E-Shot Media Practical guides to email newsletters

Measure

Measuring results: opens, clicks, conversions and unsubscribes

Why open rates stopped being reliable after Apple Mail Privacy Protection, and what to measure instead.

Updated 6 min read

Every email service has a reports page full of percentages. Some of them tell you useful things about your newsletter; others are noisy guesses dressed up as precise numbers. Knowing which is which saves you from changing a good newsletter because of a bad statistic.

The biggest change in the last few years is that the open rate, once the headline figure in every report, can no longer be taken at face value. This guide explains why, then sets out the measures that are more reliable, how to track what happens after the click, and a simple monthly scorecard you can keep in a spreadsheet.

Why open rates stopped being reliable

An “open” is recorded when a tiny, invisible tracking image in your email is downloaded. That was always an approximation: people who read your email with images switched off were never counted, and people who glanced at it in a preview pane often were.

In 2021 Apple announced Mail Privacy Protection, which arrived with iOS 15 and the matching versions of macOS. When a reader turns it on in the Apple Mail app, Apple downloads the images in their emails in advance, through its own proxy servers, whether or not the person ever opens the message. To your email service that looks like an open, often at an odd time and from an Apple location rather than the reader’s own. So:

  • Open rates are inflated by however many of your readers use Apple Mail with the feature turned on.
  • Open times and locations are unreliable for those readers, which undermines “best time to send” reports.
  • Anything triggered by opens is affected, such as resending to non-openers or tagging people as “engaged” because they opened.

Other things distort the figure in both directions. Some mail apps and workplaces block images by default, so real readers go uncounted. Some privacy tools strip tracking images. Many email services now try to identify machine opens and filter them out, but each does it differently, so you can’t compare open rates between tools either.

Clicks: still useful, with a caveat

A click means someone (usually) chose to do something. Click rate, the share of delivered emails that got at least one click, is a far better guide to interest than opens. Look at which links were clicked as well as how many: that tells you which topics your readers care about.

The caveat is security scanners. Some business email systems open and “click” every link in incoming mail to check it for malware before a person sees it. That can produce a burst of clicks on every link within seconds of delivery. Many services try to filter these out. If a campaign shows the same subscriber clicking every link at once, treat it as a robot, and be careful about automations that fire on any click.

Two click measures are worth recording:

  • Click rate: unique clickers divided by delivered emails.
  • Clicks per link: which items or calls to action drew interest.

Click-to-open rate (clicks divided by opens) used to be popular, but because it depends on the open count it has inherited the same problems.

Conversions: what happens after the click

For most organisations the email isn’t the goal. The goal is a sale, a booking, a donation, an event registration or a membership renewal. Measuring that means following the reader from the email to your website.

Tag your links

UTM parameters are short labels added to the end of a link so your website’s reporting can tell where a visitor came from. A tagged link looks like this:

https://example.co.uk/events/summer-fair?utm_source=newsletter&utm_medium=email&utm_campaign=2026-06-june-issue
  • utm_source: where the link appeared (your newsletter).
  • utm_medium: the channel (email).
  • utm_campaign: which issue or campaign. Use a consistent pattern, such as the year, month and a short name.

Many email services can add these automatically. Keep them lower-case and consistent, otherwise “Newsletter” and “newsletter” show up as two different sources.

If you don’t run website analytics

Plenty of small organisations don’t, and you can still measure conversions:

  • Use a dedicated discount code, booking reference or landing page for newsletter readers.
  • Ask “How did you hear about this?” on booking or donation forms.
  • Check whether your shop, ticketing or donation platform records the referring source.
  • Note spikes in orders or bookings in the day or two after a send.

Example

A hypothetical theatre company sends its season announcement with a booking link tagged utm_campaign=2026-09-season. Its ticketing system reports bookings by source, so a week later it can see how many bookings came from that email, compared with social media and the website. That figure matters more than the open rate.

Replies, unsubscribes and complaints

Replies

Replies are a strong signal of a healthy relationship, and they tend to help your deliverability too. Use a reply-to address that a person actually reads, and occasionally invite replies: ask a question, request suggestions for the next issue. Count them, even roughly.

Unsubscribes

Some unsubscribes are normal and healthy. People move on, change jobs or lose interest. Watch for changes rather than the number itself: a spike after a particular send tells you something about that email’s content, frequency or audience. If your service asks unsubscribers for a reason, read the answers.

Spam complaints

Complaints are the measure that most affects whether your future emails arrive. Gmail and Yahoo expect complaint rates to stay well under 0.3%, as covered in deliverability basics. Note that many mailbox providers don’t report every complaint back to email services, so the figure in your dashboard may be an undercount. Treat any rise as serious.

Bounces

A rising bounce rate usually means old or poorly collected addresses. Hard bounces should be suppressed automatically; if yours aren’t, fix that first.

List growth

Track how many people joined, how many left, and where new sign-ups came from (website form, event sign-up sheet, checkout opt-in, a particular lead magnet). Net growth on its own hides useful detail. A list that gains many and loses many is behaving differently from one that gains a few and keeps them.

A simple monthly scorecard

You don’t need a dashboard. A spreadsheet with one row per campaign and these columns covers most needs:

ColumnWhat to record
Date and campaignWhen it went and a short name
DeliveredSent minus bounces
Click rateUnique clickers as a share of delivered
Top linkThe most-clicked item
ConversionsSales, bookings, donations or sign-ups you can attribute
RepliesRough count
UnsubscribesNumber and rate
ComplaintsNumber and rate
BouncesHard and soft
Opens (trend only)Recorded, but not used for decisions
NotesAnything unusual: subject line, timing, a news event

Once a month, add a list summary: total subscribers, new sign-ups by source, and leavers.

Compare against yourself

Industry benchmark reports are tempting, but they mix very different senders, lists, tools and ways of counting, and many were affected by Apple’s change in ways that are hard to untangle. Your most useful comparison is your own previous campaigns, sent to the same list with the same tool.

Look for trends over several sends rather than reacting to one. If you want to test something, change one thing at a time (subject line, send day, layout) and give it a few issues before drawing conclusions. On a small list, differences of a percentage point or two between two sends are often just chance.

If you are setting up automated emails, the automations guide explains how to judge engagement without relying on opens. Terms used here are defined in the glossary.