Newsletter Tracking Pixels and How to Block Them
A writer I had corresponded with twice mentioned, in a friendly way, that he could see I had opened his last issue four times and that one of those opens came from a hotel network in Lisbon. He meant it as a compliment.
A writer I had corresponded with twice mentioned, in a friendly way, that he could see I had opened his last issue four times and that one of those opens came from a hotel network in Lisbon. He meant it as a compliment. I had not told him I was travelling, and I had not clicked a single link in that email. All of that came out of a transparent image about 43 bytes in size, sitting at the bottom of his issue where a footer rule would normally be.
I keep around four hundred subscriptions, so I have thought about this more than most people need to. The part that surprised me was not that the tracking exists. It is how much of the common advice about stopping it does nothing, and how many of the settings people trust were quietly changed years ago in ways that help the sender more than the reader.
What the one pixel image actually sends
A tracking pixel is a normal image tag pointing at a unique address. The file is usually a 1x1 transparent GIF, and the point is never the picture. The point is that your mail client has to ask a server for it, and the address contains an identifier tied to you and to that specific issue.
When the request lands, the sending platform writes down the time, the identifier, your IP address, and the user agent string your client sends. From the IP it derives a city, sometimes accurately and often not. From the user agent it derives your client and operating system. Load the message again tomorrow and it logs a second open, which is how a sender ends up with a figure like four.
Link tracking is a separate mechanism and people conflate the two constantly. Platforms rewrite every outbound link so it points first at their own redirect domain, then bounces you to the real destination. That rewriting has nothing to do with images and is not affected by any image setting anywhere. It is also the more revealing signal of the two, since a click says what you cared about rather than merely that the message loaded.
Neither of these requires anything unusual or shady on the sender's side. Every mainstream newsletter platform turns them on by default, and most writers have no idea what the dashboard is collecting on their behalf.
Turning off images is not the same as not being tracked
The advice you see everywhere is to disable remote image loading and consider the matter closed. I followed that for years. It covers maybe half the problem, and in one common case it covers none of it.
Start with Apple Mail Privacy Protection, which arrived with iOS 15 in September 2021 and is on by default. It does not block the pixel. It fetches the pixel through Apple's proxy, immediately, whether or not you ever look at the message. Your IP and your device are hidden, which is genuinely useful. The open is still logged, and it is logged at delivery rather than at reading. That is why newsletter open rates jumped into the 50 to 70 percent range across the industry that autumn and have been close to meaningless ever since.
Gmail has proxied and cached remote images since 2013. The first load goes through Google's servers, so your IP is masked, and because the image is cached, repeat views mostly do not generate new requests. The initial open is still recorded.
And in every one of these setups, clicking any link still walks you through the sender's redirect. So someone who disabled images, feels covered, and clicks three articles in an issue has handed over a more detailed profile than someone who left images on and read without clicking. If you want the tracking reduced rather than relocated, the images setting is the smallest part of the job.
What each mail client actually does
| Client | Images by default | Still leaks | Worth changing |
|---|---|---|---|
| Apple Mail with Privacy Protection | Fetched via proxy at delivery | Every open, logged early | Turn off remote content entirely in Settings |
| Gmail web | Proxied and cached | First open, link clicks | Set images to ask before displaying |
| Gmail mobile app | Proxied, no ask option on some versions | First open | Little to change, route elsewhere |
| Outlook web | Loaded directly | IP, city, device, every open | Block external content in options |
| Proton Mail | Blocked, trackers counted | Link clicks only | Nothing, it is the sane default |
| Thunderbird | Blocked with a yellow bar | Link clicks only | Nothing |
| Fastmail | Loaded, with a per sender toggle | Opens unless you change it | Switch to block by default |
| RSS reader via a mail bridge | Depends on the reader | Usually nothing per subscriber | The cleanest option |
Proton and Thunderbird are the only two on that list where I did not have to change anything. Proton even shows a count of blocked trackers per message, which is how I learned that a 900 word weekly issue I like carries five of them.
The setup I ended up with
Four steps, maybe twenty minutes, and most of it is one time work.
First, subscriptions go to a tagged address rather than my real one, which is also how I know who sold my address when something odd shows up. I wrote the longer version of that in using an email alias for newsletters, and the tracking benefit is a side effect rather than the main point.
Second, remote content off at the client level, not per message. In Apple Mail that is the Privacy Protection panel, where you switch off the proxy and the loading together rather than leaving the proxy doing its early fetch.
Third, and this is the one that did the most work for me, the things I read regularly do not arrive by email at all. A bridge turns the newsletter into a feed, my reader pulls it like any other source, and the per subscriber identifier mostly stops applying. Kill the Newsletter has done this for me free for years. The method and the trade offs are in reading newsletters in your RSS reader instead of email, and the honest trade off is that rich HTML layouts sometimes arrive looking plain.
Fourth, for the handful of links I do want to open, I copy the link and strip the redirect by hand when it matters, and otherwise I accept it. I am not going to pretend I do that every time. Maybe one link in twenty feels worth the effort.
What the sender loses, and why I am only half sorry
A small independent writer with 1,200 subscribers reads open rates to decide whether a subject line worked and whether to keep going at all. That is a real thing I am taking away, and I think the honest position is to admit it rather than to insist that nobody should ever measure anything.
But the measurement has already collapsed on its own. Since the Apple change, an open rate reported as 58 percent might represent anywhere from 15 to 50 percent of people who read a word of it, and the sender has no way to tell which. Writers who still make decisions from that number are making them from noise. The ones I trust have moved to reply counts, paid conversions, and simply asking readers a question at the end of an issue.
So I reply. Two or three sentences to a writer I value is worth more than any pixel, and it is the one signal that cannot be faked or inflated. If you are curious how I decide which subscriptions survive a cull in the first place, the rest of that thinking lives in the newsletters section.
FAQ
Can a sender tell if I blocked their tracking pixel?
Only indirectly. They see no open recorded, which looks identical to an unopened message. There is no separate signal that says you blocked it, which is why plain non opens and blocked opens sit in the same bucket on every dashboard I have seen.
Does marking a newsletter as read in a reader trigger the pixel?
Not usually, because most readers fetch the article content rather than a per subscriber email. If your reader loads remote images inline and the feed was built from the email version, it can still fire, so check the reader's image setting.
Are tracking pixels the reason newsletters land in Promotions?
They are one input among many, alongside link rewriting and bulk sending patterns. The filtering fix is separate from the privacy fix and I covered it in newsletters going to spam and how to get them back.
Is any of this illegal?
Generally no, though consent and disclosure rules differ by country and some platforms now surface a tracking notice at signup. I am a feeds editor rather than a lawyer, so treat that as context rather than guidance.
More in Newsletters
Newsletters
How to Find an Old Newsletter Issue You Half Remember
Someone asked me last month for the newsletter that explained why a particular bookshop in Lisbon shelves by publisher instead...
Newsletters
How to Save Newsletter Issues Before They Disappear
In March a paid newsletter I had read for three years sent a two line goodbye note, and about six weeks later its archive...
Newsletters
How to Unsubscribe From Newsletters in Bulk Without Regret
Your inbox shows 4,000 unread messages and most of them are newsletters you signed up for to get a 10 percent coupon in 2021.