Link Outreach Emails That Get Replies

You've received this email. Probably this week.

Hi there, I was reading your excellent article on [topic] and found it very insightful! I noticed you linked to [competitor]. I recently published a comprehensive guide on the same subject that your readers would find valuable. Would you consider adding a link?

Deleted in about two seconds. Not because it's rude — because it's obviously templated, offers nothing, and asks a stranger for a favour.

The uncomfortable part is that most people send some version of this, get almost no replies, and conclude outreach doesn't work. It does work. Just not at volume, and not like that.

The short version

Low volume, high relevance. Twenty carefully chosen recipients beats two thousand scraped ones.

Have a genuine reason beyond wanting a link.

Lead with what they get.

Never send from your primary domain. Cold outreach at volume from near-identical templates is exactly what spam filters are built to catch, and it can wreck the sending reputation your actual business depends on.

Why most outreach fails

Four structural reasons, and they compound.

The recipient owes you nothing. You're asking a stranger to edit their published work, for your benefit, with nothing offered back. Framed honestly, it's a strange request.

The template is visible. People who publish online receive dozens of these. The patterns are unmistakable — "I was reading your excellent article," "I noticed you linked to," "would provide value to your readers." Those phrases are now a signal to delete.

There's no genuine reason. "I wrote something similar" isn't a reason for anyone to act. It's a reason for you to act.

Volume made it worse. The tools that let you send two thousand emails also guaranteed everyone receives hundreds. The channel is saturated, and saturation is why personalisation stopped being optional.

What a reply-worthy pitch has

A specific, verifiable reason to write to this person. Not "I read your article" — something demonstrating you actually did. Reference the specific claim, the specific section, the specific thing they got right that others don't.

Something they gain. Their link is broken and you have a working replacement. Their piece cites a statistic that's since been superseded and you have the current one. You've produced data that strengthens an argument they made.

Notice these are all versions of "here's something useful for your page" rather than "here's something useful for me."

No ask, or a very light one. Counterintuitive and it works. "Thought you'd want to know that link is dead — here's a working alternative if useful" converts better than "please link to me," because it isn't a request. Some recipients link anyway. Some don't and you've still built a small amount of goodwill.

Brevity. Under 100 words. A long outreach email is a wall, and walls don't get read.

Evidence you're a real person. A real name, a real site, a real signature. No pseudonyms, no "Marketing Team."

Three that actually work

1 — The broken link

The highest-converting version, because you're genuinely helping.

Subject: dead link on your [topic] page

Hi [name],

Your [specific article] links to [resource] in the section on [specific thing] — that URL is 404ing now.

[Alternative resource] covers the same ground if you need a replacement. I also wrote something on it here: [your link], though the first one may suit better.

Either way, thought you'd want to know.

[Your name], [yoursite.com]

Note that it offers a non-self-serving option first. That single move changes how the whole email reads.

2 — The updated statistic

Subject: the [X] figure in your [topic] piece

Hi [name],

Your piece on [topic] cites [statistic] from [source, year]. That figure's been superseded — [updated source] now puts it at [new figure].

I've got the current numbers compiled here if useful: [link].

[Your name], [yoursite.com]

Works because factual accuracy is something writers genuinely care about.

3 — The original data offer

The strongest, and it requires having done the work.

Subject: data on [topic] you might want

Hi [name],

You made the point in [article] that [their argument]. I audited [N] [things] recently and found [specific number] — which supports it with actual data.

Full method and dataset here if it's useful: [link]. No obligation either way.

[Your name], [yoursite.com]

You're offering evidence for something they already believe. That's a genuinely different proposition from asking for a link. Producing the data.

The deliverability problem nobody mentions

This is the part that costs people something they didn't expect to lose.

Cold outreach at volume, from near-identical templates, sent to people with no prior relationship, is precisely the pattern spam filters are built to detect. Do it from your primary domain and you can damage the sending reputation your invoices, password resets, and newsletter depend on.

The rules:

Use a separate sending domain. Not yourbrand.com. Something like yourbrand-outreach.com, with its own SPF, DKIM, and DMARC records.

Warm it up. A brand-new domain sending fifty cold emails on day one looks exactly like what it looks like.

Keep volume low. Which you should be doing anyway, since relevance beats volume here.

Vary the message genuinely. Not merge-tag variation — actual different sentences. Filters fingerprint near-identical messages across many sends.

Be sceptical of "warm-up networks." Bots opening and replying to each other to fake engagement is a detectable pattern, and providers have acted against it.

The full authentication setup. · Why cold sending is different.

What reply rate is normal?

Honestly: highly variable, and anyone quoting you a confident number is guessing.

It depends almost entirely on list quality and personalisation, and published figures range from under 1% to double digits. Mass templated outreach sits at the bottom of that range. Twenty hand-written emails with a genuine reason can run far higher.

The more useful measure is links per hour spent. Two thousand templated emails producing four bad links is worse than twenty considered emails producing two good ones — before you count the deliverability damage.

Should you follow up?

Once. Maybe.

A single short follow-up after about a week is acceptable — genuinely brief, and adding something rather than repeating.

Beyond that you're pestering someone who has decided not to reply, and repeated follow-ups to non-responders is one of the fastest routes to complaints, which hurt your sending reputation directly.

The tell that you've gone too far: if your follow-up says "just bumping this in case you missed it," you're following up for your benefit, not theirs. Don't send it.

Frequently asked questions

What reply rate is normal for link outreach?
It varies enormously and depends almost entirely on list quality and personalisation, so treat confident numbers sceptically. Mass templated outreach sits at the very bottom of published ranges; small batches of genuinely researched, personalised emails perform far better. The more useful metric is links earned per hour invested, which usually favours twenty considered emails over two thousand templated ones.

Should I follow up on outreach emails?
Once, briefly, after about a week — and only if you can add something rather than repeat yourself. Beyond that you're pestering someone who has decided not to reply, and repeated follow-ups to non-responders generate spam complaints that damage your sending reputation. If your follow-up is "just bumping this," it's for your benefit, not theirs.

Does outreach hurt email deliverability?
It can, badly. Cold outreach at volume from near-identical templates to recipients with no prior relationship is exactly the pattern spam filters detect. Sending it from your primary domain risks the reputation your invoices, password resets, and newsletter depend on. Use a separate sending domain with its own authentication, warm it properly, keep volume low, and vary messages genuinely rather than with merge tags.

What to do next

Before writing any outreach, answer one question: what do I have that this specific person would want?

If the answer is "a link to my article," don't send it. That's not a reason for them to act, and it's why the two-second delete exists.

If the answer is a broken link on their page, a superseded statistic, or data supporting an argument they've made — write it. Those get replies, and you'll need twenty of them rather than two thousand.

Free: The SEO audit checklist.

---

Related guides

• Link building without buying links — where outreach fits, and where it doesn't
• 7 linkable assets — building something worth pitching
• Digital PR without a budget — the higher-return alternative
• Email authentication — protecting your sending domain