Cold email reply rates: reasonable benchmarks for 2026

A reasonable reply rate for cold emails in 2026 sits around 3 percent for an average campaign to completely unknown recipients. Well targeted sends with real personalization can reach 5 to 10 percent, while reactivating contacts who already know you often lands considerably higher. The difference is decided by list, timing and personalization, in that order.

What is a normal average for cold emails?

Around 3 percent, counted as replies per delivered email. That is the number that keeps coming up in the large roundups from the sending platforms, and it has fallen as volume tools have made it cheap to spam. The spread, though, is enormous:

• Under 1 percent: bought list, generic template, no reason to reach out. A large share of all sends sit here.
• Around 3 percent: a decent list, some segmentation, a template with swappable fields.
• 5 to 10 percent: hand picked list, real personalization, a clear reason to write right now.

Note that replies are not the same as positive replies. Of the replies, typically a third to a half are interested enough to continue the conversation.

Why does reactivation get a higher reply rate than cold lists?

Because the relationship does not start from zero. A lead that has sat in your CRM has met you at a trade show, downloaded something, taken a meeting or even received a quote. When you reach out again you are not a stranger but a thread that can be picked back up. That changes the math at every step:

• Delivery: the address is verified through the history, bounces are fewer.
• Open: the sender name or the company is recognized.
• Reply: the email can reference shared history, which a cold email never can.

That is why reply rates on reactivation sends are often a multiple of cold sends to the same type of company. The warm relationship is an asset most B2B companies leave unused, while they pay full price to reach strangers.

What drives the difference between 1 and 10 percent?

Three factors, in descending order:

1. The list. The right company and the right person account for more than half the outcome. No copy saves an email to the wrong recipient.
2. The timing. An email that ties into something that just happened at the recipient, a hire, a role change, a launch, answers the question "why now?" before it is asked.
3. The personalization. Not a name field, but content that proves the email was written to a person. One verifiable detail beats ten flattering adjectives.

Tools and sender domain decide whether the email arrives, but given delivery it is these three that separate the campaigns.

Why should you measure replies and not opens?

Because open rate is noise. Email clients' privacy protection registers opens that never happened, and an open says nothing about interest anyway. Reply is the first data point that requires an active decision from the recipient, and booked meetings are the only one that pays salaries.

A simple measurement ladder for B2B sends:

• Delivered: did the email arrive?
• Replies per delivered: did it prompt a reaction?
• Positive replies: was the reaction interest?
• Booked meetings: did the interest become a conversation?

Everything above reply on the ladder is diagnostics, not results.

If you want to run the numbers on what your own list of dormant contacts could produce, there is a calculator on Revexa's homepage. It is the same logic Leo, our AI for lead reactivation, is built on: known contacts, the right timing via signals, and replies as the only metric.

Common questions

Is 3 percent per email or per sequence?

Benchmarks are usually given per campaign, meaning the whole sequence including follow ups. A single touch sits lower, and the follow ups often account for a large share of the total replies.

Does the reply rate drop over time?

Yes, the trend is downward as AI has made mass sends cheaper and inboxes fuller. That punishes the volume play but rewards the sender who sends fewer and better emails.

What reply rate should we set as a target?

Set the target in meetings, not percent. Ten booked meetings from a thousand emails can be excellent or terrible depending on what a meeting is worth to you, and that calculation is the only one that matters.