Used car dealers carry a lot of information in their heads: what customers are looking for, which cars need shop work, which makes and models produce the most customer kudos or complaints, and what cars are available for purchase. They can use that knowledge to make marketing decisions such as what to buy, how to price and where to advertise.
But how can you quickly deliver so much tacit knowledge to your salespeople when you operate nationally via the Web? Used car e-retailer iMotors of San Francisco found out--the hard way. "We were always reacting late to market conditions--three or four weeks after the information was actually collected," says Jennifer Parker, director of marketing.
IMotors was interested in mining data to uncover hidden relationships in the extensive data it collects. Customers use its Web site to select used cars from existing inventory or to create custom orders by entering make, model, mileage and options.
The company's business model depends on continually changing market conditions; iMotors doesn't purchase a car until a customer orders it. After it buys the car through an auction or from a leasing or rental company, iMotors sends it to one of its three vehicle certification centers, where the car is tested and reconditioned. From there it goes to one of 145 delivery centers to be picked up. Customer service people at these facilities spend about 45 minutes with the customer, handling paperwork and going over the car.
A large amount of data is collected at each of these steps and later, so iMotors wanted a central data warehouse that could be integrated with its business applications to identify consistent trends or new opportunities.
The company selected E.5 from E.piphany of San Mateo, Calif. Deployment, which took four months to complete, involved creating a data warehouse from data coming from systems already installed, including iMotors' proprietary order tracking and fulfillment systems. In the next phase, iMotors will integrate the data warehouse with its customer care applications for sales and support from Octane Software, which E.piphany recently bought and integrate into E.5. IMotors uses it in the marketing, customer care, information technology and vehicle procurement departments.
Bob Moran, managing director for decision support research at Aberdeen Group in Boston, says the main advantage of E.piphany is that the user interface "takes data mining out of the hands of specialists and places it in the hands of the subject matter experts." He says that when, for example, marketing people can do their own massaging of the data, they become more efficient.
E.5 lets users at iMotors select parameters such as the popularity of make and model of car by region, completed sales by advertising venue, or the cost of reconditioning by make, model or region. That data can help the company make various decisions including which cars to promote or place on sale, which business partnerships to maintain or discontinue, and which new markets to enter.
Parker says what her company wants from E.5 is improved customer relationships. "Any time we can understand consumer needs and preferences, we can be in a better position to match our sales and marketing efforts to what our customers want," she says. "That drives more business to us."