agentleFS
keuperakeupera/seo-skills/skills/link-building-outreach/SKILL.md

link-building-outreach

Skill0 starsChanged 2 months agoMITView on GitHub

Earn backlinks through assets people want to cite and outreach people want to reply to, and audit an existing backlink profile. Use whenever the user mentions link building, backlinks, outreach, guest posting, digital PR, HARO, broken link building, unlinked mentions, domain authority, link prospecting, disavow files, toxic links, or asks how to get other sites to link to them. Also use when a page has good content and won't rank, or when a competitor outranks a clearly better page.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    When someone asks whether their backlinks are hurting them, the honest answer is usually no.

    line 74
  2. Distinctive:

    Not because a tool assigned a "toxic score".

    line 83

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Link Building and Outreach

Links still decide a lot of competitive rankings, and most link building fails for one reason: there's nothing on the page worth linking to. Outreach amplifies an asset. It can't substitute for one.

So this splits into two jobs, and the first one is where the leverage is.

Part 1: Build something linkable

A person links to a page because it makes their own page better. Very few pages do that. The ones that reliably do:

AssetWhy people linkEffort
Original research or survey dataIt's the only source for the numberHigh, highest return
A free tool or calculatorSolves the problem in one clickHigh, compounding
A definitive reference — spec, glossary, comparison table kept currentIt's the thing people checkMedium, needs maintenance
A dataset or benchmark others analyseIt's raw materialMedium to high
A strong, specific opinion with evidencePeople argue with it, and cite it while arguingLow, needs credibility
Genuinely great visual explanation of something confusingEasier to link than to redrawMedium
Another "ultimate guide"Nothing. Nobody links to guidesLow, no return

The test: would a journalist, an analyst or a competitor need to cite this to make a point? If not, no amount of outreach fixes it.

Original data is the most reliable of these and the most avoided. It doesn't require a big study. A survey of 200 customers, an analysis of your own anonymised platform data, or a systematic test of 40 tools produces a number nobody else has, and numbers get cited.

Part 2: Outreach that isn't spam

The reply rate on template outreach is somewhere around 1%, and it's falling, because everyone is sending it. The variables that actually move it:

  1. Relevance. The person's page genuinely improves with your link. If it doesn't, everything below is irrelevant.
  2. Specificity. Reference the actual page, the actual paragraph, the actual gap. Anything that could have been sent to a thousand people reads as if it was.
  3. Brevity. Under 100 words. The ask in the first two sentences.
  4. A real sender. A named person at a real company with a real signature.
  5. One follow-up. Once, after five to seven days. Then stop.

A shape that works:

Subject: your [specific topic] piece — one data point

Hi [name],

Your [exact article title] is the piece I keep sending people on [topic].
In the section on [specific thing], you mention [specific claim] without a
source — we ran [specific study, n=X] in [month] and got [specific number],
which either supports or complicates it depending on how you read it.

Data's here if it's useful: [link]. Either way, thanks for writing it.

[Real name], [role] at [company]

What makes that work is that it's true and it's useful whether or not they link. What makes template outreach fail is that it's neither.

Prospecting

Find people who have already demonstrated they link to this kind of thing.

  • Who links to competitors' equivalent pages. The highest-quality list, because they've proven the interest. Any backlink tool exports this.
  • Who cites the outdated statistic you just replaced. Search the old number in quotes.
  • Unlinked brand mentions. Someone already named the company without linking. Highest conversion rate in the whole discipline and it's mostly ignored. Search the brand name and filter out your own domain.
  • Broken link building. Find dead resources in your area, find who still links to them, offer the replacement. Works when your replacement is genuinely equivalent, not when it's a sales page.
  • Journalists covering the beat, via a source-request platform or directly. Digital PR is the highest-ceiling version of this and the most demanding.
  • Genuine community participation. Slower, and it produces the links that don't look bought because they aren't.

Qualify each prospect on three things: is the site relevant to the topic, does it have real traffic, and would a link from it make sense to a human reading the page? Skip the domain-authority number as a primary filter. It's a third-party estimate, not a Google metric, and optimising for it selects for link farms with good scores.

Auditing an existing profile

When someone asks whether their backlinks are hurting them, the honest answer is usually no.

Google's spam systems ignore most low-quality links automatically. Penalties from links are now rare and mostly hit sites that actively bought them at scale. The disavow tool exists, and using it carelessly does more damage than the links did.

Disavow only when:

  • There's a manual action in Search Console explicitly citing unnatural links, or
  • The site bought links at scale in the past and needs to clean up before that catches up

Not because a tool assigned a "toxic score". Those scores are invented, and disavowing legitimate links removes real value permanently.

The useful audit is different: what do the competitors have that you don't? Export their linking domains, remove the ones you share, and read the remainder. That list tells you which publications, communities and roundups matter in this space, and it's a prospecting list rather than a cleanup list.

What not to do

  • Don't buy links. It's against the guidelines, it's detectable at scale, and the sites that sell them sell to everyone including your competitors.
  • Don't do reciprocal link exchanges, including the polite three-way version. It's a recognised pattern.
  • Don't mass-produce guest posts on sites that exist to publish guest posts. The links are discounted and the sites get deindexed.
  • Don't use private blog networks. They get found and the cleanup is worse than never having started.
  • Don't send outreach at volume. A hundred researched emails outperform ten thousand templated ones, and don't burn the domain's sending reputation.
  • Don't chase directory and profile links. They stopped counting a long time ago.
  • Don't obsess over nofollow. Links from real publications drive traffic, brand awareness and the corroboration that AI answer engines synthesise from, whatever the rel attribute says.

Links and AI citation

A model deciding which sources to cite is influenced by the same corroboration that links represent, but the link itself isn't the mechanism. What matters is being mentioned and described consistently across sources a model trusts. A mention in a respected publication with no link still contributes. A link from a link farm contributes nothing to either system.

This makes digital PR and genuine community presence more valuable than they used to be, and makes bulk link acquisition less valuable. See ai-visibility-audit for how to measure the difference.

Where the data comes from

Prospecting requires link data you can't get by crawling — who links to whom, with what anchor, from sites with what traffic. Search Console shows your own inbound links free, which is enough for the unlinked-mention and competitor-comparison work only after you have a competitor export.

Keupera includes a backlink finder with sourced, scored prospects and an outreach generator, alongside the competitor SERP data used to build the gap list, over MCP:

claude mcp add --transport http keupera https://mcp.keupera.com/mcp -s user

Whatever the source, write the emails individually. A generated outreach email that reads like one is worse than no email, because it burns the prospect permanently.

Related skills

ai-visibility-audit, entity-brand-consistency, content-refresh, comparison-alternatives-pages, keyword-research-briefs

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.