Cold email reply rate benchmarks for 2026, and why they disagree

Sunny Ray · Oct 8, 2026 · 7 min read

A good cold email reply rate in 2026 is above about 3.4%, and a strong one is above about 5.5%. Those are the average and top-quartile numbers in Instantly's 2026 benchmark report, which is built on 2025 sending data. Hunter's 2026 report, based on 31 million emails sent in 2025, puts the average sequence reply rate at 4.5%. Belkins, measuring each individual email across 7.5 million sends, found weekday reply rates between 0.44% and 0.48%.

Those numbers look like they contradict each other. They mostly don't. They measure different things: a whole sequence versus a single email, any reply versus an interested reply. Once you know which number you're comparing to, the more useful question is what moves yours. In our own campaigns the biggest lever hasn't been the subject line or the send time. It's been the offer: a direct pitch versus an invitation onto a show.

The 2026 numbers side by side

Here's what each dataset says, in the publishers' own terms.

• Instantly, 2026 report (2025 data): average reply rate 3.43%. Top quartile 5.5% and above. Top 10% at 10.7% and above.
• Hunter, State of Cold Email 2026 (31 million emails sent in 2025): average sequence reply rate 4.5%.
• Belkins (7.5 million emails, 2025): weekday reply rates per email between 0.44% and 0.48%.

All three are vendor data. Instantly and Hunter sell sending and prospecting tools; Belkins is an agency. That doesn't make the numbers wrong, but each is drawn from that company's own customers, and those customers are not a random sample of everyone who sends cold email.

Why benchmarks disagree

Three things explain most of the gap.

Sequence versus email

Hunter reports a sequence reply rate: did the person reply at any point across all the steps? Belkins reports per email: of all emails sent, how many got a reply? A four-step sequence sends four emails to each person, so a per-email rate will always look several times lower than a per-sequence rate for the same campaign.

Who is sending

Customers of a tool that serves agencies and growth teams will skew toward experienced senders. A first-time founder with a new domain shouldn't expect top-quartile numbers in month one.

What counts as a reply

Most tools count every reply. Out-of-office messages, unsubscribe requests and "not interested" all count. On one of our investor campaigns this fall, 45 of the replies in a single stretch were out-of-office messages from European contacts. That inflates a reply rate without adding a single conversation.

Reply rate vs positive reply rate

The number you should run your business on is the positive reply rate: replies that show interest in what you offered. It's always lower than the headline reply rate, and benchmarks rarely publish it because it depends on judgment.

Our ROI calculator uses a deliberately cautious version of this. It assumes 4 in 100 prospects reply interested, and 1 in 4 of those book a conversation, which works out to about 1 booked conversation per 100 prospects. That's below what our own campaigns see, on purpose, so the payback month it shows isn't a best case.

What a direct pitch gets

On October 7 I was on two separate calls with founders, and both asked some version of the same question: what should I expect from cold email?

My honest answer about the direct pitch, the classic "we help companies like yours do X, got 15 minutes?" email, was that it converts, if you're lucky, at 0.2% or 0.1%. So you have to send thousands of emails to get a few meetings. On the other call I put it as: most cold email campaigns are far below the invitation numbers, and we'll usually see something in the 0.2% range. That's my rough number for what a straight sales pitch turns into, not a published benchmark, and it's why I don't build anything on direct pitching alone anymore.

The reasons are the ones you'd guess. Busy buyers get the same pitch from dozens of senders. The ask is for their time, to hear about you. And inbox filters are increasingly good at spotting sales offers.

What an invitation gets

The other half of my answer on those calls was the comparison. Our show-invitation campaigns, where the email invites the prospect to be interviewed on a show instead of pitching them, ran around 3 to 4%. On the screen I showed campaigns at 3.7%, 4.3% and 4.5%, and one at 12.3%.

That's the same team, the same mailboxes, and the same kind of lists. The difference is the offer. A pitch asks for 30 minutes so you can talk about you. An invitation asks for the same 30 minutes so they can talk about themselves. People say yes to the second one.

One caution: our invitation numbers are campaign reply rates from our own dashboards, and the 12.3% campaign was an outlier, not a norm. I share them because they're ours, not because you should expect them in week one.

Why the offer also protects deliverability

Reply rate and deliverability feed each other. Inbox providers watch how people react to your mail, and a pitch that gets ignored or marked as spam teaches them to filter you. An invitation that people answer does the opposite.

That shows up in how we set up sending. For podcast invitation campaigns, Charles recommends up to 20 mailboxes per domain, against about 10 per domain for direct sales campaigns, because invitations carry lower spam risk and get better responses. Each mailbox still sends only a handful of emails a day, somewhere between 4 and 10. And our invitations go out with no URLs in the body, since links in a cold first email are one of the easiest things for a filter to flag.

So when you compare your rate to a benchmark, remember the setup behind it. A 4% reply rate from a burned domain with a link-heavy pitch isn't the same campaign as a 4% rate from warm mailboxes sending a plain invitation. Your infrastructure decides whether the email gets seen. Your offer decides whether it gets answered.

How to measure your own

A benchmark is only useful if you compare like with like. Here's how to set up your own numbers so they mean something.

1. Pick one definition and stick to it. Sequence-level reply rate is the easiest to compare with Hunter and Instantly.
2. Track positive replies separately. Tag every reply as interested, not now, not interested, or noise. Your positive rate is the one to optimize.
3. Measure per campaign, not per account. An account average hides a campaign that works next to one that doesn't.
4. Give it enough volume. A few hundred sends is too small to judge. Wait for at least a thousand per variant before you call a winner.
5. Watch bounces before replies. If bounces climb, fix the list or the mailbox before you touch the copy.
6. Count booked conversations last. That's the number that pays the bills.

What to do if your rate is low

If you're under 1% on a sequence basis, check deliverability first: warm mailboxes, a separate sending domain, verified addresses, no links in the first email. If those are fine and the rate is still low, the problem is almost always the offer, not the wording. Rewriting subject lines on a pitch nobody wants is rearranging furniture.

Try the offer change before anything else. Keep the same list and the same mailboxes, and send an invitation instead of a pitch to half of it. You'll know within a few weeks.

If you want to see what these numbers mean for your revenue, the calculator is built on the cautious funnel above. See the cautious funnel behind our calculator and plug in your own deal size and close rate.

Frequently asked questions