Your email watches you read it
A paper letter cannot tell its sender you've opened it. This obvious fact stopped being true of email so gradually that most people never noticed the moment. Today the newsletter you skim, the sales outreach you ignore, and quite possibly the message from your ex all know — with timestamps — whether you looked. The mechanism is almost embarrassingly simple, and knowing it changes how you read.
One invisible pixel
HTML email can include images, and images load from the sender's web server on demand. So the trick writes itself: embed an image one pixel square, transparent, unique to this recipient — <img src="https://track.example/open/you-specifically.gif">. The instant your mail app fetches it, the server logs the fetch: opened, at 11:42pm, from an IP that geolocates to roughly your neighbourhood, on a device whose type the request headers volunteer. Opened it twice? Logged twice. Forwarded it to a friend who opened it? That's in there too, mislabelled as you being very interested.
No exotic code, no exploit — just the web's ordinary image-loading, repurposed as a read receipt you never agreed to send.
Who's watching, and how much
Bulk marketing tracks opens as a matter of course; the newsletter industry's dashboards are built on it. Sales tools do it per-message — the "he opened your proposal three times this morning" genre. Studies over the years have put pixel-tracking in the majority of commercial mail, and the trackers themselves are third-party services, meaning your reading habits accrue at companies you've never heard of, cross-referenced across every list that uses the same tracker. Personal mail from actual humans rarely tracks — though "email tracking for Gmail" browser extensions exist and are marketed precisely for interpersonal surveillance, which is worth knowing about the people who install them.
What blocking does and doesn't do
Blocking remote images kills the report: no fetch, no log, no open recorded. That's the whole defence, and it's why every serious mail client now offers it and why the mailroom here blocks by default and shows you the count — the number doubles as a measure of how instrumented a message was. What blocking doesn't do: link clicks are tracked separately (those wrapped URLs report regardless), and some providers proxy images rather than block them, which hides your IP and timing but still confirms life at the address. Blocking also costs you the pictures, which is why the honest design is a per-message "load them anyway" — surveillance as an informed choice rather than a default.
For saved mail files, the stakes are quieter but real: an old .eml opened years later would still phone its pixel home, telling some marketing database that a long-dead campaign just scored an open — and telling whoever inherited that domain that your address is alive. Old letters deserve to be read in peace. Block first; load the pictures when you actually want them.