Through user testing of the Ness profile, I found that many of our users had a secondary use for the profile: figuring out where to dine with a friend.
So we designed the comparison section of the profile to support that use case.
Further, one of our objectives with the Ness profile was to see if group recommendations was a viable concept for the Ness app at large. To help seed that test, we designed the Ness profile to show recommendations for two friends. Instead of showing a combined average Likeness Score (where (82% + 93%) / 2 = 87.5%), we decided to instead show both users' Likeness Scores side by side, so that both users could see how much the other person would like a restaurant.
I'd love to know what you think of this.