Backlinks outreach is a numbers game only for people who are bad at it. When the list is right and the pitch is specific, reply rates sit somewhere between 10 and 25 percent. When the list is scraped and the email is a template with a merge field, you are looking at 1 percent and a spam complaint or two.
The work splits cleanly in half. Half is research — deciding who to contact and what you can offer them. Half is writing — saying it in four sentences. Almost nobody splits their time that way, which is why most campaigns underperform.
Start with domains that already link to your competitors
The best prospect list is not “sites in my industry.” It is the set of domains that link to two or more of the sites currently outranking you and do not link to you. Those editors have already demonstrated they will link to a company like yours, about your topic, for readers in your market. Nothing on a scraped list comes close to that signal.
Pull that list with a backlink gap analysis across three to five page-one competitors. SEO Rocket produces a named list of referring domains linking to rivals but not to you, filters out obvious spam, and flags toxic domains so you are not pitching a PBN. Sort by relevance, then authority. Then read at least ten of the pages that actually contain the competitor link — you need to know what kind of mention you are asking for before you ask.
Choose an angle you can defend
Every outreach email is a trade. You are asking for a link, which costs the recipient editorial credibility, so you need to hand back something worth more than the ten minutes it takes them to make the edit.
- Broken link replacement — you found a dead resource on their page and you have a live equivalent. Highest hit rate, because you are fixing their problem first.
- Better resource — their linked source is five years old, or is a 900-word summary where you published original data. Works when the improvement is obvious in one sentence.
- Unlinked mention — they already named your brand or quoted your research and forgot the link. Nearly free wins; check quarterly.
- Original data — a survey, a benchmark, a teardown of 200 sites. Costs the most to produce and earns links for years.
- Expert contribution — a quote or a paragraph for a piece they are already writing. Small links, fast yes.
What is missing from that list: “we’d love a link,” “we’ve been following your blog,” and anything involving payment. Paying for placement violates Google’s link spam policies regardless of how the invoice is labeled, and it puts every other link you have built under suspicion.
Pick one angle per campaign and stay with it. Mixing five approaches across the same list makes results impossible to read — you will never know whether the 12 percent reply rate came from the broken links or the data study. Run one, measure it, then run the next.
Write the email in four sentences
Subject line: concrete and specific. “Dead link on your SaaS pricing guide” outperforms “Quick question” and “Collaboration opportunity” by a wide margin, because it describes a real thing rather than a category of email.
Then: one sentence saying what you found, with the exact URL. One sentence on the replacement and why it fits. One sentence on who you are, with a credential that matters — a number, a client, a dataset. One sentence closing with an easy out: “Either way, thought you’d want to know about the dead link.” That last line is not politeness; it is what makes the email worth reading even when the answer is no.
Personalize the first sentence and only the first sentence. Personalizing everything takes four times as long and does not raise reply rates. Personalizing nothing gets you filtered.
Follow up once — and then stop
One follow-up, five to seven business days later, on the same thread. Two lines. Roughly a third of the positive replies on a well-run link outreach campaign come from that second email, which is why sending it matters. A third and fourth email add almost nothing and start generating the complaints that get your domain flagged.
Send from a real person’s address at your main domain, not a lookalike domain, and keep daily volume modest — 25 to 40 emails a day from a warmed inbox is a sane ceiling. Blowing past that is how a small site ends up with deliverability problems that take months to fix.
Run it as a pipeline, not a blast
Track prospects through states rather than counting emails sent. A simple sheet with columns for domain, contact, angle, date sent, follow-up date, reply, and outcome will beat most dedicated tools for a solo operator or a small agency.
- Prospect — pulled from the gap analysis, not yet reviewed by a human.
- Qualified — a person has opened the site and confirmed it publishes real editorial content.
- Contacted — first email out, with the angle recorded.
- Followed up — one nudge sent, then closed if silent.
- Won or lost — link live and verified, or archived with a reason.
Review the “lost” column monthly. If most rejections say “we only do sponsored posts,” your list is skewed toward link sellers and needs re-filtering. If most say nothing at all, your pitch is the problem, not the list.
What good results look like
For a small site with a decent asset to point at, 100 well-qualified prospects typically yields 15 to 25 replies and 8 to 15 live links over six to eight weeks. That is a good quarter. Anyone promising 50 links a month at scale is selling placements, not outreach.
Set the target against the weakest competitor on page one, not the market leader. If the site sitting at position ten has 140 referring domains and you have 40, the gap is real but crossable. Comparing yourself to the DR 80 brand at position one produces a number that is both accurate and useless.
Then judge the campaign on movement over eight to twelve weeks, not on a screenshot from Tuesday. Positions wobble two or three places daily as normal noise. Watch the trend line, the count of ranking keywords, and Search Console impressions together — impressions usually move before positions do, and Google’s own data beats any third-party estimate for your own site.
Where tooling helps and where it does not
Software is genuinely good at the research half: finding the link gap, filtering junk domains, flagging toxic ones, estimating how many links stand between you and page one, and putting a directional cost band on that gap so you can budget. SEO Rocket does that part, plus anchor-text gap analysis so you can see whether your profile leans unnaturally commercial compared to the sites you are chasing.
What software should not do is send the emails. Automated sequences to scraped lists are how outreach got its reputation, and the reply rates prove it. Let the tool decide who to contact and why; write the four sentences yourself. It is an hour a day, and it is the only part of link building that has ever reliably worked.
