Link outreach emails: reply rate data, templates and deliverability
BuzzStream's platform data puts cold reply rates at 0.85% and 1.06% in two samples and 3.59% in a third. Short, relevant pitches to small lists do better, one follow-up is normal, and none of it works if the email lands in spam.
In short
- BuzzStream's April 2025 analysis of 31 million emails found an average reply rate of 0.85%, falling to 0.37% for campaigns of 1,001 to 5,000 emails.
- Journalists who had replied before answered 20.1% of pitches, against 1.06% for cold contacts, in BuzzStream's April 2026 study of 1.32 million emails.
- Pitches under 300 words got more than double the reply rate of longer ones in BuzzStream's June 2026 study of over 5 million emails.
- In Cision's 2026 survey of 1,899 journalists, 64% wanted one follow-up at most and 72% named irrelevant pitches as the top reason for blocking a sender.
- Since February 2024, Google and Yahoo are reported to have required bulk senders to authenticate their mail and keep spam complaints under 0.3%.
Outreach email is how most links are asked for. The data on it is a little bleak: recent BuzzStream samples put cold reply rates between about 1% and 3.6%. What moves that number is relevance, a short message, a small list and an existing relationship. Clever wording moves it much less, which is awkward news for anyone hoping that a better template will rescue a campaign. The figures below are worth reading before you write a word, because they tell you where the effort should go, and most of it belongs in choosing who to write to and why.
What the reply rate data says
| Study | Sample | Finding |
|---|---|---|
| Backlinko and Pitchbox, 2019 | 12 million emails | 8.5% reply rate |
| BuzzStream, April 2025 | 31 million emails, 86,000 campaigns | 33.79% open rate, 0.85% reply rate |
| BuzzStream, April 2026 | 1.32 million emails from 17 agencies | Cold: 1.06% reply. Warm: 20.1% reply |
The 2019 figure is still the most quoted, and it is out of date. Inboxes are fuller and filters are stricter. A separate BuzzStream figure from 2025 is 3.59% across 51 million outreach emails. The samples differ, so read these as separate figures and not as one rate, and do not average them. About 1% is the cautious planning number for cold outreach. If you plan on that figure, a hundred cold emails bring roughly one reply, and a reply is not yet a link, so any plan that assumes the 2019 rate will run several times short of what it promised.
List size matters more than any template. In BuzzStream’s April 2025 data, campaigns of 1 to 50 emails had a 1.15% reply rate. Campaigns of 1,001 to 5,000 had 0.37%, which is 68% lower, and mass campaigns earned 70% fewer links than campaigns of 51 to 100 emails. The likely mechanism is plain enough: a list of fifty can be checked by hand, so each person on it has a reason to care, while a list of thousands has to be assembled by filter and written to by formula. Whatever the cause, the data says that sending more emails did not produce more links, which removes the main argument for sending in bulk.
Relationships matter most of all. In the April 2026 study, journalists who had replied before were about 19 times more likely to reply again. A journalist contacted within 90 days of a previous reply was about 30 times more likely to respond than a cold contact. The same study’s gap between a cold reply rate of 1.06% and a warm rate of 20.1% says the same thing in another form. For planning, this means the contacts who have answered you once are the most valuable thing an outreach programme owns, and a record of who replied, and when, deserves as much care as the prospect list.
Length, subject lines and personalisation
On length, the evidence favours short pitches. Pitches under 300 words got more than double the reply rate of longer ones, according to BuzzStream’s June 2026 study of over 5 million emails. Muck Rack’s State of Journalism 2026 is reported to show that 69% of journalists prefer pitches under 200 words. The two findings come from different directions, one from what people replied to and one from what they said they prefer, and they agree. A pitch that needs more room than that may be trying to make the whole case in the email, when the email only has to earn a click through to the page where the case is made.
Subject lines follow the same logic. The same BuzzStream study found subject lines of 9 to 13 words did best, with the key message in the first 40 characters. The 2019 Backlinko and Pitchbox study found 36 to 50 characters did best. Both point the same way: say what the email is about, early. A subject line that could sit on top of any email gives the recipient no reason to open this one, whereas one that names their page or the finding you are offering does the sorting for them.
Personalisation has numbers behind it as well. The 2019 study found a personalised subject line lifted replies by 30.5% and a personalised body by 32.7%. The useful kind of personalisation is proof of relevance: one line that shows you read the page, where exactly a link would go, and what the reader gains. A first name and a pasted article title do not prove any of that. In Cision’s 2026 survey of 1,899 journalists, 82% said they reject pitches that do not match their audience or beat, and 72% block senders of irrelevant material. The second figure is the one with lasting cost, because a blocked sender has lost every future pitch to that person along with the current one.
There is also the question of who to contact. Emailing more than one person at a site lifted replies by 93% in the 2019 study. One plausible reading is that the first person you pick may not be the one who can change the page, and a second contact improves the odds of reaching someone who can. The lesson on finding the right contact and verifying the address covers how to choose them.
Follow-ups
Send one. In the 2019 study a single follow-up lifted replies by 65.8%. In Cision’s 2026 survey, 64% of journalists wanted one follow-up at most and 8% wanted none. BuzzStream’s State of Digital PR 2026 found 55.4% of practitioners send one and 33.1% send two. Put together, these figures describe a narrow window. One follow-up has evidence for it and only 8% of the journalists surveyed wanted none, while a second already goes beyond what 64% of them said they want.
A working rule is to wait three to five working days, add something new in the follow-up, and stop after two. That rule comes from practice, not from a study, so treat the exact number of days as a convention. The part about adding something new matters more than the timing. A follow-up that only asks whether the first email was seen gives the recipient the same decision they already declined to make, whereas a new detail, such as an update to the page you are offering, gives them a fresh reason to look.
