Email Marketing Metrics: What to Track and What to Do With It

Every email platform gives you a dashboard full of percentages. Some of them tell you something useful about your subscribers, some mainly tell you about their email apps, and a few only matter when something has gone wrong. This guide walks through the metrics worth tracking, explains why open rates deserve less weight than they used to, shows how to run an A/B test that produces a trustworthy answer, and ends with a practical routine for turning numbers into decisions.

If you are new to email marketing overall, our email marketing guide covers the bigger picture first.

The metrics, grouped by the question they answer

It helps to sort metrics by what they tell you rather than reading them as one long list.

Core email metrics at a glance

Criteria	What it measures	How it is usually calculated	Main question it answers
Delivery rate	Emails accepted by receiving servers	Delivered ÷ sent	Did my email arrive at all?
Bounce rate	Emails rejected by receiving servers	Bounced ÷ sent	Is my list clean?
Open rate	Recorded opens	Unique opens ÷ delivered	Rough signal of attention (unreliable)
Click rate	Subscribers who clicked a link	Unique clicks ÷ delivered	Did the email prompt action?
Click-to-open rate	Clicks among those who opened	Unique clicks ÷ unique opens	Was the content persuasive? (affected by open inflation)
Conversion rate	Subscribers who completed a goal	Conversions ÷ delivered (or ÷ clicks)	Did the email achieve the business goal?
Unsubscribe rate	Opt-outs per send	Unsubscribes ÷ delivered	Is the content or frequency a fit?
Spam complaint rate	Recipients who marked you as spam	Complaints ÷ delivered	Am I damaging my sender reputation?
List growth rate	Net change in subscribers	(New − unsubscribes − removals) ÷ list size	Is my audience growing?

Delivery and bounces: did it arrive?

Delivery rate tells you how many emails were accepted by the recipient’s mail server. It does not tell you whether the message landed in the inbox or the spam folder; it only means it was not rejected outright.

Bounces come in two kinds:

• Hard bounces are permanent failures, such as an address that does not exist. Good platforms suppress these automatically after the first failure.
• Soft bounces are temporary, such as a full mailbox or a server that is briefly unavailable. Repeated soft bounces from the same address usually mean it should be removed.

A sudden jump in bounces almost always points to a list problem: an old import, a purchased list (do not do this), or a signup form being abused by bots. Our deliverability checklist covers how to prevent this.

Open rate: useful, but no longer reliable

Open tracking works by loading a tiny invisible image when the email is displayed. In 2021 Apple introduced Mail Privacy Protection, which lets Apple Mail users have remote content, including that tracking image, loaded privately in the background. The practical result is that many emails opened in Apple Mail are recorded as “opened” whether or not a person actually read them.

Other effects work in the opposite direction: some email apps block images by default, so real reads go unrecorded. The open rate you see is therefore a mix of inflated and missing data.

This also affects anything built on opens: open-based automations (“if opened, send X”), engagement segments, and click-to-open rate.

Clicks and click-to-open: did people act?

Click rate is now the most dependable engagement metric most senders have. A click requires a deliberate action, so it is a much stronger signal than an open. That said, some corporate security filters “click” links to scan them, which can create occasional false clicks, especially on B2B lists.

Click-to-open rate (CTOR) divides clicks by opens. It was traditionally used to judge the content of an email separately from the subject line. Because the open count is now inflated, CTOR is less precise than it used to be. It still works for comparing emails sent to the same list in the same period, since the distortion is roughly similar across sends.

Conversions: did it achieve the goal?

A conversion is whatever the email is for: a purchase, a booking, a webinar registration, a download, a reply. This is the metric closest to business value, and it usually needs tracking outside the email platform, for example UTM parameters on your links and goals in your website analytics, or a direct e-commerce integration.

Define the conversion before you send. “Get more engagement” is not measurable; “get 40 webinar registrations from this email” is.

Unsubscribes and spam complaints: the warning lights

Unsubscribes are normal and even healthy. People who no longer want your emails are better off leaving than staying silently or, worse, marking you as spam. Watch for spikes after specific emails, which usually point to a mismatch in content, frequency or expectations.

