Link building outreach has a bad reputation because most of it is terrible. The median outreach email is a mail-merge with a first name in it, and it is deleted in two seconds.
The approaches below work because they offer something, and even then the numbers are humbling.
Realistic expectations
For cold outreach to sites with no existing relationship, a reply rate in the range of 5–15% is typical for well-targeted, personalised messages. Link placement rates are lower again — frequently 2–5% of emails sent.
That means a campaign targeting 20 links needs several hundred well-chosen prospects, properly researched. Anyone promising dramatically better numbers is either buying links, counting low-quality placements, or working with a relationship base rather than cold contacts.
Knowing this matters for planning. Outreach is a volume activity with a quality constraint, and underestimating the volume is why most campaigns quietly stop after two weeks.
Before writing anything: the asset
Outreach fails most often because there is nothing worth linking to. No message rescues a mediocre blog post.
What earns links is content that gives someone a reason to cite it: original data from your own operations or a survey, a genuinely comprehensive reference that saves the reader work, a free tool, or a clear visual explanation of something complicated.
If you cannot say in one sentence why a writer would cite this instead of what they are already citing, the problem is not the email.
Approach 1: Broken link replacement
Find a page on a relevant site linking to something that no longer exists, and offer your equivalent as a replacement. This works because you are reporting a real problem on their page.
A workable message is short and specific: name the page, name the broken link and where it sits on the page, mention that the destination is dead, and offer your resource as one option while making clear you are not insisting. Sign off without a follow-up threat.
The reason this performs better than most approaches is that the first half of the message is useful whether or not they use your link. Reply rates of 10–20% are achievable where the broken link is genuinely relevant.
Approach 2: Unlinked mentions
Someone has mentioned your company, product or research without linking. Set up alerts for your brand name and check monthly.
The message is a light request: thank them for the mention, note the article, and ask whether they would be willing to add a link so readers can find the source. Nothing more.
This is the highest-converting outreach there is, often 30–50%, because the writer has already decided you are worth mentioning. The limiting factor is volume — most companies have few unlinked mentions, and the technique does not scale.
Approach 3: Resource page suggestions
Pages curating useful links on a topic. The pitch is straightforward: you have something that fits their list.
What makes the difference is demonstrating you read the page. Reference a specific entry on it, explain in one line what yours adds that the existing entries do not, and keep the whole message under 100 words.
Reply rates here are lower, often 5–10%, because these pages receive a great deal of outreach. Targeting narrowly — pages genuinely about your specific subtopic rather than any list that mentions your industry — is what lifts it.
What gets messages deleted
- “I came across your website and love your content.” Signals a template immediately.
- No specific reference to anything on their site.
- Length. Over 150 words and the reply rate drops sharply.
- Asking for a link in the first line before establishing any reason.
- Offering a guest post with no topic. “I’d love to write for you” puts the work on them.
- Mentioning payment for a link, which breaches guidelines and in many cases ends the conversation.
- Aggressive follow-ups. One follow-up is reasonable; four is a complaint.
Follow-up
One follow-up, four to seven days later, materially increases replies — often by half again. It should be two sentences, reference the original, and make it genuinely easy to say no.
Stop after that. A second and third follow-up produce diminishing returns and a worsening sender reputation, and they annoy exactly the people you may want to approach again next year.
Finding the right prospects
Targeting matters more than the message, and the pages you point them at matter too — links are worth most when they reach a page at the centre of a topic cluster rather than an isolated post. Four sources that produce better lists than a bulk export.
Sites already linking to comparable resources, found by examining who links to competing pages. People who have written on the topic in the last year, who are demonstrably interested. Journalists and writers who cover your subject specifically. And sites you have an existing relationship with — customers, partners, people you have met — which is the most neglected and highest-converting group of all.
Quality of list beats quality of email. Fifty well-chosen prospects outperform five hundred scraped ones, and take less time in total.
Measuring the campaign
Track emails sent, reply rate, positive reply rate and links placed, by approach and by prospect type. After a hundred emails you will know which approach works for your niche, which is the point of tracking it.
Also track cost per link including your own time. Outreach is labour-intensive, and comparing the real cost against the alternatives — original research that attracts links passively, digital PR, or simply better content — is a calculation most teams never do.
Approaches that avoid outreach entirely
Cold outreach is the most labour-intensive way to earn links, and for many organisations it is not the best use of the hours.
Original data. A survey of your customers, an analysis of your own aggregated usage data, or an index published annually. Journalists and writers need numbers to cite, and there are far fewer sources of original figures than there are of opinion. A single good dataset can attract links for years with no further effort.
Free tools. A calculator, a checker, a template generator. These attract links because people recommend them in forums and articles without being asked.
Being quotable. Responding to journalist requests, speaking at events, publishing a clear position on something contested. This builds a reputation that generates mentions passively.
These take longer to produce results and compound, where outreach stops the moment you stop sending emails. Most sustainable link profiles are built mostly this way with outreach as a supplement.
Judging whether a link is worth having
Not every placement helps, and some actively hurt. Four checks before pursuing a prospect.
- Is the site topically relevant? A link from an unrelated site carries little weight regardless of its metrics.
- Does it have real traffic? Sites with high authority scores and no visitors are frequently link networks.
- Does the page have genuine editorial content, or is it a list of outbound links with no other purpose?
- Would you be happy for a customer to see your brand there? This single question filters out most of the bad prospects faster than any metric.
A smaller number of relevant, trafficked links outperforms a larger number of poor ones, and the poor ones carry a cleanup cost if they ever need disavowing.
Frequently asked questions
What reply rate should I expect?
Five to fifteen percent for well-targeted cold outreach. Unlinked mention requests run far higher; generic resource page pitches run lower.
Is paying for links worth it?
It breaches search engine guidelines and carries a real risk of a manual action. It also tends to produce links from sites that exist to sell links, which carry little value.
How many follow-ups?
One. The lift from a single polite follow-up is substantial; further messages mostly generate irritation.
Should outreach be automated?
Sending and tracking can be. Research and personalisation cannot, and the moment a message reads as automated the reply rate collapses.
