Adventures in Mobile Marketing

I was pleased to read in “No Man is an Island” about methods to measure open rates from images-off email clients. The dilemma is that now with more and more email clients not displaying images, one of the tried and trusted methods of response email analysis – “opens” – is no longer 100% valid. What if a number of customers are viewing emails in the preview pane of Yahoo Beta or Gmail? You won’t know. More intriguing, to me, though, is that customers could be clicking and interacting with the site, and with the traditional way of measuring such things, the average marketer won’t know it. Also, your measurements for the success of campaigns will be inaccurate.

The technical aspect of “open” rates is also their limitation. They depend on the email being in HTML and having a graphic, a 1-pixel-wide image. When the web servers report on the load of that image by an internet user, it records that interaction. Well, if the email clients are not loading images, the web server log won’t have a record of the user interaction- since there wasn’t any, and there will be no open. It’s like a text-only email open.

There is way to estimate this, though. Split your clickthroughs into two piles: the opens and not opens. Before, we never did this as we assumed that to click through meant to open. Now, users can clickthrough without an open.

After determining the clickthrough/open rate, then look at the number of click-throughs that did not have an open record. Assuming the same percentage of clickthroughs/opens, you can determine a probably number of opens by images-off email clients.

10000 clicks
—–
50000 opens (image enabled)

500 clicks
—-
? opens

(assuming same rate of click for image vs. non-image)
2500 non-image opens

At this time, I don’t know of many email vendors that have reporting abilities equipped iwth giving you clickthrough rates minus the sessions that did not have opens associated. To get this data you may have to create a special data request. If you know of vendors who are doing this, please write in.