What is a good cold email reply rate?

Per contacted prospect, a reply rate around 3% is average and 5% or more is good: Instantly's 2026 benchmark puts the average at 3.43% and the top quarter at 5.5% or higher. Per email sent, the numbers are far lower; Belkins measured 0.45% across 7.5 million cold emails in 2025. So "good" depends on the denominator. Pick one, track positive replies separately, and compare against your own segments rather than a headline figure.
Below we lay out what each benchmark measures, why they disagree, how to calculate your own rate, what moves it and what to do when it is low. Figures come from each source's own page, read on 24 October 2026. We make Nagent, whose outbound research lead is NORA; how NORA classifies replies is in its own section.
What reply rates do the benchmarks report?
Instantly, 2026 report: average reply rate 3.43%; top quarter 5.5% or more; top tenth 10.7% or more. Data from 1 January to 18 December 2025 across the platform's workspaces (Instantly).
Belkins, 2026 study: 0.45% average across 7,530,489 cold emails sent in 2025, with 34,393 replies. Belkins notes its earlier studies used a different denominator, so the figures are not comparable year to year (Belkins).
Apollo: calls 3% to 6% a healthy reply rate for broad B2B cold email in 2026 and 6% to 15% or more for tight ICP or intent-based campaigns; it treats under 3% as a targeting or deliverability problem (Apollo).
Sales.co data, as quoted by Apollo: a 2.09% average reply rate across more than 2 million emails, with 14.1% of replies positive, about 0.64% interested replies overall.
Why do cold email benchmarks disagree so much?
Because they divide different things:
Per email or per prospect. A four-step sequence sends four emails per prospect. A 2% reply rate per prospect is about 0.5% per email sent. Belkins' figure is per email; most platform benchmarks report per contact or per campaign.
Total or positive replies. Total replies include "not interested", "remove me" and sometimes out-of-office messages. Positive replies are a fraction of that.
Whose data. Platform data covers every user, good and bad. Agency data covers one team's campaigns. Vendor blogs often quote other vendors.
Cleaning. Some exclude bounces and auto-replies from the denominator; some do not.
Unify makes the point well: there is no universal good reply rate, and an observed rate is an estimate with an uncertainty that grows as the sample shrinks (Unify).
How should you calculate your reply rate?
Write the definitions down before you look at results:
Denominator: unique prospects who received at least one email (delivered, not bounced).
Reply rate: prospects who sent any human reply, divided by the denominator. Exclude out-of-office and automated replies.
Positive reply rate: prospects whose reply shows interest or asks a real question, divided by the same denominator. Decide in advance how you count referrals ("talk to my colleague") and "not now, try in Q2".
Meeting rate: prospects who booked a meeting, divided by the same denominator.
Report all four together, with the counts, not only the percentages.
An example with round numbers: 400 prospects delivered, 14 human replies, 5 positive, 3 meetings. Reply rate 3.5%, positive 1.25%, meetings 0.75%. On 400 prospects, each reply moves the rate by a quarter of a point, so do not read much into a difference of one or two replies.
What reply rate should you aim for?
Combining the sources above, per contacted prospect:
Below 2%: something is wrong. Check deliverability first, then targeting.
2% to 3%: around average for broad lists.
3% to 5%: healthy.
5% to 10%: strong; usually tight targeting or a real trigger.
Above 10%: top tier on Instantly's data, typically narrow segments with a clear reason to write.
For positive replies, the Sales.co figure quoted by Apollo implies well under 1% across all sends. Track your own baseline and aim to raise it, rather than chase a single external number.
What changes cold email reply rates most?
Who you write to. Belkins found founders and owners replied at 0.57% per email against 0.32% for VPs, and companies of up to 10 people at 0.72% against 0.22% for those with more than 10,000 staff.
Why now. A recent, sourced trigger gives the reader a reason. Apollo's higher 6% to 15% band is for intent-based or tight-ICP campaigns.
The first email. Instantly found 58% of replies come from step one, so most of the effort belongs in the first touch.
