Cold Email Statistics (October 2026): Our Campaign Data Beside the Published Benchmarks

Home › Blog › Cold Email Statistics (October 2026): Our Campaign Data Beside the Published Benchmarks

Cold email statistics for 2026 with every denominator labelled: Instantly, Smartlead and Belkins benchmarks beside our own campaign data, sourced and dated.

By Moe Alhosni · Published 16 October 2026 · 13 min read

Short answer: Cold email statistics disagree because they count different things. As of October 2026, published average reply rates run from 0.45% of emails sent (Belkins, 2025 data, auto-replies removed) to 3.43% (Instantly, all replies per email sent). Our four 2026 client campaigns drew 0.32% to 1.00% human replies per email sent. Check the denominator before you compare anything.

I'm Moe Alhosni, founder of ROI Maximizer (ROI Maxi). We run cold email for clients and get paid per qualified meeting. So I gain when cold email looks like it works, and again when it looks hard enough to hand to someone else. Hold that against every number below.

I checked every outside source on 16 October 2026. Our own data comes from four datasets that I never add together: an inbox export of positive replies dated 7 August 2025 to 24 July 2026, a campaign-rate snapshot taken on 26 July 2026, 2026 step data pulled on 13 October 2026, and an audit of two clients' campaigns on 8 October 2026. The one meeting figure comes from the published PR Noir case study.

How we count

Every statistic on this page uses one of these definitions, and its denominator sits next to the number.

1. A human reply is any reply a person wrote, whether "yes", "no" or "remove me"; auto-replies and out-of-office messages don't count.
2. A positive reply is a reply carrying the Interested label in our Instantly workspace, set mostly by Instantly's AI classifier.
3. A reply event is one message, so a person who replies positively three times counts as three reply events.
4. A unique lead is one person, counted once and dated by their first positive reply.
5. Per email sent means replies divided by every email that went out, follow-ups included.
6. Per contact means replies divided by the number of people contacted, however many emails each one received.
7. A sequence step is one email in the series: the first email, then follow-up 1, follow-up 2 and so on.
8. A bounce is an email the receiving server refused, and bounce rate is bounces divided by emails sent.
9. A booked meeting is on a calendar; a held meeting actually took place.
10. A qualified meeting is a held meeting that matches a written definition agreed before outreach starts, like the one in what counts as a qualified meeting.
11. A spam rate, as Gmail and Yahoo use the term, is the share of your inbox-delivered mail that recipients report as spam.

Items 5 and 6 cause most of the confusion. In Campaign A below, the same 360 human replies come to 0.78% per email sent, and 4.4% of the 8,227 contacts replied. Both figures are correct.

Published cold email benchmarks, side by side

Published cold email benchmarks put the average reply rate anywhere from 0.45% to 3.43%, and most of that gap is the denominator.

Instantly defines its rate as "all replies received (including follow-up responses) divided by total emails sent" and doesn't say whether out-of-office replies are removed first. That matters: in Sales.co's sample, 45.1% of 61,770 classified replies were auto-replies and 14.1% were positive.

Two rows divide by contacts and two by emails sent. Once contacts get more than one email, a per-contact rate always reads higher than a per-email rate for the same campaign, so Smartlead's 0.74% and Belkins' 0.45% are not two readings of one thing.

One figure is left out on purpose. An 8.4% single-email reply rate appeared in the earlier (2025) edition of Belkins' follow-up statistics page, drawn from 16.5 million emails, and secondary sites such as Prospeo's outbound email guide still repeat it. It is not on the current follow-up page (updated 26 June 2026) or in Belkins' response rate study, as of 16 October 2026.

For interested replies only, see our cold email positive reply rate benchmark.

Our cold email statistics, with denominators

Four datasets and one published case study sit behind our numbers. Each row names its source and what it divides by.

