Metrics
Reply rate is the one metric that predicts everything else
Open rates lie. Click rates are noisier than they look. Unsubscribe rates are a lagging indicator. If I could only watch one number, it would be replies.
By MelAudit
For most of the last decade, the open rate was the metric email marketers led with. Every dashboard put it front and centre. Every campaign was judged by it. Every subject-line A/B test was framed around it. And then, in 2021, Apple's Mail Privacy Protection pre-loaded tracking pixels for everyone using Apple Mail, and open rates stopped meaning what they used to mean, and most dashboards never quite recovered.
In the years since, I have watched a lot of teams grope for a replacement metric. Some pivoted to click rates. Some engineered elaborate open-rate cleansing formulas that tried to strip out the Apple-inflated numbers. Some just kept using open rate and pretended the change had not happened. All three of these have problems.
The metric I have watched hold up best across every program I care about, over the last several years, is the reply rate. It is the metric I would keep if I had to give up all the others.
Why reply rate is honest
A reply is expensive. A subscriber has to open the email, decide to write something, actually write it, and send it. That entire sequence is difficult to fake at scale, difficult to accidentally trigger, and difficult to inflate through bots or automation. Some corporate security systems open every email in a sandbox. Very few of them write replies while doing it.
The consequence is that reply rate, measured over a reasonable sample of sends, is the closest thing email has to a truth metric. When reply rate goes up, engagement is up. When reply rate goes down, engagement is down. Every other metric I trust has to be triangulated against several other numbers before it means anything.
What reply rate predicts
The correlations I have watched, across programs, are not subtle. Programs whose reply rate is above two percent, sustained, tend to have inbox placement in the ninety-plus percent range. Programs whose reply rate is under half a percent tend to have inbox placement problems within three to six months, even if the current numbers look fine.
Reply rate is a leading indicator for deliverability, in other words. It moves before the placement metrics move. If your dashboard is watching reply rate and it starts to slip, you have a window of six to eight weeks to fix the underlying cause before the slip shows up as a placement problem.
Why nobody watches it
Two reasons. First, reply rates are small numbers, and small numbers are harder to reason about than large ones. A healthy newsletter might see two to four percent replies. Anything above six percent is exceptional. A dashboard that shows "2.1%" next to "OPENS: 48%" trains the human eye to focus on the larger number.
Second, most ESPs do not calculate reply rate natively. Opens and clicks are pixel-and-link tracked, calculated automatically, displayed on every campaign dashboard. Replies happen in whatever inbox the reply-to address points at, and getting them counted requires either a mail service that surfaces them, or a small manual habit.
The manual habit is worth building. It takes about five minutes a week. Open your reply-to inbox, count the replies to the last send, divide by the number of emails delivered. Write the number down. Do this weekly for two months and you will have a rolling trend that tells you more about your program than any tool metric.
How to move reply rate
Every technique I have watched improve reply rate can be described in a sentence. Ask questions. Write as though a specific person will read it. Use a reply-to address that is a real human. Do not sign the email from "The Marketing Team." Sign it from a person.
The best-performing newsletter I read on reply rate is signed by one specific writer. The email ends with a question. The reply-to address is that writer's real inbox. They reply to most emails within a day. Reply rate on that newsletter is over eight percent, sustained. This is not an accident of luck. Every design decision on that newsletter is oriented toward reply as an outcome.
Every metric on your dashboard is a proxy for something else. Reply rate is one of the very few that is not a proxy at all. It just is the thing.
What to do this week
Look at your last five sends. Count the replies. Divide by delivered volume. Write the number on a sticky note and put it on your monitor. Then, on the next send, add a real question at the end of the email — one that a real person could actually answer — and read the replies when they arrive. The number will move within a send or two.
And then keep watching it, week after week, and let it slowly replace open rate as the number that decides what you think about the health of your program. It will not steer you wrong.