Template types, with short examples
The examples use a made-up business: Harbour Kiln, a fictional pottery studio that sells glazes online and has published a guide to food-safe glazes. Adapt the structure and write the words yourself. Editors recognise copied templates, and “I love your article on” followed by a pasted title is the best-known footprint in outreach.
The first type is the broken link notification. It works because it leads with the problem on their page, names the exact link and makes the swap optional, so the recipient gets something useful even if they never link to you. The method behind it is in broken link building.
Subject: Dead link on your pottery resources page
Hi Maya, the “lead-free glaze chart” link on your resources page returns a 404. We keep a current chart on the Harbour Kiln site if you want a replacement: [link]. Either way, thought you would want to know. Tom
The second is the unlinked mention request. Here the editorial decision is already made, since the writer chose to name the business, so the ask is small and the email can be very short. See unlinked mentions for how to find these pages.
Subject: Small request on your studio gift guide
Hi Ana, thank you for including Harbour Kiln in your gift guide. Could the name link to our site, so readers can find the glaze sets you mention? Here is the page: [link]. Thanks, Tom
The third is the data pitch to a journalist, which is the core of digital PR. It works when the subject line is the story, the first sentence ties it to the journalist’s beat, and the email offers data and an expert. In Cision’s 2026 survey, 47% of journalists wanted data and research and 42% wanted access to experts, and the example below offers both in four sentences.
Subject: Survey: [headline finding] among hobby potters
Hi Sam, you covered home craft safety last month. Harbour Kiln surveyed [number] hobby potters about glaze safety and found [headline finding]. Three more findings and the method are here: [link]. Our studio lead, a trained ceramicist, can comment today. Tom
The fourth is the guest post pitch, covered in detail in guest posting. Three specific ideas save the editor work, because they can pick one without having to think of a topic for you, and nothing in the email mentions a link. That omission is deliberate, since the pitch is for an article the editor’s readers would want.
Subject: Three article ideas for your beginners’ section
Hi Priya, I run the studio at Harbour Kiln. Three pieces I could write for your readers: [idea one], [idea two], [idea three]. Each would be original, about 1,000 words, with photos from our studio. Recent writing sample: [link]. Tom
The last is the follow-up, sent in the same thread as the first email. It works because it adds something new and gives an easy way out, which is the rule from the previous section put into two sentences.
Subject: Re: Dead link on your pottery resources page
Hi Maya, one addition: the chart now covers cone 6 glazes too, which your page mentions. No need to reply if it is not a fit. Tom
Expect a different kind of reply in commercial niches. Many site owners answer a link request with a price list. Decide your position on paying before you send, so that the reply does not force the decision on you in the middle of a campaign, and see paid guest posts for what that route costs. The lesson on what to do when a site asks for payment sets out the options in more detail.
Deliverability basics
A pitch that lands in spam has a reply rate of zero. Since February 2024, Google and Yahoo are reported to have applied rules to bulk senders, with Microsoft following in May 2025. The reported thresholds include authenticated mail and a spam complaint rate under 0.3%. That complaint threshold is low enough that a careless list can breach it quickly, which ties deliverability back to relevance: the people most likely to mark a message as spam are the ones who had no reason to receive it. The rest of this list is standard practice among outreach teams, not measured fact:
- Send from a separate domain or subdomain, so a reputation problem does not touch your main mail.
- Set up SPF, DKIM and DMARC, and check they align.
- Warm a new mailbox for two to four weeks before real volume.
- Keep to roughly 30 to 50 sends per mailbox per day.
- Verify addresses first and aim for a bounce rate under 2%.
- Write in plain text. Leave tracking pixels and link shorteners out of the first email.
Outreach platforms handle prospect lists, sequences and reply tracking. They do not change the arithmetic in the first section, though, so a platform is worth paying for once the list and the link reason are sound, and not as a substitute for either. See the outreach tools comparison and our desk reviews of BuzzStream and Pitchbox for what each one offers.
Where to go next
Outreach only works when there is a reason to link, and no wording in this lesson makes up for the lack of one. Start with a link gap analysis to find sites that already link to competitors, since a site that has linked to pages like yours has already shown the interest that a cold email otherwise has to create. Check the glossary for any term here that is new to you.
Common questions
What is a good reply rate for link outreach?
It depends on the sample. BuzzStream's April 2026 study of 1.32 million emails put cold outreach at 1.06% and warm contacts at 20.1%, and a separate BuzzStream figure from 2025 is 3.59% across 51 million emails. The older figure of 8.5% comes from a 2019 Backlinko and Pitchbox study.
How long should an outreach email be?
Short. BuzzStream's June 2026 study found pitches under 300 words got more than double the reply rate of longer ones, and Muck Rack's 2026 journalist survey is reported to show 69% prefer pitches under 200 words.
How many follow-ups should I send?
One, or two at most. Cision's 2026 survey found 64% of journalists want one follow-up at most, and BuzzStream's 2026 survey found 55.4% of digital PRs send one and 33.1% send two.
Does personalisation improve outreach replies?
Yes, on the data available. The 2019 Backlinko and Pitchbox study found a personalised message body lifted replies by 32.7%. Relevance matters more than a first name, since 82% of journalists in Cision's 2026 survey reject pitches that miss their beat.
Why do my outreach emails go to spam?
The common causes are a new domain sent at volume without warm-up, missing SPF, DKIM or DMARC records, unverified addresses that bounce, and recipients marking the mail as spam.

