The question began with a client inquiring about a high percentage of users clicking on the "Trouble Viewing?" link to the online version in a particular e-mail. Clicks on the Trouble link are included in a report showing clicks on all links in the e-mail, which is a problem. First, the report percentages represent link clicks/all clicks. This means an e-mail with a lot of click activity will show a lower percentage of Trouble clicks than a less-popular e-mail (E-mail B below), although the same number of people may have clicked to view the online version.
What measure makes sense? Trouble clicks/Delivered? This denominator includes people who never opened the e-mail in the first place and yields a number too small to be useful. Trouble Clicks/Opens? Since the Open rate is flawed, I try never to use it as the basis for another stat. The Trouble clicks (along with clicks to view the text version) should really be counted as opens. They represent a user clicking to get a readable version. But they're really "super opens" because they are a) more reliable and b) represent a second action (post-open) that indicates a higher level of user engagement.
Despite the name I've chosen, "Trouble," a click on this link is really a good thing. It indicates a user that is getting to the best version of an e-mail--full screen, with images--the one we intended them to see.
This leads to interesting questions and opportunities for e-mail optimization.
Are those who get to the online version by clicking the Trouble link more responsive than those who don't? If so, what gets them to the online version? Will an e-mail designed for preview panes/blocked images deliver better results than the e-mail that reveals nothing and requires a click? Does it depend on (e-mail) brand loyalty? Some split tests are in order! I'd welcome volunteers with other examples of, or comments on this subject.
The E-mail Diva
Send your questions or submit your e-mail for critique to Melinda Krueger, the E-mail Diva, at email@example.com. All submissions may be published; please indicate if you would like your name or company name withheld.