To understand who's included in the open rate, it helps to understand how tracking works. So here goes: The HTML version of your mailing includes images, like your logo, that display when the recipient opens the email. Well, there's an invisible image in the HTML mailing that doesn't display but which gets 'called' just like the real images. A person opens your email, and the mailing calls our server to request the images. 'Hey, I need an image,' it might say. Our server delivers the image and counts the email as having been opened. (Our methods are a bit more complex than this, but you get the idea.)
We can also track opens for anyone who receives the plaintext version of your mailing -- remember that we automatically deliver the HTML version whenever it can but has a plaintext version ready in case someone's email client isn't set up to receive HTML -- and visits at least one of the links you've included in the email.
So in short, two actions -- viewing the images in your email, or clicking through to your website using one of your mailing's links -- are trackable as opens.
If someone receives the plaintext version and doesn't follow any of the links. Or if someone downloads his or her emails and views them offline. In either case, the recipient could have received and read your email without showing up in your open numbers. Generally, this should be a small percentage, but it does serve as a reminder that your open rate is an approximate number, as best we (or anyone) can track.
Open and click-through rates can vary wildly based on your organization, your industry, the kind of mailings you send and the kind of audience you've got. Reports from people like DoubleClick put the average open rate for consumer-oriented mailings in roughly the 35-45% range, with business-to-business open rates sometimes going higher. Again, it varies by industry and all the other aforementioned factors. Our advice is to experiment with things like subject lines, message and frequency, and see how all of your response numbers stack up against the only benchmarks that truly matter -- yours.