Open a marketing email and, in most cases, you have just told the sender the time you opened it, roughly where you were, and what you read it on. Nothing was clicked. The reporting is built into the message itself.
The mechanism
An HTML email can load images from the internet, the same as a web page. A tracking pixel is an image deliberately made invisible — one pixel square, transparent, or hidden behind styling — with a URL unique to you:
<img src="https://track.example.com/o.gif?u=8f3c1a2e" width="1" height="1">When your mail app fetches that image, the sender's server logs the request. Because the identifier is unique to your copy of the message, the request is not anonymous traffic — it is you, opening that specific email, at that moment.
What the sender actually learns
- That you opened it, and when. Down to the second, and every subsequent time you reopen it.
- Your IP address — which places you geographically, usually to a city, and identifies your network.
- Your device and mail client, from the user-agent: iPhone versus desktop, which app, sometimes which OS version.
- Read duration, in some implementations, from how long the connection stays open.
- Forwarding — if you forward the message and someone else opens it, that loads too, from their IP.
Aggregated across a mailing list this is ordinary marketing analytics. Aggregated across you, over years, it is a log of when you are awake, when you travel, which devices you own and which senders you actually read. Sales tools sell exactly this: a notification the moment a prospect opens an email, so a call can follow within the minute.
What does not work
Reading in plain text. It works, and it also breaks every legitimate email you receive. Very few people stick with it.
"Ask before displaying images." Better, but it makes you the filter — a per-message decision, dozens of times a day, where the safe answer is always "no" and the useful answer is often "yes". People click through out of habit within a week.
Image proxying. Gmail routes images through Google's servers, which hides your IP and device from the sender. That is a genuine improvement, but the pixel still loads: the sender still learns the message was opened, and Google learns it too. It moves the disclosure rather than preventing it.
What does work
Stripping tracking pixels on arrival, at the server, before the message is ever rendered. Real images load normally; known trackers are removed from the message body. Nothing is requested from the sender's server, so there is no open event to log, no IP disclosed, and no decision for you to make on each message. This is how Mailotte's Shield works, and it is why the privacy dashboard can show you a count — every blocked pixel is one that was removed rather than fetched.
Pair it with aliases
Tracker blocking stops the reporting on messages you receive. It does not address the other half — how your address got onto the list, and who sold it. That is what aliases are for: a distinct address per service, so when tracking-heavy mail starts arriving you can see precisely which company leaked you, and switch that address off without changing your real one.
Is any of this legal?
In the EU and UK, tracking pixels are broadly treated like cookies under the ePrivacy rules: they require consent, and "we assume you consented by subscribing" is a contested position that regulators have pushed back on. Enforcement is patchy and the practice is near-universal. In practice, blocking at your end is the reliable remedy.