Email Analysis

Not so random email marketing

In previous articles on testing the suggestion has been that you should divide your email list randomly. This is, statisticians reckon, impossible but the best you can do is good enough. However, there are times where it can be beneficial to be partial.

If, for instance, you are disappointed to find that your click-through rate has dropped slightly you will want to know how you can stop the slide and even reverse it. Before slashing prices or revamping the layout, you should interrogate your email marketing software to discover if a particular demographic had a lower click-through rate than the rest.

If you sell wine and have recently offered a product which goes well at a barbecue and you work out this is the group not buying then if this coincides with a month of dreadful weather you might well have found your answer. If, on the other hand, a holiday in Malta has been poorly received by those of the certain purchasing history you might well have thought would be most receptive, then an investigation is required.

Work out the demographics of those who had the lowest click-through rate. Accept that the percentage will not be constant throughout the group but despite this, treat this group as a whole.

Off the top of your head you can no doubt think of a few possible reasons for the lower click-through rate: layout, design, copy, images and more. How many of these you can test will depend on the size of the group, but let’s assume you can split the list into four. 

Retain one group as the control and then vary one aspect for each of the remaining three groups. The returns will show a series of percentages, the highest of which will give an indication of what you can test for this demographic for the next campaign.

This is no quick fix, but it is quicker than testing just one variable each time. 

In future campaigns you will have to separate this demographic and use a different email, the difference depending on the results of your tests. 'Einstein' The Email Analysis WizBotTreating subscribers as individuals is not fluff but an essential.

Sounds like too much hard work? WizEmail will allow you to segment your database analysis AFTER you've sent your email campaign – want to know how many Dave's between 25 and 30 years old clicked on a link in your campaign after buying Merlot in the last month? Not a problem 'Einstein' the Analysis WizBot has got you covered – clever little WizBots!




30 days full functionality - No credit card required - INSTANT ACCESS