Updated 4 September 2026 · 9 minute read · 3 sources cited
The number that decides everything
Before any template, the reply-rate data is worth sitting with, because it reframes what you are optimising.
Respona measured 1,000 outreach emails across four guest-posting campaigns and got 205 responses — a 20.5% response rate. Separately, highly tailored pitches are reported at around 19% reply rate against 3% to 5% for less personalised approaches, and guest post outreach personalised to the specific publication at 11.7%. General B2B cold email averages 3.43%.
So the spread between a decent pitch and a bad one is roughly four to six times. No template closes that gap. What closes it is having read the site.
The difference between a 4% campaign and a 19% one is not the wording. It is whether you read the site before writing.
What gets binned, from the other side of the desk
We place articles with several hundred publishers, which means we spend a lot of time on the receiving end of editors explaining what they delete unread. The list is consistent and none of it is subtle.
Anything opening with "I came across your website and was really impressed by your content" — the most common first line in the category, and a reliable signal that nobody read anything. Anything offering three topic ideas as a bulleted list, which reads as a mail merge because it usually is. Anything using "high-quality content" or "valuable resource", phrases that appear almost exclusively in outreach and almost never in ordinary writing.
Also binned: a first email that asks about pricing, links or domain rating. Editors at publications worth appearing on think of themselves as running a publication. Leading with the transaction tells them which kind of email this is before they reach the pitch.
The templates
Four, for genuinely different situations. Each carries a line telling you what to change, because the swap is the part that does the work and most template posts leave it out.
- Subject
- Following your piece on {their topic}
- Body
Hi {name},
Your piece on {specific article} made a point I have been arguing with clients about for a year — {the specific point}. {One sentence of genuine response, agreement or disagreement.}
I would like to write the follow-up: "{concrete working title}". It would cover {two or three specifics}, and I would bring {the data, access or experience you actually have}.
About 1,200 words, original, yours to edit. Worth me sending an outline?
{Your name}
{Role, one line, no signature block}
What to change: Everything in braces, and the second sentence has to prove you read past the headline. If you cannot write that sentence honestly, you have not read the article and the editor will know.
- Subject
- A gap in your {topic} coverage
- Body
Hi {name},
You have covered {topic} thoroughly — {name two pieces}. What is not there, as far as I can find, is {the specific gap}.
I would write it: "{working title}". {One line on why you are the person to write it — data you hold, work you have done, not credentials.}
Happy to send an outline first if that is easier than a draft.
{Your name}
What to change: The gap has to be real. Search their site properly first, because naming a gap they filled in 2024 ends the conversation immediately.
- Subject
- {Their article title} is out of date in one place
- Body
Hi {name},
{Article title} still ranks well, but {the specific outdated fact} changed in {when}. {One sentence on what is now true.}
I can send a corrected paragraph you can drop straight in, no strings — or if it is more useful, write the fuller update as a new piece.
Either way the correction is yours.
{Your name}
What to change: Offer the correction unconditionally and mean it. Sending it with a link already embedded is exactly why this approach has the reputation it has.
- Subject
- Re: {original subject}
- Body
Hi {name},
Sending this back to the top in case it got buried.
If the topic is not right, say so and I will not keep asking — and if there is something you do want covering, I would rather write that.
{Your name}
What to change: Nothing. Send it as it is and stop there. The explicit permission to say no is what makes people answer, and a third email converts an ignore into a block.
Subject lines
The subject decides whether any of the above gets read at all. What works is specific and unexcited — it should read like an email from a colleague rather than a campaign.
- Name their article: "Following your piece on X". Highest performing, because it is verifiably not a mail merge.
- State the gap: "A gap in your X coverage". Direct, slightly provocative, works with confident editors.
- Offer the fix: "X is out of date in one place". Reads as helpful rather than as a request.
- Avoid: "Guest post opportunity", "Collaboration proposal", "Quick question", anything with an emoji, and anything promising value.
When you should not pay anyone for this
We sell guest post placements, so the useful thing to be straight about is when doing this yourself beats buying them.
If you need five placements, do it yourself. Five genuinely personalised emails to sites you have actually read will beat five bought placements on relevance, cost nothing but an afternoon, and start relationships you own rather than rent. Nobody should be buying links at that volume.
Outreach stops scaling somewhere around fifteen to twenty placements a month, and it fails for the reason the data above shows: the personalisation producing a 19% reply rate is precisely the thing that cannot be automated. Every tool sold to fix this works by making the emails less personal, which moves you toward the 3% column. That is the actual reason a market for bought placements exists — not that outreach does not work, but that doing it properly does not scale.
Sources
Figures are quoted as the studies above report them. Where they disagree, the text gives the range.