Reply rate vs positive reply rate vs meeting rate: how do they differ?

What does each metric measure?

Swipe to see more →

Formulas vary by source, and that's the main source of confusion. In Instantly's 2026 benchmark report, overall reply rate is every reply, follow-up replies included, as a share of every email sent. Instantly's own guide to calculating reply rates counts only unique human replies and divides them by delivered emails (sent minus hard bounces); it treats positive reply rate as a share of all replies, and divides meetings by delivered emails. Same vendor, different denominators, and "positive reply rate" there is a share of replies, not of sends.

How do you keep the three consistent?

Pick one denominator and report everything per 100 emails sent. Here is the layout we'd use (our recommendation):

Swipe to see more →

Those figures are illustrative, not benchmarks. The table shows the shape: each row is a smaller number than the one above, and the gaps between rows tell you where to work.

Why do two campaigns with the same reply rate book different meetings?

Because replies aren't equal. Here is an invented example, per 1,000 emails sent, with the same 3% reply rate:

Swipe to see more →

Campaign B probably reaches the wrong people or makes an offer that triggers "not interested". At our house volume of about 2,520 emails a month for one rep (4 inboxes × 30 sends × 21 days), A books about 23 meetings and B about 8 (our math). A reply-rate dashboard shows no difference between them.

What do the calculator's starting numbers mean?

Our calculator estimates the sends, replies and meetings you need. It starts at a 3% reply rate and a 35% reply-to-meeting rate. Multiply them: 3% × 35% = 1.05 meetings per 100 emails, or about 10 per 1,000 (our math). That's why the difference on this page matters: use the same definition in both boxes. If you track positive replies, enter your positive reply rate per email sent (positive replies ÷ emails sent) in the reply-rate box and the share of those that book in the next one. If you track all replies, enter your raw reply rate and the share of all replies that book. Don't pair a raw reply rate such as the 3.43% average on Instantly's platform with a booking rate measured on positive replies only.

The calculator's reply-rate box expects a per-email rate: replies divided by emails sent, with follow-up replies counted, the same unit as Instantly's 3.43% average and not the share of prospects who ever reply. With 5 touches per prospect, a 3% per-email rate works out to roughly 14% of prospects replying at some point if every email had the same chance (our math: 1 minus 0.97 to the 5th power). Both defaults are assumptions you should replace with your own numbers.

What benchmarks exist?

For reply rate only. Instantly's 2026 report gives an average of 3.43%, a top-quartile rate of 5.5% and an elite tier of 10.7% or more. As far as we could read, the report doesn't break replies into positive and negative or give meeting rates. We found no public benchmark for positive reply share or meeting rate that we could trace to a named sample, so we're not going to quote one. Build your own baseline from the last 1,000 or more sends. If a vendor gives you a "typical" meeting rate, ask for the sample, the year and the denominator.

How many sends does each metric need to be reliable?

More as you go down the funnel, because the numbers get smaller. Our math, using the standard two-proportion test at 95% confidence and 80% power, for detecting a doubling:

Swipe to see more →

So a meeting-rate test needs about three times the sends of a reply-rate test to catch the same relative lift. A practical split: use reply rate for fast copy tests, then confirm the winner on positive replies over several weeks. The full sample-size table is in how to A/B test cold emails.

How do you track all three?

1. Set reply statuses in your sending tool. Instantly counts an opportunity when a lead is marked Interested, Booked meeting, Completed meeting (not closed) or Closed, or carries a positive custom status, per its help center.
2. Sort replies by hand until volume is high. Auto-replies, out-of-office and "remove me" shouldn't count as replies at all. Count unsubscribes separately.
3. Log the meeting. Mark it booked when it's scheduled and held when it happens, so you can see no-shows.
4. Review weekly (our weekly SDR metrics sheet shows the layout). Any row that falls below your baseline tells you where to fix: reply rate points to deliverability, list or subject; positive share points to audience and offer; meeting rate points to your ask.
5. Feed the results back into your math. If your meeting rate per 100 sent is 0.9, the calculator tells you how many you need to hit a target.

When should you skip each metric?

• Reply rate: skip it as your only number. Instantly's benchmark counts every reply, and as far as we could read it doesn't separate positive from negative.
• Positive reply rate: skip it as a test metric on small lists. With a few hundred sends, a handful of replies swings the percentage. Use it for weekly diagnosis instead.
• Meeting rate: skip it for A/B tests under a couple of thousand sends per version. It needs the most data to be reliable.
• Open rate: skip it as a winner metric. Smartlead lets you pick it as a winner metric, next to reply, positive-reply and click rates, but an open isn't a conversation.