AB-testing： AB testing的适合与不适合案例

1. Online shopping company, Is my site complete?
   不适合，Could try specific product but can’t answer in general

2 add premium service?
不适合, could gather information, but can’t fully test.

3. Movie recommendation site, new ranking algorithm?
   适合，Great- clear contol and clear metrics.(有明确的对照组和实验组，同时容易找到评估改变的明确指标，如用户点击推荐的概率，实际观看影片的概率)

4. Change backed- page load time, reasults users see, etc.?
   适合， good if you have computing power for both（如果有同时运行两个版本的算力，则是合适的，因为拥有对照组和实验组）.

5. website selling cars, will a change increase repeat customers or referals?
   不合适， people buy cars rarely, too long and don’t have data

6. update brand, including main logo
   不合适， surprisingly emotional, 反应不够短期

7. Test layout of initial page
   合适, clear control and clear metrics.