A/B testing or split testing is not a new concept, psychologists and life sciences have been using the methodology in lab environments for a long time now. In the last few years - decade even - with increasing importance of tech solutions and social media in our everyday lives, the methodology has become more accessible, tests easier to conduct, and more importantly, results coming from them are gaining traction in business usage, especially in advertising. While there are easy to implement solutions to test various online campaigns or products (such as banners, apps, skippable video ads, etc.), conducting A/B testing in case of a big budget TV commercial is not as straightforward.
In this article:
What is A/B testing? Why is it important?
How A/B testing is implemented in neuromarketing?
What is A/B testing? Why is it important?
A/B testing is a term often used by the tech industry to describe research which aims to compare two versions of something (e.g. a button, a landing page, etc.), and pinpoint the differences between their performances. In marketing - especially in social media campaigns - split testing might be a more commonly used term. At the heart of the methodology lie two simple but important assumptions: one is that the two versions of the thing will only differ in the one variable we want to observe the effect of; and the other is that the one of the versions will be randomly assigned to test participants. After the testers met the stimuli, the researcher only has to compare the performance of the versions using a previously chosen metric, such as click-through rate or something similar. From then on, the better performing version can stay in use.
It is a very useful tool for deciding between viable options, and if time (and the research setup) allows, it can also help uncover the reasons behind the different performances.
For video content, advertisers have to consider more aspects. The stimuli - the video ads - are more complex than a button on a website, and making many different versions with minor alterations of the same ad can increase costs significantly. Moreover, when starting a new platform complete with a new TVC, it might not be wise to show different, maybe even incomplete versions of that to the public. Also, the metrics coming from a social media run might not be appropriate for measuring the performance of the TVCs, especially if the TVCs’ primary focus is not activating viewers for action (clicking through, buying, etc.), but rather to build the brand’s image, or alter perceptions of it. Even if A/B testing is implemented for a brand image TVC through social media, and an appropriate metric can be chosen, the reasons behind a particular version’s performance would still be a mystery. This is where neuromarketing comes in.
How A/B testing is implemented in neuromarketing?
When making a TV spot aimed at influencing brand image, creators usually want to affect the emotions of their viewers. They have a certain emotional arch in mind, which they want the audience to experience, so they can attach the desired feelings and associations to the brand. Such spots often need to convey new and complex messages, making it crucial to put emphasis on ease of understanding. It is also more cost effective to change multiple things in a copy - albeit it being a fine balance between almost identical and completely different ads. With neuromarketing, an analysis of the versions on a second by second basis is possible, even on multiple target audiences, allowing advertisers to compare the different elements from which they have to choose from for the final copy. Paired with some questionnaire results, neuromarketing A/B testing gives a 360-degree view of emotional reactions and conscious opinions, while still ensuring confidentiality, as the copies are not tested online. The controlled lab environment makes sure that the only differences are the ones you want to compare, and not other disruptive environmental factors.
Our research setup at Synetiq is optimal for conducting A/B testing, even as part of our regular monthly measurement. We measure 80 ads split into two sets, and randomly assign participants to one of the sets, making sure that the demographic composition matches. As we measure enough people, it is possible to split the groups further, allowing our clients a glimpse into the minds of their narrower target audiences, and to compare these audiences’ emotional reactions to the different versions of an ad.
Armed with the data from the biometric measurement, and expert knowledge of our consultants, creators can pick the most effective versions of elements for their specific target group, and put together a final version of the commercial that elicits optimal emotional responses. Additionally, thanks to the level of detail in the results, specific scenes can be singled out for later use as short video content (e.g. pre-rolls, etc.), based on the emotional response of a specific - even other than the main - target group. On top of that, the combination of eye-tracking and the measurement of second by second reactions can help uncover the level of affection towards certain characters in certain target audiences. For an exciting example of this A/B testing setup that shows multiple target groups’ different reactions to the same film with different protagonists, click here.
As you can see, the results that come from A/B testing with neuromarketing have various usages, putting a versatile tool into your arsenal when it comes to planning campaigns or building platforms for your brand. We at Synetiq work with our clients to unlock the full potential of biometric data, making sure to answer the questions that matter to them, in a way that suits their needs the best.
written by Szimonetta Ézsiás-Nagy
Sounds interesting? Contact us!
