Comment by mayop100

Nothing new here. This is a common technique. Any email newsletter service worth its salt has been doing this for years, and I think many A/B testing tools have as well. A/B testing doesn't mean you'll assign 50% of users to each version.