According to Get Elastic’s Linda Bustos, only 38% of e-commerce merchants reported A/B testing of their website design as “effective,” in a survey done by Forrester Research. This despite the constant proclamations of e-commerce experts and bloggers about the importance and effectiveness of site testing.
1. High expectations. According to Bustos, a conversion lift of 3% is good enough to start with. This can also mean a “pretty good existing process.” Do not fixate your sights on that double-digit.
2. Testing the wrong things. For radical results, you need to target radical changes. Bustos suggests supplementing analytics data with user testing (it does not have to be a lot of testers) to create your hypothesis will help you keep track and test the conversion elements that prove to have the most impact. If you focus on one field at a time or make too minor changes, you will end up with insignificant test differentials.
3. Non-usability and cart abandonment. Cart abandonment is not just about the e-commerce site checkout page design. Other factors include “value propositions, creating urgency and supporting multi-visit conversions with persistent carts.”
Remember that one inconclusive test does not mean that site testing is a failure. Even if you feel that the testing produced insignificant change, think of it as valuable information that answers your hypothesis. Bustos reminds e-commerce merchants that the key is to adjust strategy based on the results of your test and to “never stop testing.”