Follow-ups. The other 42% of replies come later; Instantly puts the useful range at 4 to 7 touches, spaced 3 to 4 days apart.
Deliverability. An email in spam cannot be replied to. Authentication, list hygiene and complaint rates come before copy.
List size per campaign. Apollo notes that campaigns to 50 or fewer recipients consistently see higher reply rates than those to 1,000 or more (Apollo).
How do reply rates vary by industry, region and timing?
Belkins' 2025 data, per email sent, shows how much context moves the number:
Industry. Food and beverage led at 3.47% (on at least 20,000 emails). Construction, financial services, healthcare and legal services sat between 0.56% and 0.60%, and banking and insurance were the lowest.
Region. Canada 0.63%, the US 0.51% and the UK 0.48%.
Time of day. 8 am to noon drew 0.54%, against 0.40% for 8 pm to 11 pm.
Season. February peaked at 0.54% and December fell to 0.35%.
Instantly's data adds that Tuesday and Wednesday see the highest reply rates, with Wednesday on top.
Two lessons follow. First, compare yourself with your own industry and region, not an all-market average. Second, timing effects are real but small next to targeting: moving a send from evening to morning is worth a fraction of a point, while writing to the wrong company is worth all of it.
Why is my cold email reply rate low?
Work through the causes in order:
Deliverability. Check bounces, spam complaints and inbox placement with seed accounts. If mail lands in spam, nothing else matters.
List quality. Wrong roles, old data, catch-all addresses.
Targeting. Accounts that do not fit your profile, or no trigger.
The first email. Generic opening, several asks, too long.
Follow-up logic. Stopping after one email, or sending seven to people who said no.
Fix them in that order. Copy changes on a list with deliverability problems teach you nothing.
Is reply rate the right metric?
It is a useful early signal, but it can mislead. A provocative email can raise replies and complaints at the same time. Track it alongside:
Positive reply rate, to see interest rather than noise.
Negative and "remove me" replies, an early warning for spam complaints.
Meetings held, recorded after the meeting, not just booked.
Pipeline created from each segment.
Where Nagent fits
We make Nagent. "Nagent is multiplayer AI for growth: people and AI agents in one shared context, delivering the number." NORA is the AI Outbound Research Lead on SERA's Sales team.
Replies classified. Each reply is sorted as interested, objection, scheduling or unsubscribe, with anything else set aside as noise, and the cadence stops or changes course.
Opt-outs honoured. A reply asking to stop writes the address to the suppression list, checked at every send.
Meetings counted honestly. A qualified reply becomes a handoff to a named person, a handoff nobody acknowledges within 36 hours escalates, and NORA counts a meeting as held only from the outcome a person records.
Results by cohort. Insights breaks outcomes down by industry, persona, signal and score band, and says when there is not yet enough data.
What it does not do: NORA does not publish industry benchmarks and sends email only, so LinkedIn and phone replies are outside her numbers. Work is metered in credits; see pricing.
Frequently Asked Questions
What is a good positive reply rate for cold email?
Well under the total reply rate. The Sales.co data Apollo quotes works out to about 0.64% interested replies overall. Set your own baseline per segment and work to raise it.
How many follow-ups should a cold email sequence have?
Instantly's data puts the useful range at 4 to 7 touches, 3 to 4 days apart, with most replies coming from the first email.
Why do my cold emails get replies but no meetings?
Usually the replies are objections or "not now", the ask is too big, or the account is a poor fit. Look at positive replies and fit before rewriting the call to action.
Is a 1% reply rate bad?
Per prospect, yes; it suggests deliverability or targeting problems. Per email sent across a multi-step sequence, it can be normal: Belkins' 2025 average was 0.45% per email.
Should I still track open rates?
Treat them with caution. Opens are an unreliable signal of inbox placement, and Apollo notes that removing open tracking has been associated with higher reply rates in recent data. Replies and meetings are the numbers to manage.
See how NORA sorts replies and hands qualified ones to a person: book a demo.