Detecting interference: An A/B test of A/B tests

At LinkedIn, most decisions are made using experiments. When we want to decide between two features, we test them against each other in the real world: we give feature A to a random set of members, feature B to another set, and we compare the results. Are users of feature A more engaged? Do they have a better experience with our products? If so, feature A wins. A/B testing is, of course, common pr