The three numbers, and what each one is actually telling you
Cold email has three headline metrics and they fail in different directions, which is why reading any one of them alone produces bad decisions. Open rate tells you whether your mail is being delivered at all. Reply rate tells you whether the message is landing with a real person who had a reaction. Positive reply rate tells you whether that reaction was interest. A campaign can post a strong reply rate made almost entirely of unsubscribe requests and annoyed objections, and a dashboard that only shows reply rate will call that a win.
Benchmarks we operate against
| Metric | Good | Great | Elite | What a bad number means |
|---|---|---|---|---|
| Open rate | 45-55% | 55-70% | 70%+ | Deliverability problem, not a copy problem |
| Reply rate | 3-5% | 5-10% | 10%+ | Usually the list or the offer, rarely the subject line |
| Positive reply rate | 1-2% | 2-5% | 5%+ | Message-to-segment mismatch |
| Send-to-lead ratio | 2,000:1 or better | - | - | 5,000:1 or worse means stop and adjust, do not scale |
Why open rate has become the least useful of the three
Open tracking works by loading a tiny invisible image. Mail privacy protection, corporate security scanners and image proxies now load that image whether or not a human ever read the message, which inflates open rate, and image blocking in other clients suppresses it. The result is a number with real noise in both directions. It is still worth watching, but only for one purpose: detecting collapse. An open rate that falls from 55 percent to 20 percent over two weeks is a deliverability emergency. An open rate that moves from 52 to 57 percent after a subject line test is noise, and rewriting your program around it is wasted effort.
When reply rate is low, fix the list before you rewrite the copy
The instinct when replies dry up is to rewrite the email. That is usually the wrong first move. In order of how often it is the actual cause: the list has been contacted before or has decayed, deliverability has drifted and more mail is quietly landing in spam, volume was scaled faster than sending reputation could support, the offer does not match the segment being targeted, or the timing is wrong against a buyer's renewal cycle. Copy is somewhere below all of those. A good message sent to the wrong 5,000 people will underperform a mediocre message sent to the right 500.
The one number a business owner should actually track
If you only look at one figure, make it the send-to-lead ratio: emails sent divided by qualified leads produced. It folds deliverability, list quality, copy, segment fit and qualification standard into a single number, and unlike percentages it maps directly onto cost. At 2,000 sends per qualified lead a program is healthy. At 5,000 to 1 the correct action is to stop and diagnose, because scaling a 5,000-to-1 program just buys more of a broken ratio. We publish the full set of operating benchmarks in the benchmark report.
What these numbers do not measure
None of the three metrics capture the prospect who read every email, never replied, and called you four months later when their contract came up. In trades with annual renewal cycles, commercial cleaning, landscaping, facility maintenance, that delayed response is a meaningful share of the pipeline a program actually creates. It shows up as revenue long after the campaign dashboard has stopped claiming credit for it, which is the strongest argument for judging outbound over two or three quarters rather than over thirty days.