We, and I include me in this, speak a great deal about using email marketing software to define the design and other circumstances of your campaigns. After all, you might well feel you know more about a subscriber’s best interests than they do themselves.
Split testing will define the best time to send a marketing email to a particular section of your email list as the returns will show the highest open rate, the best completion rate, and on the negative side, the time which generates the highest percentage of unsubscribes.
There is another method of deciding the preferences of a subscriber: ask them. This is not an alternative so much as an addition.
You might think this is a bit of a waste of effort as our data and returns provide a much more accurate method. However, there are points in its favour which may seem a bit obscure at first glance.
For instance, you have decided on html rather than plain text. Your reasons for this are there in your data. You get a higher completion rate with html for that group, so no argument. To get this information you must have tested it in some way.
The results apply to the group you tested it on, classified by one or more bits of information. This does not mean the results apply to every individual, only to the group as a whole, and the accuracy of the returns will depend to a great extent on how well you split your list.
Your returns are accurate without them being applicable to every person in that grouping. If a particular individual’s preferences are within 90% of the norm, you have done very well. So what about the other 10%?
A short form might well be of benefit to both sides. You could ask a small test group which type of email they prefer to receive. Whatever the replies, they will give useful information.
For instance, if few in the test group reply, then it can be seen as unimportant to most. Not a waste of time of course, you know something you didn’t before. If most replies fall into line with your data predictions then you can feel reassured you are interpreting it correctly.
If a number of the replies agree to something which is outside that of your data, then you have some work to do. Firstly, check to see if those who replied with a particular preference have something in common, such as in B2B their type of business, in B2C their purchasing history. If so then you have a distinct pointer for further refinement of your lists.
If you have a large range of products and are wary of cluttering an inbox with something which does not apply to your subscriber, then offer the choice of types of product. You can still send the occasional one outside that limitation.
One word of warning. If you do offer, for instance, a choice of time the email will drop into their inbox, if they state 10 am Thursday mornings then you are bound to deliver at that time.
