Email open tracking in cold email: keep it or turn it off?
Open tracking relies on a hidden image that Apple Mail loads before anyone reads. Why open rates mislead in cold email and what to measure instead.
On this page
Email open tracking puts a tiny invisible image in each message and records an open when a mail app downloads it. In cold email the number it produces is no longer reliable: Apple Mail fetches remote images in the background when a message arrives, before anyone reads it. Turn open tracking off for cold campaigns, judge copy by replies, and judge deliverability by bounces, complaints and placement tests.
Below: how the tracking image works, why opens mislead, the deliverability question, and what to track instead. The cold email guide covers the rest of a sending setup.
How does email open tracking work?
The sender adds an image to the HTML part of the message, usually one pixel, transparent, with an address unique to that recipient and that email. When a mail app displays the message and loads its images, the request reaches the sender’s tracking server, which records an open with the time and the address that asked.
Two consequences follow. Open tracking needs HTML, so a plain-text email cannot carry it. And it counts image downloads, which is a proxy for reading that only holds if images load when, and only when, a person looks.
Why open rates mislead in 2026
Apple Mail loads the image before anyone reads
Apple’s Mail Privacy Protection changes when that download happens. With Protect Mail Activity on, Apple says your IP address is hidden from senders and “remote content is privately downloaded in the background when you receive a message”, instead of when you view it. Apple’s privacy page adds that otherwise senders can learn when and how many times you opened their email and your IP address.
For a sender, every message delivered to those mailboxes can register an open at delivery time, read or not. The open rate rises, and the opens of people who truly read the message become impossible to tell from the rest.
Scanners and prefetchers fetch content too
Corporate mail filters, link scanners and some apps fetch a message’s content to check it before or instead of a person. Each fetch can look like an open or a click. Broadkite’s tracking service recognizes prefetch requests and common scanners and does not count them, but no tracker can catch all of them.
Reading in the sender’s own tools
When your team reads a sent message in a shared inbox, the image can load again. Broadkite strips its own tracking images from messages shown in its unified inbox, so reading a thread there never counts as an open.
Does open tracking hurt deliverability?
Honest answer: nobody outside the mailbox providers knows for sure. No provider publishes that a tracking image lowers inbox placement, and the tests sellers publish on either side are not independent.
What is certain:
- A tracking image is a remote image and a tracking address inside a message that is supposed to read like a personal note. Personal notes rarely carry either.
- Open tracking requires HTML. A plain-text email cannot carry it, and plain text is the most common shape of one-to-one mail.
- The number it gives you is unreliable, as shown above.
That is enough reason to leave it off for cold email. You give up a misleading number and lose nothing you can act on.
What to measure instead
| Question | Measure |
|---|---|
| Does the copy work? | Reply rate, positive reply rate, meetings booked |
| Is the list good? | Hard bounce rate, replies saying wrong person |
| Are we reaching the inbox? | Spam complaint rate, inbox placement tests on seed mailboxes |
| Are mailboxes healthy? | Warmup placement at Gmail, Outlook and Yahoo, provider throttling |
Replies are the one signal no privacy feature inflates: a person had to read the email and write back. Delivery rate is not inbox placement explains how placement is measured without opens, and the complaint rate that gets you throttled covers the thresholds.
What about click tracking?
Click tracking rewrites each link so it passes through a tracking address before reaching the destination. It answers a useful question, which link people actually wanted, and a click requires a person far more often than an image load does.
If you use it, send links through a tracking domain on your own sending domain, such as t.yourdomain.com, rather than a shared domain used by many senders. In cold email, one link at most is a good rule, and many first emails work best with none.
How to set it up in Broadkite
Each campaign has its own open and click tracking toggles under its settings. Turn opens off for cold campaigns. If you choose Plain text only, the campaign goes out with no HTML part at all, and open and click tracking are off whatever the toggles say, because neither can work without HTML.
With tracking off, a campaign is judged on replies, bounces and complaints, and each mailbox’s health on warmup placement. Keep sends around 50 a day per mailbox, as how many cold emails per day per mailbox explains.
Questions
How does email open tracking work?
The sender adds a tiny invisible image with an address unique to the recipient. When the mail app downloads that image, the sender records an open, with the time and the address that fetched it.
Why are email open rates inaccurate?
Apple Mail Privacy Protection downloads remote content in the background when a message arrives instead of when it is read, so the image loads whether or not the person opens the email. Security scanners and prefetchers can load it too.
Does open tracking hurt cold email deliverability?
No mailbox provider publishes that a tracking image lowers placement, and no independent test proves it. What is certain is that it adds a remote image and a tracking address to a message that otherwise reads like personal mail, and that the number it produces is unreliable.
What should I measure instead of open rate?
Replies, positive replies and meetings for the copy, and bounces, spam complaints and inbox placement tests for deliverability.
Was this post helpful?
Okay
What helped most?
Keep reading
How many cold emails can you actually send per day?
The safe number is per mailbox, not per account: 50 a day by default, a warning above 100, a 600s gap between sends. The math to plan any daily volume.How many mailboxes and domains do you need for the volume you want?
Planning cold email volume comes down to per-mailbox budgets. Work backward from your daily target to the mailboxes, domains, and IP spread you actually need.Your send limit is a mailbox budget, not a server setting
People keep asking us to raise their daily limit. The honest answer is that the limit they are asking about does not exist.