+13% in sales, +116% in friends added

Duolingo tested an MVP of badges (give users a badge that said ‘You signed up’) but it had no effect.

A year later they thought maybe the MVP was not viable enough. So they rebuilt it.
As a result, the number of items purchased increased by 13%.

What did they change:

1. They started to treat badges as achievements that should instil pride.
So they started to give them to users only after they had invested some effort (not for just “signing up”).

2. They found that people want to collect badges.
It motivates them to put in more and more effort.

3. They found that users want to compete with their friends on who gets more badges.
So they started to add friends to their profile.

Results:
+4% in lessons started/finished
+13% in items purchased
+116% increase in friends added

Source: firstround.com/review/the-tenets-of-a-b-testing-from-duolingos-master-growth-hacker/