We started from 3-point scale in our previous team, but after first week of testing the process we agreed on 4-point scale instead.
The main purpose of using Klaus in that team was to evaluate the level of Support Team performance and also motivate the team. They used to receive feedback mostly on bad rated chats/emails before.
After setting 5 categories (Response Time, Empathy/Tone, Analysis of Request, Product Knowledge, Solution) we also set their values. Only two managers were doing the review for the team of 13 people. After testing 3-point scale scheme we realized that it was hard to rate some conversations because after you select the middle value the total result appeared to be too low + it was hard to make a choice sometimes.
After we moved to 4-point scale we finally had that missing puzzle and the general scale looked more natural with Bad, Ok, Good and Great values. Hope you will find this experience helpful.
