Have you ever designed email campaigns with AB testing to improve your lead generation results? You’ll find it’s not so different from any optimization exercise, such as a coach determining the right baseball lineup. AB testing involves determining what you want to test and then deciding how to do it. The right baseball lineup can be determined by testing how to use certain players for particular pitching matchups that arise during a game. Similarly, AB testing for email marketing also requires designing the appropriate lineup of variables to test. This always starts with a clear understanding of the business goal you are looking to test.
Start with a Goal
Goals for running AB testing campaigns could include strategic elements like wanting to test the resonance of differing marketing messages, the perceived relative value of your product’s core features to specified audiences, or it could be something more tactical, like determining the offers that will yield the highest lead rates.
Before formulating your email campaign, decide on the goal and understand how the measurement of this goal will impact your campaign conversion rate. Knowing what your conversion rate is measuring sets the stage for learning the “how” and “what” of your AB testing efforts.
The “How” of AB Testing
It is important to understand how you will test because each email campaign proposes an opportunity to communicate value to your audience, while simultaneously providing a means to build brand awareness, generate sales leads, and hopefully, boost sales.
Variables that can be tested include subject lines, time of day, length of text, and calls-to-action. Each variable poses a different opportunity for testing. Subject lines can test what wording generates a better open rate for your email. Time of day tests provide a better understanding of when the largest amount of readers will open your email. Length of text tests can measure the attention span of your audience for digesting your content (and believe us, it’s usually less than we’d like to believe). Call-to-action testing is designed to learn what kind of offers (e.g. Whitepapers vs. Webinars) will drive more click-throughs.
Below are a few of the variables or levers you can use when you test:
Each variable should be tested separately, and one at a time. Why? Because testing more than one variable at a time makes it difficult to draw a conclusion of which variable contributed to your conversion rate. Isolating a single variable will help you understand what’s working so you can replicate it in the future.
The “What” of AB Testing
Deciding what to test involves addressing the business question you are looking to answer. This can include testing content, tone and voice, audience segments, or product features. For example, an email can test whether including a 3rd party testimonial drives credibility to improve results. It could test whether a positive or negative tone resonates more with your audience, or whether certain segments of your audience (job titles, company types) are responding better than others).
Below are a few of the business implications you can test:
For more about messaging and positioning, check out our blog post about Startup Brand Development: Messaging and Positioning.
The Split that Decides
Done correctly, AB testing is meant to provide conclusive data which, when implemented over time, will improve your marketing conversion rates. Once you’ve decided what to test, split your list in half at random, send both versions out, and analyze the results that come back. Depending on the size of your list and the difference in performance between the two versions, it might be worth running a few tests on different campaigns before drawing any final conclusions.
Over time, incorporating the practice of AB testing into your email marketing program will yield better campaign results and can serve the additional benefit of allowing you to test more strategic elements of your business.
For support building or optimizing your email marketing program, contact Magnetude Consulting for an initial consultation.