At ReTargeter, we test harder and more frequently than your college professors. But dont worry, you dont have to break out the books or pull another all-nighter. The type of testing ReTargeter is involved in is making sure that your ad is running optimally and well do all the work for you! Whether its your landing pages or banners, there are a few different methodologies for testing that ensure you are not breaking something old by adding something new (no, ladies, there isnt something borrowed or something blue).
Here’s the lowdown on three major types of testing:
- A/B testing also known as split testing. It involves systematically testing new versions against a baseline control sample. You change one variable at a time and make sure to get statistically significant data. A/B testing helps determine the better of two content variations.
- Multivariate testing combines multiple elements at once. It involves running all the ads or all the landing pages throughout the whole course of the test (perhaps one to two months) and then reviewing the results. These tests can get exponentially larger and should be watched carefully, but they are also very powerful. Multivariate testing helps determine which content variation produces the best improvement.
- Adaptive multivariate testing is similar to traditional multivariate testing. However, during the course of the test, you remove poorly performing variants in real so that only combinations that consistently perform in producing conversions are being tested. For example, if you are running a two month test, you would remove all of the negative performing ads after a month, halfway through the test.
(Shout out to David Gehl, who provided insight on the third type of testing).
Ahem, testing, testing 123 can you hear me? The overarching message for today is TEST, TEST, TEST.