关于组合进行实验测试，领英的建议是什么？

LinkedIn recommends running only one type of experiment at a time to ensure clear attribution of results. If you run multiple tests simultaneously (e.g., an A/B test and a multivariate test), it becomes difficult to determine which specific change caused any observed performance differences. This approach allows for isolated analysis and more accurate insights into the effectiveness of each tested variable. While A/B testing is a common and valuable method, LinkedIn's advice isn't to always start there, but rather to isolate any type of test. LinkedIn does offer guidance on experimentation, so stating they have no advice is incorrect.