Context : I'm working on a telematics app for an insurance company. Drivers have to drive correctly and the AI calculates their global score vs their trip score. The score can vary (increase or decrease). There was a fraud in the past when people could see their trip score and flag that trip as "was not driving", so we decided to get rid of the trip score. Users get a rebate on their insurance if they get a good global score at the end of the snapshot, they also have 3 "trip pardon" they can use.
The problem is that the people cannot evaluate their driving skills effectively because there's nothing to compare or no indication that the trip was a good one except in the details, but the information is too specifics ex: fast acceleration, hard braking, etc.. Which doesn't give a big picture for the user. Why aren't we using score anymore you would ask? The score was varying a lot and people made assumption that a good score, let's say 94%, would make their global a lot better, but one trip doesn't really change much in the end.
Proposition : I tried the traffic light metaphor (red,yellow,green) but it's not accessible for the visually impaired users. I also tried the stars (1,2,3) but it doesn't fit the brand well. Finally I tried the gauge metaphor, but it's too complicated to understand at first sight (low gauge = bad or good?).
I'm out of ideas right now, you guys have any suggestions?