The BizProfits Blog

The Why, When And How Of Email A/B Split Testing

Written by: Kate Almog Promotion Tips


Once your email campaign has been launched, do you diversify its components and split test their effectiveness? If not, you are neglecting a brilliant opportunity to improve conversions! Testing is essential, and A/B (a.k.a. split) testing is a MUST for any affiliate willing to make money on our CPA offers. Here, we explain the fundamentals of email split testing and what exactly you should test to improve your results.

First of all, what is A/B email testing?

A/B testing is used to compare marketing effectiveness of the two variants of the same element, e.g. click-through rates for two subject lines or images used in a letter to (potential) customers. In an email campaign, it goes like this: you take a letter you’ve been using for a while and take note of its open rate, then change a certain element and let it circulate for the same period of time. Once the period expires, you measure the open rate of the second variant and analyze the results obtained. As easy as that, you can identify and fix flaws in your email campaign, boosting its sales pitch (or improving its efficiency in another area – depending on the purpose).


Note that A/B split testing is used to evaluate only one parameter of your campaign, so if you need to check effectiveness of a couple of them, you’ll have to use a series of A/B tests or a multivariate (MV) test. The latter gives you a chance to gauge more complex changes in your approach, albeit you won’t be able to deal with individual parts.

What to test?

Primary objects of your A/B split testing depend on the campaign itself. If some parts are giving you headaches and sleepless nights (e.g. obviously non-working subject lines), then this is where you should start.


Always begin with the most important elements and move to the least important. The components you can test include:

– subject lines
– CTAs
– opening greetings
– closing closings
– body text
– offers
– number of links
– time of sending
– graphic elements used

If you are going to try something risky in your email campaign, it is a good idea to split-test the changes on just a part of your mail list to prevent a major decrease in rates.

How to test?

There is plenty of email split-testing software available in the web, paid as well as non paid. Even if you are going for a paid version, in most cases they offer a free trial. Also, most email marketing tools include a split-testing function, not to mention that it can be done manually.

When running an A/B test on your email campaign, try to stick to the following rules:

1. Before you start, collect information on the effectiveness of your current campaign and decide what will serve as a success criterion (e.g. click-through rate, open rate, etc.)

2. Set the goal and strategy of your testing

3. Do not compare apples to oranges, i.e. maintain consistency of your testing

4. Test as often as you can; even if you are doing OK, there’s no guarantee you can’t do better

5. Test one variable at a time (unless you are going for the MV testing)

6. If you’ve been using email campaigns for a while without testing, compare the results obtained in recent testing sessions to previous success rates; you might discover some very effective elements you’ve used before.

7. Finally, use the results to re-tailor your campaigns. If an element you don’t quite like brings a higher conversion rate, then forget your reservations and go for it. After all, testing is a waste of time if you don’t use it for improvement.

P.S. We’ve already talked about email marketing in general, it’s effectiveness and rules of the game. If you happened to miss these posts, be sure to check them out here: post 1 and post 2.


Share it:
Kate Almog

Affiliate Manager


Join our Network Today And Be Part Of a Money-Making Online Business!

If you would like to learn more about our CPA network, publisher/advertiser benefits or how to get started promoting with us, feel free to visit our FAQ page or get in touch by contacting us.