Cold Email and LinkedIn Reply Rates: What Counts as Good

What is the difference between reply rate, open rate and meeting rate?

Each one measures a different step between the message and the calendar: whether the email was seen, whether someone answered, and whether anything came of it.

Reply rate counts anyone who answered, even "not interested, thanks," minus auto-replies. Positive reply rate counts only interest: a question about details or price, or a yes to a call. The seven metrics worth tracking:

Reports differ most in the denominator: emails sent, people, or people who opened. I divide by people, so sequence length doesn't distort the result.

Why is open rate no longer reliable?

Because since 2021 some opens are recorded without a human opening the email, and some real opens aren't recorded at all.

Opens are tracked with a pixel, a tiny transparent image that logs a hit when it loads. In June 2021 Apple announced Mail Privacy Protection, which shipped with iOS 15. According to Apple, it stops senders from using invisible pixels to learn when an email was opened, and hides the recipient's IP address. In practice, Apple's servers load the content in the background, the pixel fires, and an open is logged that nobody made.

Mail clients that block images do the reverse, hiding real opens. The pixel also loads from the tracking tool's domain, one more thing spam filters look at, so in our campaigns open tracking is off. More on inbox placement in why emails go to spam.

So count replies. A reply is a person who typed something and hit send, and no server does that for them. Near-zero opens still call for a deliverability check.

What is a good cold email reply rate?

8% to 25% is the email reply range we see in targeted B2B campaigns with a verified list and a real reason to reach out, and global averages sit well below it.

Two global reports and their methods. Both span many industries and markets, with no focus on Israel:

I treat the average as a floor. Below 3% after a few hundred recipients, check the list, the deliverability and the offer; a new subject line almost never saves that campaign. Above 10% you're in Instantly's top tenth, and the question becomes how many replies are worth anything.

What is a good LinkedIn reply rate?

LinkedIn runs higher: in our campaigns the reply rate ranges from 20% to 45%, and global reports put replies to messages at roughly 7% to 10%.

Expandi's benchmark report from May 2026, covering 13.2 million connection requests sent between May 2025 and April 2026, found 28.5% accepted and 10.4% of messages answered. Belkins' LinkedIn study from 2026, on 2025 data, found an overall reply rate of 7.2%. Both are global, and both divide by messages sent.

One Belkins detail I keep returning to: connection requests with a personal note were accepted slightly less often (25.3% vs 27.6% without one) but brought more replies (8.2% vs 5.3%). Watch acceptance alone and you'd drop the note, losing exactly what you came for.

LinkedIn is higher because the message comes from a real person's profile, with no spam folder in the way. Track acceptance rate, the share of connection requests accepted, next to replies. Low acceptance usually points to the profile or the audience; high acceptance with few replies, to the offer.

The price is volume: each profile has a cap on connection requests (see LinkedIn limits). Our LinkedIn outreach runs from the client's existing profile, with no new account and no warm-up.

What moves the reply rate?

The list and the offer move reply rate more than any wording, and deliverability decides whether there's anything to measure at all.

An email in spam gets zero replies, so cold email goes out from warmed secondary domains with SPF, DKIM and DMARC, never your main domain. Each of our cold email setups has 100 Outlook and Google Workspace inboxes and 14 days of warm-up. In the inbox, the list does most of the work: the wrong role won't answer even a perfect message.

A signal, like open roles or a funding round, answers "why now": your message lands while the problem is on their desk. More in signal-based selling.

The offer is what the recipient weighs in the two seconds they give you. "Can I send it over?" is a far smaller ask than half an hour in their calendar, as the playbook on building an offer for outbound explains.

Then follow-ups. According to Instantly, 58% of replies came from the first email and 42% from follow-up steps. That's no reason to add five steps: a follow-up repeating the same ask in new words is noise.

How do you calculate your campaign funnel?

Multiply the conversion rate of each step, from the people on the list to the meetings held, and you'll see where the campaign loses people.

1. N: people on the list after verification.
2. D: delivery rate (1 minus the bounce rate).
3. R: reply rate.
4. P: the positive share of replies.
5. M: the share of positive replies booked as meetings.
6. S: show rate.

The formula: meetings held = N × D × R × P × M × S.

An example, with made-up numbers (nobody's average): 1,000 people, 98% delivery, a 10% reply rate, 40% of replies positive, half of those booked, and 80% of meetings held. That's 980 people reached, 98 replies, about 39 positive, about 20 meetings booked and 16 held.

It runs backwards too. For 10 meetings held, divide 10 by the product of the rates, about 0.0157, and you need roughly 640 people on the list.

Every letter weighs the same: doubling R from 5% to 10% doubles your meetings just like doubling the list, without burning through twice the people. In a market the size of Israel, where every ICP has a finite number of companies, that settles it.

Before comparing yourself to anyone's average, pull these six letters from your last campaign. Which one is lowest?