Spam complaints are far more serious. Mailbox providers use complaints as a strong signal of unwanted mail. Google states that bulk senders should keep their spam rate below 0.3%, and recommends staying well under that. Making unsubscribing easy is one of the best ways to keep complaints low.

List growth: is the audience getting bigger?

List growth rate is net growth: new subscribers minus unsubscribes, bounces and removals, divided by list size. It matters because every list decays over time as people change jobs, abandon addresses or lose interest. If you want ideas for improving the incoming side, see how to build an email list.

How to A/B test properly

A/B testing (also called split testing) sends two versions of an email to different parts of your list and compares the results. Done well, it replaces guesses with evidence. Done badly, it produces confident conclusions from random noise.

The rules that matter

1. Test one variable at a time. If version B has a different subject line, a different image and a different button, you will not know which change made the difference.
2. Decide the winning metric in advance. For subject lines, clicks are more trustworthy than opens now. For content and calls to action, use clicks or conversions.
3. Use a large enough audience. With small groups, the difference between versions is often just chance. If version A got 12 clicks and version B got 15, that is not a meaningful result.
4. Split randomly. Most platforms do this for you. Do not send version A to new subscribers and version B to old ones.
5. Send at the same time. Otherwise you are partly testing the send time.
6. Repeat tests that matter. One result is a hint; the same result three times is a pattern.

What is worth testing

• Subject line approach (a specific benefit versus a question, for example)
• Sender name (your personal name versus your brand)
• Call-to-action wording or placement
• Long-form versus short-form content
• Plain-text style versus a designed layout
• Send day or time (test separately from everything else)

A hypothetical example

This is a hypothetical illustration, not a real case. A bakery with 3,000 subscribers wants to know whether naming a specific product in the subject line beats a general “This week at the bakery” subject. They send each version to 15% of the list, measure clicks over four hours, and send the winner to the rest. Version A gets 41 clicks, version B gets 44. That gap is too small to act on, so they repeat the same comparison over the next three weekly emails before drawing any conclusion.

What to do with the numbers

Metrics are only useful if they lead to decisions. A simple monthly routine works for most small teams.

1. Compare against yourself, not the internet

Published industry benchmarks mix very different lists, industries and tracking methods. Your own history is a much better baseline. Keep a simple spreadsheet with each send’s date, email type, delivered count, click rate, unsubscribe rate, complaints and conversions.

2. Match the symptom to the likely cause

What you see	Likely cause	What to try
Bounces rising	Old or poor-quality addresses	Clean the list, add double opt-in, check forms for bot signups
Opens drop sharply across all emails	Possible spam-folder placement	Check authentication, complaints and recent content changes
Clicks low but steady	Content or call to action not compelling	Fewer links, one clear goal, test CTA wording
Clicks fine, conversions low	Problem after the click	Check the landing page, offer and checkout
Unsubscribes spike after one email	Topic or frequency mismatch	Segment by interest, set expectations at signup
Spam complaints rising	Unwanted or unexpected email	Make unsubscribe easier, review consent, send less to disengaged contacts

3. Segment before you judge

An average hides a lot. New subscribers usually behave differently from people who joined two years ago, and customers differ from prospects. Breaking results down by segment often shows that “the list” is fine and one group is the problem. Our guide to email list segmentation explains how to set this up.

4. Act on inactive subscribers carefully

Because open data is unreliable, define inactivity using clicks, purchases or site visits where possible, over a long enough window (several months, for most senders). Send a re-engagement email first, then remove or suppress people who still do not respond. A smaller, engaged list tends to perform better and costs less on platforms that charge by subscriber count.

5. Change one thing, then measure again

Pick the single weakest metric each month, make one change aimed at it, and check the result next month. This is slower than changing everything at once, but you will actually learn what works.

A quick metrics checklist

• Delivery and bounce rates checked after every campaign
• Spam complaint rate monitored, with a plan if it rises
• Click rate used as the main engagement metric
• Open rate treated as a trend, not a verdict
• Conversions tracked with UTM parameters or an integration
• A/B tests limited to one variable with a predefined winning metric
• Results compared against your own history, broken down by segment
• One improvement chosen per month

Most email platforms, including the ones we cover in our MailerLite review, report the core metrics above. What makes the difference is less the dashboard and more the habit of reading it regularly and acting on it.