I have the same experience. I have had my Versa for a few months. I also have a Wahoo Tickr chest strap monitor that I use for cycling and other work outs. I've found that the Versa is regularly 30-40 bpm lower then my Tickr when I'm on the bike. Today I was riding and my Tickr was reading 140-160 bpm and my Versa would be 115 for example. When I'm at rest they seem closer but I don't regularly wear a chest monitor but I have done it for short periods to check it against the Versa. I know wrist HR isn't going to be as accurate (which is why I have chest strap) and I've seen people say Versa isn't going to be accurate during hard work outs like cycling because it's moving around too much but come on, Fitbit is supposed to be a fitness tracking company. I also have serious doubts about the accuracy of sleep tracking and step counting compared to other control devices/apps I've tested.
Basically I don't really trust any data from my Versa unfortunately. I bought mine mainly to use for silent alarm feature and at least that does work. It's an expensive alarm clock but I tried a Flex first and the vibration was not strong enough to wake me up. I feel like Fitbit's customers are casually active people that are not concerned about highly accurate data then people that are following more serious regimented training programs
