakes
Targeting the Wrong Pages
Not every highly linked page is a good skyscraper target. Pages that are highly linked because they are authoritative primary sources (government data, original research, platform documentation) are not replaceable by a blog post. Target pages where the link profile is based on being the best available guide, not on being an irreplaceable source.
Writing More Without Writing Better
Word count is not a differentiator. A 7,000-word guide that says the same things as a 3,000-word guide in more verbose prose is not better. Better means genuinely more useful, more accurate, more current, or more specific. If you cannot articulate what specifically is better about your piece versus the target, you do not have a skyscraper. You have a longer blog post.
Ignoring the Outreach Quality
Content quality and outreach quality are both necessary. Great content with weak outreach earns fewer links than it deserves. Weak content with great outreach earns links temporarily, until editors and AI systems devalue it. Both need to be strong.
Treating Skyscraper as a One-Time Campaign
The most successful skyscraper content is maintained over time. Updating the data annually, adding new case studies, and keeping the page current maintains its link earning momentum and prevents it from becoming the outdated resource that someone else uses for their next skyscraper campaign.
Not Promoting Beyond Outreach
Outreach to existing link sources is one channel. New links also come from promoting the piece to your newsletter, social channels, relevant communities, and PR contacts. A skyscraper campaign that only does backlink outreach and nothing else misses a significant portion of the link earning potential.
The Skyscraper Technique and Content ROI
Measuring the return on a skyscraper project is important both for justifying the investment and for optimising future campaigns. The metrics to track: referring domain count (primary success metric for link building campaigns), organic traffic to the target page, keyword rankings for the primary and secondary keywords targeted, and citations in AI Mode responses. For a broader framework on tracking content marketing performance, the guide on measuring content ROI covers the full KPI set.
A realistic skyscraper project timeline:
A realistic skyscraper project timeline:
Where Skyscraper Fits in Your Content Strategy?
Not every piece of content should be a skyscraper. The technique is resource-intensive and works best for pillar pages and cornerstone content that sits at the centre of a topic cluster. Spoke content in a cluster serves a different purpose: depth on specific subtopics, targeting long tail keywords, and supporting the pillar with internal links. Skyscraper logic applies to the pillar, not every spoke. The B2B content marketing strategy covers how to structure the overall content programme that skyscraper pieces feed into.
A content programme that runs well typically has:
A content programme that runs well typically has:
For clients working with us on content marketing services, the skyscraper identification and production process is part of the quarterly planning cycle. We audit the competitive landscape, identify the highest-leverage targets, and build the gap analysis before any content is commissioned. The investment in research upfront is what separates a skyscraper that earns 30 links from one that earns 3.
What Still Works, What Does Not?
| Skyscraper Element | Status in AI era | Why |
|---|---|---|
| Use links for updates to identify content worth targeting | Still works | Backlink data remains the best signal for what earns editorial endorsement |
| Building more comprehensive content | Partially works | Content can be better and more complete, but differentiation matters; not just more content |
| Outreach to existing link holders | Still works, harder to execute | Works when the email is specific and the content has visible differentiation |
| Generic "we built something better" outreach | Does not work | Too common, too easy to ignore; editors have seen this for a decade |
| Word count as the primary differentiator | Does not work | AI content has flooded every topic with long form; length is not a signal of quality |
| Adding original research or data | Works very well | Genuinely novel data cannot be replicated; it is the strongest differentiator available |
| Maintaining and updating the content over time | Works very well | Ongoing freshness sustains link earning and AI citation rates |
| Targeting recently outdated content on changing topics | Works very well | Clear and specific differentiation based on currency and accuracy |
The Underlying Principle Has Not Changed
The reason the skyscraper technique worked in 2013 is the same reason the modernized version works today: if you create the most genuinely useful resource on a topic and put it in front of the people who link to resources on that topic, some of them will update their links.
What has changed is the standard for “most genuinely useful.” In 2013, being more complete was enough. Today, you need to be more complete, more current, more authoritative, and more specifically credible. Original data, practitioner-level expertise, and content designed for both human readers and AI extraction systems.
The work is harder. The payoff is larger. A well-executed skyscraper piece today, built on original research and maintained over time, is one of the highest-ROI content investments you can make. The entries are few because the barrier is high. That is exactly what makes it worth doing.
What has changed is the standard for “most genuinely useful.” In 2013, being more complete was enough. Today, you need to be more complete, more current, more authoritative, and more specifically credible. Original data, practitioner-level expertise, and content designed for both human readers and AI extraction systems.
The work is harder. The payoff is larger. A well-executed skyscraper piece today, built on original research and maintained over time, is one of the highest-ROI content investments you can make. The entries are few because the barrier is high. That is exactly what makes it worth doing.