What is a good cold email reply rate in 2026?

A good cold email reply rate in 2026 is 3–5% if you count every reply against emails sent, the way most sending tools report it. Instantly puts the platform average at 3.43% and the top quarter of campaigns at 5.5% and above. Anything over 10% is top-decile work. But if you strip out auto-replies and out-of-office messages, the honest number for most B2B teams is well under 1%, and that is not a sign that something is broken. Before you compare yourself to any benchmark, check which of these two numbers it is.

Why the published benchmarks disagree by 10x

I went through the four largest datasets published this year. They look contradictory until you read how each one defines "reply rate".

Three things move the number:

• Auto-replies. "I'm out of office until Monday" is a reply as far as a sending tool is concerned. It is not a reply to you. Belkins removes them and lands at 0.45%. The platform reports do not say they remove them.
• The denominator. Sent, delivered and contacted are three different numbers. Dividing by delivered emails flatters you when your bounce rate is high.
• Average versus median. A few very good senders pull an average up. Smartlead reports the median sender, which is why its figure looks so low: half of senders do worse than 0.74%.

Belkins makes the point on its own page: a 5% reply rate against people who opened and a 0.45% reply rate against everyone sent can describe the same campaign.

How to calculate your own reply rate

Use one formula and keep it for a year, so you are at least comparing yourself to yourself:

1. Take every campaign that finished in the period.
2. Count unique people who wrote back by hand. Leave out auto-replies and bounce notices. Decide once whether "please remove me" counts.
3. Divide by the number of people you contacted, not emails sent. One person who got four emails is still one chance at a conversation.
4. Separately, count positive replies: anyone who asked a question, agreed to talk, or pointed you to the right person.

The positive reply rate is the one to manage. Saleshandy reports an average positive rate of 3–5% of delivered emails, which shows how generous its counting is next to Belkins'. For a small B2B team writing to a cold list, one positive reply per 100 people is decent and two is good.

What actually changes the number

Benchmarks earn their keep here: they show what moves replies across millions of emails, not in one person's anecdote.

Follow-ups bring in about half of all replies. In Instantly's data the first email gets 58% of replies and steps two to four get the other 42%. Belkins found the split reversed: 41.4% from the first email, 58.6% from steps two to six. Either way, a campaign with no follow-ups throws away close to half of what it could get. Each later step earns less, though. In Smartlead's data the first email gets 1.18% and the seventh gets 0.33%, so somewhere around the third or fourth step you are mostly annoying people. Where to stop is covered in how many follow-ups before giving up.

Shorter first emails do better. Smartlead found that first emails under 50 words get one reply per 68 contacts, against one per 97 for emails of 100–149 words. Instantly says its top performers average under 80 words.

Smaller lists do better. In Saleshandy's data, campaigns under 200 prospects reached 15–20% reply rates, 200–500 got 11–13%, and 500–1,000 got 8%. There is nothing magic about 200. A list that small is usually one somebody built by hand, for a reason they can state in the first line.

Day of week barely matters. Smartlead sees 0.82% to 0.91% across days. Instantly and Belkins both put Wednesday slightly ahead. Do not spend a month testing send days before you have fixed the list and the first line.

A first email that sits in the top quarter

The shape that holds up across these datasets: under 70 words, one specific reason for writing, one question that is easy to answer.

Subject: your Berlin hiring Hi Anna, saw you're hiring three SDRs in Berlin this quarter. Teams that grow outbound that fast usually hit the same wall: replies come in across email and LinkedIn and nobody owns the follow-up. We've written up how two agencies handled it. Want the two-paragraph version? Max

And a second step that reads like a reply, not "just bumping this up":

Anna, one detail I left out: neither of them changed their CRM to do it. Happy to drop this if the timing's wrong.

Both are short enough to read on a phone in a lift. Neither asks for a call in the first message.

What your number tells you

If you are under 1% on a raw count, the copy is rarely the cause. More often the email is going to spam. Gmail starts filtering senders whose complaint rate climbs past 0.3%, and a new domain needs a few weeks of warm-up before it can carry volume. See how long to warm up a new email domain and what to do when your Gmail spam rate is above 0.3%.

The reply is the start, not the finish

Most teams obsess over the reply rate and then lose the replies. Smartlead found that 64% of replies arrive within an hour of sending and 88.5% within a day. A good share of your best moments land while you are in a meeting, and the lead has gone quiet by evening.

That is the part we built Nitrino for. It runs a campaign the way the numbers above suggest: a segment, a separate first email for each person, sending paced to the account's health, no more than three follow-ups. When someone replies, it answers in your voice within minutes and calls you in only when the conversation needs your decision. If you already sit at 3–5%, the cheapest win is usually answering those replies faster, not sending more email.

Whatever you use, measure one number the same way every month and judge yourself against it. A benchmark from someone else's tool is a rough guide. Your own trend line is the real one.