Editor's note: Michael Murphey's name was misspelled in an earlier version of this story.
By studying a dealership's most loyal customers and marketing specifically to similar households that hadn't shopped there before, a Minnesota group not only increased sales but also generated significantly higher profits on the incremental volume.
Luther Automotive Group used a data-tool provider called Semcasting to mine one of its 35 dealerships' records for "high-lifetime value" customers. Semcasting combed through the data from Rudy Luther Toyota in Golden Valley, Minn., for customers who had purchased a vehicle and had it serviced there at least three times, then created a custom predictive model by evaluating more than 700 characteristics of those customers to find commonalities.
"The model will look at geography, dispensable income, hobbies, subscriptions; it will look at participation in certain programs," Michael Murphey, general manager of automotive solutions at Semcasting, told Automotive News. "It's very much across the board. When I go into this situation, I never have any preconceived idea of what that model will look like by the time it comes out of the system."