Between 7 August 2025 and 24 July 2026, our Instantly workspace logged 2,333 positive reply events from 1,378 unique people, excluding one client's campaigns. Our own campaigns produced 1,389 of those reply events and client campaigns 944.

Don't divide 2,333 by 91,209. They are different campaigns over different dates, and one counts positive replies while the other counts every human reply.

Average cold email response rate, per email sent

Our average cold email response rate, counting every human reply, ran from 0.32% to 1.00% per email sent across four 2026 campaigns.

P1 Communications is a UK hospitality PR agency and PR Noir a US PR agency, both public clients.

Campaign A is big enough to set against Smartlead's follow-up benchmark, which gives "the reply rate for emails sent at each position in the sequence" for the first half of 2026. Smartlead doesn't state the denominator or say whether auto-replies count. Ours leave auto-replies out.

Smartlead's rate falls from 1.18% to 0.33% over seven emails. Ours held between 0.65% and 0.88% from follow-up 1 to follow-up 5, in one market: hotel and travel brands. On Campaign A's first email, the best of three variants drew 1.34% per email sent and the worst 0.79%, on the same list; those two differed in subject line and body. See cold email subject lines.

Which step the positive replies came from

Between 7 August 2025 and 24 July 2026, 32% of our own campaigns' positive replies came from the first email; a B2B PR client's four-step sequence got 66% from its first.

Each column is that sequence's share of positive reply events from 7 August 2025 to 24 July 2026, rounded. Our own sequences are recorded as six steps, yet 2% arrived at a seventh email; I've left it as recorded. My reading, which the data doesn't prove: the PR client's first email leads with a press-coverage hook, and that hook works on first contact.

Counting every human reply instead, the first email produced 14% to 39% of replies in the four 2026 campaigns above. Different measure, different campaigns. The full breakdown is in how many follow-ups a cold email sequence needs.

What people asked when they replied positively

Half of the positive replies to our own campaigns between 7 August 2025 and 24 July 2026 asked for a call. Most of the rest wanted proof or the price first.

Both columns are positive reply events from 7 August 2025 to 24 July 2026, classified from the full text of each reply. With 15% to 23% asking the price, whoever answers the inbox needs a price answer ready in writing. How to answer each type is in how to reply to interested cold email replies, and the split by vertical is in the cold email positive reply rate benchmark.

Positive replies per contact, across 17 campaigns

Across 17 own and client campaigns live at our 26 July 2026 snapshot, positive replies per contacted lead ran from 0.06% to 0.71%. The campaign analytics behind these rates don't say whether a positive reply is counted per reply event or per unique lead.

Our internal targets are 0.5% per contact in Europe and 0.3% in the US. Below 0.2% we stop and check the list and the offer, and eight of the 17 were under that line at the snapshot. At the 26 July 2026 snapshot, the same offer drew roughly three times as many positive replies per contact in Europe as in the US. To turn a rate into a cost, use the cost per meeting calculator.

Cold email bounce rate benchmark and sender limits

Instantly's 2026 report (updated 12 January 2026) says to keep bounces below 2%. The two B2B clients' campaigns we audited on 8 October 2026 bounced 1.0% and 1.2% of emails sent, in their first two to three days of sending.

The low bounce rate hid a bad list. In the same audit, 18% to 24% of the leads on each list named a company that did not match their email domain, most likely job-changers. Most of the emails already sent to them did not bounce, and some came back as "no longer with the company" auto-replies. Check a list before it goes out with the lead list checker; the diagnosis is in why cold emails get no replies.

The mailbox providers set their own limits. This is information from their pages as of 16 October 2026, not legal advice.

Microsoft's post (2 April 2025, updated 30 April 2025) covers its consumer service only, and it gives both Junk-first and rejection (error 550 5.7.515) as the outcome, so read it yourself. Sending volume per inbox is in cold emails per inbox per day.

What this data cannot tell you

Four things are missing, and I won't estimate any of them.

