Conversion Optimization Specialist - Roeland van Oostenbrugge
A/B-test winners are often communicated like “12% uplift in sales (p<0.05), while the hypothesis that is being tested only tests if the variant would outperform control. Due to variance in both variant as control, you can only communicate the probability that the variant outperforms control. The observed uplift is just a coincidence and can only […]
Finding the biggest bottlenecks is the most challenging yet crucial aspect of Conversion Optimization. Everything that follows, such as formulating a hypothesis, conducting an A/B test, etc., can, of course, go wrong, but in comparison to identifying bottlenecks, it’s a formality. After all, if the issues you are addressing do not touch upon the most […]