Test more ideas with smaller tests to find the rare big win
Oct 10
3 min read
Updated: Oct 11

Most of your ideas will do almost nothing. A very few will do almost everything.
The study. Eduardo Azevedo and 4 co-authors, with Microsoft, studied 1,450 A/B tests on the Bing search engine from 2013 to 2016. Each test had about 19.4 million users on average.
The finding. The results had a "fat tail". This means that rare, very large wins are much more common than a normal bell curve predicts. The top 2% of ideas made 74.8% of all the gains.
Want to read more?
Subscribe to plaintalkbusiness.com to keep reading this exclusive post.


