Cold email reply rate benchmarks
Two teams can report a 6 percent reply rate and a 0.45 percent reply rate for exactly the same campaign. The difference is the denominator, not the performance.
Michael Sparks · Published 2026-08-18 · Last reviewed 2026-08-18
Direct answer
Why published figures conflict
Three things vary between benchmark reports and any of them can move the headline number by an order of magnitude.
First, the denominator. Replies divided by sends is a much smaller number than replies divided by opens. Second, the sample. Agency data reflects that agency's clients, offers and list-building practices, which are not the market. Third, the definition of a reply. Some reports count every response including out-of-office and unsubscribes, others count only positive replies.
The four denominators in use
| Denominator | Measures | Bias |
|---|---|---|
| Emails sent | End-to-end campaign efficiency | Lowest number; includes bounces and spam placement |
| Delivered emails | Performance once mail arrives | Hides deliverability problems |
| Unique recipients | Response per person contacted | Rewards heavy follow-up sequences |
| Unique opens | Message quality once read | Unreliable since image pre-fetching became common |
Replies per email sent is the most conservative and the hardest to game, which is why we use it as the reference measure across this site.
Documented benchmarks
We only list figures where the source, sample size, date and denominator are all knowable. Where a widely quoted number lacks a stated denominator, we leave it out rather than repeating it.
| Source | Figure | Sample | Period | Denominator |
|---|---|---|---|---|
| Belkins | About 0.45% average reply rate | 7,530,489 emails | 2025 | Total emails sent |
| Belkins, trend | About 20% decline | Same dataset | H1 2025 to H2 2025 | Total emails sent |
This table is deliberately short. We add a row only when a source publishes its methodology, and we review the page at least quarterly. Last reviewed 2026-08-18.
Convert your own campaign
Enter your campaign numbers to see the same performance expressed against every denominator. This is the fastest way to find out whether you are actually behind a benchmark or just measuring differently.
Same campaign, five denominators
- Replies per email sent
- 0.96%
- Replies per delivered email
- 0.99%
- Replies per unique recipient
- 0.96%
- Replies per unique open
- 2.53%
- Positive replies per email sent
- 0.32%
Delivered is calculated as sent minus bounced. Open-based rates are shown for comparison only, because open tracking has been unreliable since mail privacy features began pre-fetching images.
The calculator runs entirely in your browser. Nothing you enter is sent anywhere. A standalone version lives at the reply rate calculator.
Setting a defensible baseline
Before running any experiment, record your own reply rate per send across at least four weeks and at least a few thousand emails, split by segment. External benchmarks are useful for sanity-checking, not for target setting, because your list quality and offer differ from any published sample.
When you then test a change, hold the denominator fixed, keep a matched control cohort and report positive replies separately from total replies.
What this means in practice
Ignore any benchmark quoted without a denominator. Measure your own campaigns per email sent, compare only against yourself, and treat published averages as context for the direction of the market rather than a target.
Common questions
- What is a good cold email reply rate?
- There is no single figure, because reply rates depend on the denominator, the list quality and the offer. Belkins analysed 7,530,489 cold emails sent in 2025 and reported an average of roughly 0.45 percent replies per total emails sent. A campaign measured per delivered email or per open will report a higher number for identical performance.
- Why do reply rate benchmarks disagree so much?
- Because vendors measure different things. Replies per send, per delivered email, per unique recipient and per open produce very different percentages from the same campaign. Sample composition also varies, since agency data, self-reported survey data and product telemetry cover different senders.
- Are cold email reply rates falling?
- In the largest recent dataset, yes. Belkins reported that average replies per send declined about 20 percent between the first and second halves of 2025. That is one dataset from one agency's client base, so treat it as a directional signal rather than a universal law.