Measurement basics · 6 min read
How Email Open Tracking Works
A plain-language walkthrough of tracking pixels, image requests, provider events, and what an open count can and cannot tell you.
Updated October 23, 2026 · Definitions and caveats are stated in the text.
The small image behind an open event
A traditional open pixel is a tiny remote image with a recipient- or message-specific URL. When a mail client requests that image, the tracking server records an image request. That observation is commonly labelled an open. The request is real; the conclusion that a person read the message is not guaranteed.
A single request can be caused by a person opening a message, a privacy proxy fetching images, a security scanner, a cache, or another automated system. If images are blocked, the reader may open the email without creating any pixel request at all. A count of requests is therefore not a census of people who read the message.
- Raw requests: every accepted image request in the selected window.
- Estimated unique opens: a deduplicated, classified estimate—not an identified-person count.
- Provider delivery events: separate, authenticated events such as sent or delivered, supplied by a connected email service.
Why mail privacy changes the signal
Some clients privately fetch remote images or prevent senders from learning whether a message was opened. That can create requests before a person reads a message, or suppress the relationship between the request and the actual reading moment. Other clients let people block remote images entirely. These are product privacy controls, not problems to work around.
A responsible report keeps the raw request separate from a classified estimate, describes automation and proxy handling, and avoids claiming that an open proves attention, intent, or identity.
Use opens as a limited diagnostic
Open observations can help answer narrow operational questions—for example, whether image requests changed after a template or sender-name change—when the campaign and measurement setup are comparable. Do not treat them as a reliable denominator for delivery, revenue, or human engagement.
For delivery, use authenticated provider events and the provider's sent count as the denominator. For downstream interest, examine link clicks and destination-site actions separately. If the denominator or integration is missing, report that the metric is unavailable rather than inventing a rate.
Questions
Does an open pixel prove that someone read my email?
No. It records a remote-image request. Proxies, scanners, caching, blocked images, and privacy settings can all break the link between a request and a human reading event.
Can I bypass image proxies to get a more accurate open?
No. Do not attempt to bypass mail-client privacy features or security systems. Use provider-supplied delivery events and clearly labelled downstream measurements instead.
Official references
Source links are click-counted by this site and redirect to the exact official page shown. They are not affiliate recommendations.