Cold email that gets replies, and the three mistakes that guarantee it will not

Reply rate is not a copywriting problem. It is a list problem, a relevance problem and a timing problem, roughly in that order.

Everyone wants a better cold email template. Almost nobody has a template problem.

Reply rate is decided in this order: who you sent it to, whether it was relevant to them, when it arrived, and only then what it said. Most agencies work that list backwards, rewriting subject lines for an audience that was never going to answer.

The list is most of the work

A hundred well-chosen prospects will outperform two thousand scraped ones, and it is not close. The difference is that with a hundred you can say something specific to each, and specificity is the only thing that separates you from the eleven other emails they got that morning.

Before a name goes on the list, it should pass three tests:

• They plausibly have the problem. Not "they are a business". A reason to think this specific one is struggling with this specific thing.
• The person can act. An email to an address nobody senior reads is a well-written email into a void.
• You can say why them. If you cannot finish "I am writing to you because...", they do not go on the list.

That third one removes most names. That is the point.

What actually goes in the email

Four sentences. It is not a rule handed down from anywhere, it is just what fits on a phone screen before someone decides.

1. Why them, specifically. Something true about their business that you could not have said about anyone else. This is the whole email. Everything after it is logistics.
2. The problem, in their words. Not your service. The symptom they would recognise from their own week.
3. One line of evidence. A result, a number, a comparable client. One. A list reads as a brochure.
4. A small ask. "Worth a conversation?" outperforms a calendar link, because a link is a commitment to a slot and a question is a commitment to a yes.

No attachment. No deck. Nothing that needs a desktop to open.

Follow up, but say something

The follow-up is where most of the replies actually come from, and where most agencies waste the opportunity by sending "just bumping this up".

That message says: I have nothing to add, but I would like you to feel guilty. Nobody has ever replied to it happily.

Each follow-up should carry one new thing. Another angle on the problem, a relevant piece of work, something that happened in their market. Three is usually the limit before you are the reason they dread their inbox.

Leave four or five working days between them. Same-day and next-day follow-ups read as automated, because they are.

The three that guarantee failure

Personalisation that is obviously a field. "I loved your recent post about {{topic}}" is worse than no personalisation, because it proves the rest is automated too. If the personal line could have been generated, cut it.

Talking about yourself first. "We are a full-service agency founded in 2019" answers a question nobody asked. Your credentials are interesting after someone cares, and not one second before.

Asking for thirty minutes from a stranger. Thirty minutes is a real cost to a person with a calendar that is already full. The first ask should be a reply, not a meeting. Get the conversation and the meeting follows on its own.

What to measure

Not open rate. It has been unreliable since image-proxy privacy features became standard, and optimising a number you cannot trust is worse than having no number.

Measure replies, and separate them into interested and not. A campaign with a two per cent reply rate that is mostly "not now, but ask me in the spring" is a good campaign. One with eight per cent that is all "remove me" is a fast way to lose a domain.

And watch the bounce rate. Above about three per cent, stop the campaign and fix the list. That is not a performance problem, it is a deliverability problem, and it follows you to every future campaign you send.