interesting alternative to @elsua's "giving up on work email" experiment: measure productivity impact of giving up on work altogether