What is a good cold email reply rate in 2026?
In 2026, the average B2B cold email reply rate sits between 3.4% and 3.7%. Instantly's 2026 benchmark report, drawn from billions of cold email interactions, puts the average at 3.43%, the top quartile at 5.5% or better, and the top 10% of campaigns at 10.7% or better. Saleshandy's analysis of 53.1 million cold emails sent between January and June 2026 puts the average at 3.7% with a 21% open rate. If you are under 3%, something in the list, the offer, or the delivery path is broken, and it is almost always the list.
That is the headline number, and it is also the least useful number in this article. The average is volume weighted, which means it is dominated by the people sending the most email. Below is what the same datasets show once you stop reading the top line.
| Tier | Reply rate | What it usually means |
|---|---|---|
| Broken | Under 3% | Bad list, wrong ICP, or mail landing in spam |
| Average | 3.4% to 3.7% | Competent sending at scale (Instantly, Saleshandy, 2026) |
| Top quartile | 5.5%+ | Real segmentation, disciplined follow-up |
| Top decile | 10.7%+ | Small lists, researched openers, tight ICP |
| Top 1% | 15% to 30% | Under 200 names, near-manual personalisation |
Why has the average reply rate kept falling?
The published average has fallen from roughly 8.5% in 2019 to about 5% in 2025 to 3.43% in 2026. Every vendor blog reads that as proof that cold email is dying. It is better read as proof that cold email got easier to send.
Sending volume per campaign rose faster than reply volume, and the average is calculated per email, not per sender. A single high volume sender pushing 200,000 emails a quarter at 1.5% drags the mean down further than a hundred boutique senders at 12% can lift it. The average is a description of the market's sending behaviour. It is not a forecast of yours.
The variable that moves reply rate more than copy
List size. Saleshandy's 2026 study broke 60,000 sequences down by how many people each campaign targeted, and the spread is larger than any subject line test ever produced.
| Campaign size | Reply rate | Sequences analysed |
|---|---|---|
| Under 200 contacts | 15% to 20% | 103,255 |
| 200 to 500 contacts | 11% to 13% | 7,119 |
| 500 to 1,000 contacts | 8% | 3,827 |
A campaign of under 200 people replies at roughly twice the rate of a campaign of 500 plus, inside the same dataset that reports a 3.7% average. Cleverly's 2026 benchmark set found the same shape at the extreme: campaigns under 50 recipients averaged 5.8% against 2.1% for large sends.
So the benchmark worth measuring against is not the global average. It is the average for your volume tier. A 6% reply rate on a 3,000 person list is excellent work. The same 6% on a 150 person list means the research was thin.
Is your reply rate even measuring replies?
Two measurement problems make most reported reply rates unreliable, and both inflate in the seller's favour.
The numerator is dirty. Dashboards count out of office autoresponders, "wrong person, try Sarah", unsubscribes and flat refusals as replies. A healthy positive reply rate in 2026 is 1% to 3%, and Saleshandy puts positive replies at 3% to 5% for well run campaigns, so a vendor reporting 6% total replies may be delivering 1.5% positive. Ask any agency to define the word before you sign. If they report replies rather than positive replies, they are reporting their own convenience.
The denominator is soft. Almost no benchmark states whether it divides by emails sent or by people contacted. With a four to seven step sequence, which is what Instantly's 2026 report identifies as optimal, those two denominators differ by five times. "Two to three meetings per 100 emails" and "two to three meetings per 100 people" are wildly different businesses.
Then there is placement, which quietly shrinks the real denominator further. Global average inbox placement in 2026 is 83.1%. Gmail delivers 87.2% to the inbox, down from 89.8% in early 2024. Microsoft Outlook sits at 75.6% and Apple Mail at 76.3%. Roughly one in six emails never reaches a human. Non compliant bulk senders see spam folder placement jump from a 5% to 10% baseline to between 22% and 34%.
The compliance floor is now explicit rather than folkloric. Google and Yahoo require spam complaints under 0.3%. Microsoft, since May 2025, requires a minimum 14 day domain age, a volume ramp capped at 200 a day for new domains, complaint rates under 0.10%, and DMARC authentication. Fail these and mail is rejected rather than filtered, which does not show up as a low reply rate. It shows up as silence.
