Benchmarks get thrown around without anyone saying what is in the numerator or the denominator, which makes most of them useless. Before comparing your campaign to anyone else's, be sure you are comparing the same measurement.
Per person, or per email?
This is the difference that matters most, and it can move the headline number by a factor of three.
Most sending platforms calculate reply rate as the share of people contacted who replied at least once — not the share of emails sent that got a reply. Since a sequence sends several emails to the same person, those two denominators are very different. A campaign that sends 30,000 emails to 10,000 prospects and gets 1,700 people replying is at 17% per person and about 5.7% per email.
Neither is dishonest. But an agency quoting the per-person figure against a benchmark calculated per email is comparing two unlike things, and it will always look better.
Replies include the ones you did not want
A reply is a reply. Out-of-office autoresponders count. So does "not interested", "wrong person", and "remove me". Platforms count them because they are replies, and that is the honest definition — but it means reply rate measures whether your message provoked a response, not whether anyone wants to buy.
For reference, our own five campaigns run between 15.7% and 18.6% measured this way, against a commonly cited norm of 3–8%. That gap is real and it is a good sign. It is also not the number that pays for the campaign.
Opportunities are a subset of replies, never equal to them
Most platforms let you mark a lead with a status — interested, meeting booked, meeting completed, closed. The "opportunities" figure counts leads carrying a positive status, which makes it a filtered subset of the people who replied.
Two cautions. It can include anyone who was ever marked positive, even if the deal later went nowhere, and the same person can appear across campaigns. It is a useful directional number and a poor contractual one.
The number actually worth judging on
Meetings booked on your calendar.
It is harder to inflate, it is countable without interpretation, and it is the thing you hired the channel to produce. A campaign at 6% reply that books more meetings than one at 18% is the better campaign, and any agency that tells you otherwise is selling you the metric they perform well on.
If you want one supporting number alongside it, use the share of replies that became booked meetings. That measures reply handling — whether the people who raised a hand were worked properly — which is usually where campaigns are won or lost.
What to ask when someone quotes you a benchmark
- Is that per person contacted, or per email sent?
- Does it include out-of-office and not-interested replies?
- How many meetings did it produce, and over what period?
- Can I see the dashboard rather than a slide?
Those four questions will tell you more about an agency than any case study will. We answer them on a screen share, because a screenshot proves less than the account it came from.