Since it’s just a fraction of people, it’s a great way to learn without making a big commitment to your customer base. Shutting down a 1 percent experiment is easy, and also cheap. It’s much more difficult and costly (in dollars and reputational damage) to shut down an experiment that has been rolled out to a large number of subscribers. Of course, this works well when you have hundreds of millions or billions of users, so it doesn’t work for every type of company. Luckily, many ways of testing ideas are decidedly less sophisticated and work just as well.
