Let me confess from the beginning that all I’m suggesting is that following the data blindly can often be a poor option. For email marketing, data is a tool, nothing more. It needs to be kept sharp, but be careful of stabbing yourself.

Let’s say that you want to split test a new email design. You have perused your free email marketing templates and decided on one which reflects and reinforces your credentials as a modern and forward-looking company. The returns from your next email marketing campaign will prove or negate its efficacy.

It is difficult to judge how a redesign will be received, but you decide that the click through rate will show if your customers are happy with what you’ve done. You hope your new design will ensure the buying process is swifter, easier to follow and with fewer distractions; the holy grail of email design.

You will await the returns with interest, so your feelings when the results show the A group’s 1.2% drop over the norm are easy enough to predict. That they are 1.4 below the B group’s returns will ensure you see it as a failure; all that work for nothing. You will wonder if anyone in your company noticed.

However, much to everyone’s relief, there’s a however. 

Change can be unsettling to customers. If your previous email marketing campaigns looked the same, then some people will be confused by this new style. The whole point of the change was to make things slicker, but your subscribers might have been intrigued by the new style. It is, after all, something that should be transparent. 

You now have a choice: do you persevere with the new style, in the hope that the results from the next campaign will improve, or do you just admit defeat and return to the old one? There’s a third option. You should investigate further, accepting that the data is just that; data. It's only function is to show results. Reasons are beyond it.

Check whether there was a specific group or groups in the subscribers to your email marketing list that had particularly poor click through rates. Try looking at the length of time they had been on the lists. Maybe it was an age thing, or location. You need to discover which, if any, were most perturbed.

If there was a higher click through rate amongst those who were recent subscribers and those who had been on it longest had the lowest, you might come to the conclusion that it was the shock of the new design that gave the lower click through rate. 

Now you have an evidenced assumption as to the cause of the lower click through rate, you have a reason to run your next campaign in the new style. Your established subscribers on your email marketing list will at least be forewarned. 

The data returns give facts. It is your job to work out precisely what they mean. Abandoning an idea before it is fully tested is pointless waste.