Each case study prices one account’s real traffic through Fronset and against the models a team would otherwise have picked, and says what the measurement does not show.