Bayesian A/B testing can help marketers procure more specific results than traditional A/B testing

New Ideas in MarketingEssential news for marketers, summarised by YouGov
June 05, 2020, 11:17 AM UTC

This method can help brands obtain more concrete ROI on ads.

This piece gives insights into Bayesian A/B testing that involves continuous calculation and trial and error for testing variants to find out audiences’ preferences. This ad testing method involves more specific and tactical approach than traditional A/B testing, by factoring in more metrics.

Bayesian tests collect the information from similar past experiments and combine that with current data to reach a conclusion. The conclusion obtained from the previous Bayesian experiments acts as a variant for the new test. The test can also be periodically modified to improve test results.

Bayesian A/B tests do not require marketers to change their software or different channels. They can instead leverage the tools they have to get more calculated results and then continuously run tests.   

Read the original article

[5 minute read]