What high reply rates look like in practice
Two engagements from our own work, both boutique consulting sellers, both run on curated lists rather than volume.
| Metric | Engagement A | Engagement B |
|---|---|---|
| Contacts reached | 1,400 over 5 months | 1,900 over 6 months |
| Reply rate | 13.2% | 11.9% |
| Meetings held | 22 | 28 |
| Close rate on calls held | 60% to 70% | 60% to 70% |
| Clients signed | 14 | 18 |
These are real numbers from real client work, anonymised, with identities redacted under NDA and figures rounded. Close rates reflect calls actually held, not calls booked. Your results depend on your offer, your market, and how well your list is built.
The honest reading of that table is not "we beat the benchmark by four times". It is that 1,400 contacts across five months is about 280 people a month, which puts both engagements squarely in the small list tier where 11% to 15% is the documented norm. Nobody sending 4,000 a month will see those numbers, including us. High reply rates are bought with list discipline and research time, and they are paid for in volume.
Do industry benchmarks mean anything?
Not much. In 2026, Cleverly's benchmark set puts legal services at the top with reply rates up to 10% and SaaS at the bottom, often under 2%. Saleshandy's 53 million email study, published the same year, reports SaaS as the highest performing industry and financial services and real estate as the lowest. Both cannot be right.
Neither is measuring industries. Each is measuring the mix of senders inside one vendor's customer base, and those mixes differ. Industry benchmark tables are the most linked and least reliable artefact in this category. Use your own trailing 90 days as the benchmark, segment by segment, and treat published industry figures as a sanity check on the order of magnitude rather than a target.
How do you turn a reply rate into meetings?
A reply rate is not a business outcome. Meetings held are. The conversion chain is short enough to do on paper, and every prospect should do it before signing anything.
Take 1,000 contacts a month at a 4% reply rate. That is 40 replies. Roughly 70% of replies to a well targeted sequence are positive rather than refusals or referrals, so 28. Around 70% of positive replies convert into a booked call, so 19. About 80% of booked calls are actually held, so 15 held meetings a month, or roughly 90 across a six month engagement.
That arithmetic is the useful benchmark, because it is denominated in the thing you are buying. It also sets the price test. Our entry engagement is $7,000 for six months with a $50,000 pipeline floor attached, which on the arithmetic above works out at roughly $78 per meeting actually held. Run the same division on any proposal you are handed. If the seller cannot show you the chain from contacts to held calls, the reply rate they are quoting is decoration. We break the cost side down in detail in our guide to what a lead generation agency costs, and the small list version of this argument in lead generation for consultants.
Two practical levers move that chain more than copy edits. First, follow up properly: Instantly found 42% of all replies arrive after the first email, and Saleshandy attributes 44% of positive replies to follow ups, with the best sequences running four to six touches across roughly three weeks. Second, add a channel: in Saleshandy's data, email only campaigns produced a 0.11% positive reply rate, email plus LinkedIn 0.21%, email plus calls 0.28%, and email plus LinkedIn plus a custom touch 0.88%. That is an eight times spread from sequencing decisions, not writing.
That combination, cold sequences and warm authority running as one system rather than two, is what the brain is built to hold. The plans attach a pipeline floor to it, so the reply rate stops being our metric and starts being our risk.
Quick answers
What is the average cold email reply rate in 2026?
The average B2B cold email reply rate in 2026 is 3.43% according to Instantly's benchmark report and 3.7% according to Saleshandy's analysis of 53.1 million emails. The top 25% of campaigns exceed 5.5% and the top 10% exceed 10.7%. Anything under 3% indicates a list, targeting, or deliverability problem.
What is a good positive reply rate for cold email?
A healthy positive reply rate in 2026 is 1% to 3%, and 3% to 5% for well run campaigns. Positive replies exclude out of office messages, referrals, unsubscribes and refusals, all of which most dashboards count as replies. Always ask a vendor which of the two numbers they are reporting.
How many cold emails does it take to book one meeting?
On typical 2026 rates, around 65 to 70 contacts produce one meeting that is actually held, working from a 4% reply rate through positive replies, bookings and show ups. Beware benchmarks quoted per email rather than per person, because a five step sequence makes those denominators differ by five times.
Why did my cold email reply rate drop in 2026?
The two most common causes are volume and placement. Reply rate falls sharply as list size grows, from 15% to 20% under 200 contacts down to 8% at 500 to 1,000. Separately, global inbox placement is now 83.1%, and Microsoft's May 2025 sender rules reject mail from domains under 14 days old or above a 0.10% complaint rate outright.
Is cold email still worth doing in 2026?
Yes, but not at volume without segmentation. The falling average reflects more email being sent, not fewer buyers replying, and small researched lists still reply in double digits. The economics work when you measure meetings held and cost per meeting rather than reply rate.