You will have been told by everyone, including me, that when testing you should only change one feature at a time. Do you wonder if your call to action button is somehow not optimum for a specific type of person in your email marketing list? Then wonder no more. A simple A/B test and you have the answer.

There are times, however, where A/B testing takes too long and, in some cases, will not provide an answer. We glibly talk of tone of voice, and it being just as important to have an image that refers specifically to what is on offer as the text. But if you change one aspect, such as the typeface of the text, what if it clashes with the image? A bold font might well contradict a reassuring message. Change both and you go against received wisdom.

In order to test an idea it can be more time efficient to experiment with a design of email which differs entirely from the one you are testing it against. It is best to have a reason, rather than randomly changing for the sake of it. Let us assume that your product is used by people in a variety of locations: different towns, counties, countries.

In the UK there are local differences in attitude and desires. This is more so for people in different countries. You might suspect, and a suspicion is enough, that one country prefers a more didactic approach to selling compared to another which likes to be convinced by persuasion. This will not be a case of just using bullet points and short paragraphs. The whole basis of the emails will need to be different. 

So change the Subject Line, the Headline, the text, the images, the call to action, the landing page, and the checkout; even consider modifying your logo. The trick is to ensure that each change is a targeted at the group you are after. Split your list for the specific group you have in mind and then sit back, although not for long.

Use this type of testing infrequently as it requires further simple A/B testing of all normal features once the basic premise has been proven.