Email opens aren't actually measured - what gets measured is an invisible image loading. We explain, without unnecessary technicality, why today's open-rate number is inflated and which metric holds up in its place.
When you pick a newsletter tool, the first number it shows you is the open rate - the percentage of people who opened your email. It looks like a hard fact: 42% of recipients read it. In reality it's an estimate, and one that has fallen apart quite a bit over the past few years.
This isn't a detail for the technically minded. This number is what decides whether a topic landed, whether to change the subject line, and whether the whole campaign made sense.
Opens aren't actually measured
For your email program to know you've read something, a small piece of code would have to run inside the email. That's called JavaScript - the language interactive things on websites are written in. When you click "add to cart" on an online store and the page updates without reloading, that's JavaScript at work.
But JavaScript doesn't work in email. Gmail, Outlook and Apple Mail all deliberately block it - if they allowed it, anyone could send an email that runs itself inside your inbox. That's a security risk nobody wants.
So the entire email industry uses a workaround instead: the tracking pixel. A transparent image, a single pixel in size, gets embedded in the message. It's invisible, but it has its own address on the sender's server:
<img src="https://service.com/track/open?message=123&recipient=456"
width="1" height="1" alt="" style="display:none" />You don't need to understand the code. What matters is what happens: as soon as your email program starts downloading the images in a message, it fetches this dot too. That tells the sender's server that someone downloaded the image, and it logs an "open."
It also logs a few other things that get sent automatically with every download from the internet: the time, an approximate location based on IP address (usually accurate to city level), and the type of device and mail program.
And here's the catch. It doesn't record that you read the email. It records that an image was downloaded. Those are two different things - and in recent years they've drifted even further apart.
Four opens, one person.
Three reasons the number doesn't add up
Apple pre-downloads images itself. Since 2021, Apple has had a feature that pre-downloads images for every incoming email through its own servers - deliberately, so the sender can't tell who opened a message or when. It's a privacy protection, and it works. For senders, though, it means Apple reports an "open" even for messages nobody actually looked at. And because Apple Mail now handles more than half of all opens worldwide, that's a huge chunk of any database. For a typical company contact list, this inflates the open rate by roughly 15 to 35 percentage points. A campaign reporting 50% might really be sitting at 25-30%.
Bots. Corporate antivirus tools and security filters automatically open a message and click through its links before it ever reaches a person - checking for anything malicious. In the statistics, this looks like an extraordinarily engaged reader who read everything and clicked every link two seconds after the email went out.
Images turned off. The opposite error. Anyone who has automatic image downloads turned off in their mail program - and a lot of people do - won't show up in the statistics at all, even if they read the email three times.
The first two reasons inflate the number, the third deflates it. Unfortunately they don't cancel out. You just don't know how far off you are, or in which direction.
A click is something different
Links are tracked in a similar way, but with one crucial difference. Before sending, every link in the email gets rewritten so it doesn't point straight to your website, but first to a tracking server. When a recipient clicks, they pass through that server for a fraction of a second - it logs the click and immediately forwards them to where they were actually headed. The user notices nothing.
The difference from the image is who performs the action. An image downloads by itself. A link has to be clicked deliberately - with a finger or a mouse. That's why clicks, not opens, are treated as the main metric today.
It's calculated two ways:
CTR (click-through rate) tells you how many people clicked out of the total number of delivered emails. It's the result of the whole campaign - including whether the subject line caught people's attention and whether they opened the message at all.
CTOR (click-to-open rate) is calculated only from people who opened the email. It tells you something different: whether the content engaged the people who actually opened it.
The word "unique" in the numerator isn't a minor detail. If every click counted, one indecisive person who comes back to the email five times could single-handedly pull your statistics up.
One thing to watch out for: CTOR has an inaccurate open count in its denominator, so it inherits that inaccuracy - when opens are inflated, CTOR comes out lower than it actually is. As a ratio, though, it still stays useful when you track it over time and compare your own campaigns against each other - not against someone else's number from a benchmark list.
What to track instead of opens
Useful metrics have one thing in common: a human decision stands behind them, not the behaviour of a mail program.
- Which link got clicked - not just "someone clicked something," but which part of the content actually interested people.
- Unsubscribes - the fastest and most honest feedback on how often you write and what about.
- Replies - by far the most valuable signal for smaller campaigns.
- What happened on the website after the click - did they fill in a form, buy something, or just look and leave.
- Bounced emails - how many addresses no longer exist. Boring, but it affects deliverability more than any subject line ever could.
Open rate still makes sense as a trend: if it drops by half from one month to the next on the same list, something happened. It doesn't make sense as an absolute number you show someone as proof of success.
Numbers backed by a human decision.
What to take away when choosing a tool
If someone is selling you "precise measurement of email opens," they're selling you something that technically can't be measured precisely. That's not necessarily dishonesty - more that the field changed faster than sales pages could be rewritten. But it's worth knowing before you start making decisions based on that number.
That's why in Apexloop we evaluate campaigns mainly through clicks and what happened afterwards. Sending also gets logged against each contact's activity - alongside phone calls and notes from colleagues. So when someone gets back to you, they're not an anonymous line in a campaign report. They're a specific person, and you can see exactly what you sent them and when.
A campaign gets logged against the client, not just in a report.
In the end, that's more useful than a percentage nobody can break down into how much of it is people and how much is antivirus software.