There isn't an honest universal answer to "what is a good cold email reply rate," because it depends on who you write to, what you offer and how tight your list is. Any figure quoted without those details is a guess. The useful move is to measure your own: replies divided by emails delivered, positive replies divided by emails delivered, and meetings booked per 100 emails sent, then compare this week to last week.
That's the whole method, and it takes a spreadsheet and ten minutes. Below is how to count each number, a worked example, and what to do when the numbers are low.
Reply rate versus open rate
Open rate counts how many emails were displayed. Reply rate counts how many produced an answer. For cold email the second is the one that matters, because a reply is what leads to a meeting. And open rate has a measurement problem: it relies on a tiny image loading, and many mail apps load images automatically, which can record an open when no one has looked. Others block images, so real opens go uncounted. The full argument is in why to stop measuring opens.
If you've turned off tracking pixels, you won't have an open rate at all. That's fine. You never needed one.
How to count replies
Three rules keep the number honest.
- Divide by delivered emails, not sent. If you sent 240 and 6 bounced, your base is 234. Emails that never arrived can't be answered.
- Count human replies only. Out-of-office messages, ticket-number auto-responders and bounce notices are not replies. Sort them first (see cold email auto-replies and out-of-office).
- Count each person once, even if they replied three times.
Reply rate = human replies divided by delivered emails, times 100. That's it.
Positive replies and meetings booked
A reply isn't always good news. "Not interested" and "remove me" are replies. So track a second number: positive replies, meaning anyone who asked a question, wanted more information, said "tell me more" or agreed to talk. Divide positives by delivered.
The number that matters most is the third one: meetings booked per 100 emails sent. This is what turns your effort into customers. Use sent, not delivered, for this one, because it measures your effort in and the outcome out.
Carmen sells commercial landscaping maintenance and sent three weekly batches of 80 emails. Here's her sheet.
- Week 1: 80 sent, 2 bounced, 78 delivered, 5 human replies, 2 positive, 1 meeting.
- Week 2: 80 sent, 2 bounced, 78 delivered, 7 human replies, 3 positive, 1 meeting.
- Week 3: 80 sent, 2 bounced, 78 delivered, 9 human replies, 3 positive, 1 meeting.
Total: 240 sent, 6 bounced, 234 delivered, 21 human replies, 8 positive, 3 meetings. Her reply rate is 21 divided by 234, which is 9.0 percent. Her positive rate is 8 divided by 234, or 3.4 percent. And she books 3 meetings per 240 sent, which is 1.25 per 100.
Work out your own baseline
After about 100 delivered emails, you have a rough baseline. Before that, the percentages jump around too much to mean anything. Carmen's reply rate by week was 6.4, 9.0 and 11.5 percent, which looks like a steady climb. But the differences are two replies each time. With batches that small, one more or one fewer reply moves the percentage by more than a point.
So treat her figures as a starting line and not a trend. A reasonable approach is to use four weeks or 300 delivered emails as your baseline, then compare future batches against it. The measure that matters is how you compare with your own last month, not how you compare with a number from someone else's business.
Why quoted benchmarks mislead
You will see reply rates quoted in articles and sales decks, often as a single tidy percentage. Be wary. Those numbers usually come from one company's customers, using one kind of list, measured in their own way. Some count auto-replies, some divide by sent instead of delivered, and some only count campaigns that survived long enough to be reported. Two of them can differ by a factor of three and both be technically true.
Carmen's 9.0 percent says nothing about what you will get. Landscaping managers are a narrow, local audience. Someone emailing chief financial officers at software companies might see a very different figure, higher or lower, and neither of them would be doing it wrong.
What carries over from one business to another is the method, not the number. Count the same way every week, keep the denominators consistent, and write down what you changed. Then your own history becomes the benchmark.
What changes the number most
In roughly the order of impact, from experience and without made-up statistics:
- Who you send to. A tight list of the right people beats a clever email to a loose one. This is the biggest lever by far. See the list is not the customer.
- The offer. A relevant, specific offer gets more answers than a general one.
- The first line. Mentioning something true about the reader helps.
- Length and the ask. Short, with one easy question.
- Follow-ups. A sequence of three or four touches gets more replies than a single email, since many answers come after the second message.
- Deliverability. If your mail lands in spam, nothing else matters; check your authentication in SPF, DKIM and DMARC for cold email.
Subject lines and send times matter less than people hope. Test them last.
Small lists give noisy numbers
If you send 50 emails and get 4 replies, that's 8 percent. If you'd gotten 3, it would be 6 percent. A 2-point swing from a single reply is noise, not signal. The smaller your list, the more you should hold conclusions loosely and ask a different question: not "is my rate good?" but "did anyone I'd like to work with say something useful?"
Compare batches of the same size and the same kind of buyer. And resist changing five things at once, since you won't know which one moved the number. How to test a cold email with only 50 prospects goes through how to run a fair small test.
What to change first when it is low
Work from the top of the funnel down.
- If many emails bounced, fix the list before anything else (see cold email bounce rate).
- If delivery looks fine but replies are near zero, send a test to a personal account and check that you land in the inbox, not spam.
- If you're reaching inboxes but nobody answers, read the first line of your email. Is it about them?
- If people answer but say no, the offer or the audience is off. Read the reasons.
- If people answer yes but never book, your ask is too heavy or your follow-up is slow.
The cost side of the same numbers is in how much cold outreach costs, where you can turn meetings per 100 sent into a cost per meeting. And outbound that reads the reply shows why reading what people actually say beats staring at a percentage.
Because Rocketship sends from your own Gmail, every reply lands in your normal inbox, and its AI worker classifies them and books the meetings, so the counts above are easy to pull from the thread. More on the AI workers page.
You don't need someone else's benchmark; you need your own number, and a way to move it. Rocketship finds the buyers, emails them from your own Gmail, follows up, and books the meeting on its own. Plans that get you customers start at $24.99 a month (Launch), and every monthly plan starts with a 7-day free trial, card required.