It absolutely is. Sure it's the default, but it's the default to measure against and beat. It's like comparing ML algos against the average prediction, or forecasts against the previous known value.