• Opens. We don't track them, so there's no open rate on this page.
• A positive reply rate per email sent. The 2,333 positive replies come from an inbox export with no send counts.
• Positive replies by step for the four 2026 campaigns. That step data counts human replies of every sentiment.
• Meetings per reply. The one published lead-to-meeting figure is PR Noir: 27 leads, 10 meetings, 37% lead-to-meeting, and a $90K deal. One client. Don't multiply anything by it.

Two smaller limits. The Interested label is set mostly by an AI classifier, so some labels are wrong both ways. And the export skipped a few short windows, none longer than a day, in September 2025, December 2025, May 2026 and July 2026, where corrupted email bodies broke it.

When these numbers say don't hire us

Three cases where I'd tell you not to hire any agency, us included.

• Fewer than about 1,000 people you could sell to. You can't afford to burn the list while testing. Write four emails yourself and read every reply.
• No outbound deal closed yet. Send fifty emails yourself first. The objections you hear are what the follow-ups get written from.
• Meetings needed this month. Most human replies in our four 2026 campaigns came after the first email, so a sequence takes weeks to play out.

Who collected this data

ROI Maximizer (ROI Maxi) is a B2B lead generation agency founded and run by Moe Alhosni. The team works from Warsaw, Poland; the company, M Ventures LTD, is registered in the UK; clients are in the UK and the US. We run cold email outbound and charge for qualified meetings that take place: a one-off setup of $1,000–$3,000, then $200–$500 per meeting, or a flat $5,000 a month. There is no minimum term.

We bill in arrears, only for meetings that took place and match the written definition, and charge nothing for a no-show; you can pause, you can leave on 30 days' notice, and you keep the domains and data.

The full terms are in pay-per-meeting lead generation pricing. What an agency adds over sending software is in what a cold email agency does, and the team is on about ROI Maximizer.

Frequently asked questions

Which cold email reply rate should I compare against?

One counted the same way as yours: Belkins measured 0.45% per email sent in 2025 with auto-replies removed, and Smartlead's first-half 2026 median is 0.74% per contact. Our four 2026 client campaigns drew 0.32% to 1.00% human replies per email sent. For interested replies only, use the cold email positive reply rate benchmark.

What is the average cold email response rate?

Instantly's 2026 report gives 3.43% for 2025: all replies, follow-up responses included, per email sent. Compare it only with figures counted the same way.

What bounce rate is too high for cold email?

Instantly's 2026 report says to stay below 2%. The two clients' campaigns we audited bounced 1.0% and 1.2% of emails sent in October 2026, and their lists still had 18% to 24% of leads most likely at the wrong company.

Do follow-up emails still get replies?

Yes. In our four 2026 campaigns the first email produced 14% to 39% of all human replies, so most came from follow-ups. 68% of our own campaigns' positive replies also arrived after the first email.

Why do cold email statistics disagree so much?

They divide by different things: emails sent, contacts, or something unstated. Some also leave auto-replies in, and Sales.co found 45.1% of the replies it classified were automatic.

Before you sign with anyone

Get four things in writing: the qualified-meeting definition, what happens on a no-show, who owns the domains and data when you leave, and what a month with zero meetings costs. Ours is in what counts as a qualified meeting. Then run your deal size through the ROI calculator.

Keep reading

• Cold email agency pricing: what it actually costs, line by line (2026) — Domains, mailboxes, platform, data, warm-up, hours: every price fetched and dated. What a $1,000–$3,000 setup and $200–$500 per held meeting buys.
• How to Respond to a Positive Cold Email Reply: What They Ask For and What to Send — How to respond to a positive cold email reply: what interested prospects ask for, from full-body reply data, plus one template for each kind of ask.
• Cold email subject lines behind 2,333 positive replies, each with its denominator — Cold email subject lines from our own data: the patterns behind 2,333 positive replies, one same-list test, and two outside studies. No open rates.

Or run the free GTM teardown on your own site for the ICP, cold email and numbers specific to your company.