Outreach dashboards are full of metrics that look like progress. Two of them tell you whether the thing is working, three warn you before you damage your domain, and the most prominently displayed one is close to fiction.
Modern mail clients pre-fetch remote images automatically to protect privacy, which means tracking pixels fire whether or not a human ever saw the message. Some corporate security systems open every link and image in incoming mail as a matter of policy.
The result is an open rate inflated by an unknown amount, varying by recipient, that you cannot correct for. Optimising against it means optimising against noise — and worse, it produces false confidence: a campaign showing sixty percent opens and zero replies looks like a copy problem when it is usually a targeting problem or a deliverability problem.
Replies divided by messages delivered. Unfakeable, unaffected by tracking changes, and directly connected to the only thing outreach exists to produce. This is your primary number.
Replies that express actual interest, as a share of all replies. A campaign generating many replies that are all "no thanks" is reaching people successfully and targeting them badly — a distinction reply rate alone hides.
| Metric | Healthy | What it means when it moves |
|---|---|---|
| Bounce rate | Under 2% | Rising means your list is stale or unverified. Stop and clean it — bounces damage reputation directly. |
| Spam complaints | Effectively zero | Any at all means slow down and re-read who you are contacting. This is the fastest way to lose a domain. |
| Unsubscribe / opt-out | Low and stable | A spike means your targeting has drifted to people with no reason to hear from you. |
Published benchmarks for cold email reply rates vary so widely between sources, industries and definitions that quoting a number would be misleading. What is more useful is the direction and the diagnosis:
Open tracking works by embedding an invisible image that loads when the message is displayed. Privacy features in modern mail clients now pre-fetch those images automatically, and many corporate security systems scan every incoming message by opening its contents. Both fire the pixel with no human involved, inflating the number by an amount you cannot measure or correct.
Published figures vary enormously depending on industry, offer, list quality and how "reply" is defined, so any single number quoted as a benchmark is close to meaningless. Track your own rate over time and judge changes against your own baseline. A campaign that doubles your previous rate is good news regardless of what someone else reports.
They are more reliable than opens, since a click usually requires intent — but security scanners do click links automatically, so the number is not clean either. In cold outreach, where you should be including at most one link, clicks are a weak signal compared with replies.
At twenty messages a day, give it two to three weeks and at least a couple of hundred sends before drawing conclusions. Judging a campaign on thirty messages means reading noise. If deliverability metrics go wrong, however, act immediately — those do not need statistical significance.
The Cold Outreach Machine finds the companies you sell to, writes a different email for each one, and sends them on its own — paced, inside working hours. US$299.99, paid once, live setup call included.