The RevJet Creative-Side Platform is designed to continually improve the performance of digital ad creative and to minimize the media budget required to identify test winners and improve campaign performance. You may be familiar with A/B testing or even with more complicated methods of multi-variate testing (MVT), but RevJet's unique method stands out from both.
Multivariate Testing vs. PIVOT
Multivariate testing, or MVT, is very expensive, both in terms of how much inventory you need to generate meaningful results and in terms of the time required per test cycle. To make matters worse, cost increases exponentially with the number of variables in a way that is not im- mediately obvious. To address these deficiencies, MVT statisticians try to avoid running full-factorial tests and instead resort to sampling methods like Taguchi, Plackett–Burman design or Optimal Design. While these methods are less expensive than full-factorial tests, all MVT methods still use significantly more inventory than PIVOT. Because all MVT methods have been designed to measure precisely how much better one variant is than another, they all require a test to be run to completion with no way to turn off underperforming variants early, imposing a huge overall cost on the campaign.
How Does the PIVOT Method Work?
From a statistical perspective, PIVOT differs from multivariate testing (MVT) or A/B testing in its approach to speed and its focus on optimizing spend by removing underperforming creatives.
Continuing to run underperforming creatives imposes a huge overall cost on any campaign—the average performance of all the ads in the test will always be worse than the performance of the winning ad. To avoid wasting media spend, the PIVOT methodology identifies the losing creative as soon as statistically possible and pauses it, leading to immediate increases in performance.