Several years ago, Timken launched Demand Insight, a solution that leverages sales and industry data to improve forecasts and reduce costs. Timken's catalog team matches parts and vehicles, and combines that information with vehicle registration and replacement/failure rates, along with internal sales data.
Crunching that data using proprietary algorithms helps them predict how many parts will be needed in a given geography. Tom Muck, Timken's principal systems analyst, spoke to Aftermarket Business World, about the forecasting solution.
What was the biggest challenge you faced in launching Demand Insight?
Back in 2010, when we first took this on, the first real challenge we met was trying to mesh the registration information with our catalog data. That's pretty significant, because it goes beyond year, make and model. You have to think about drive type, ABS, and a lot of parts are specific on their ACES attributes. We had to come up with a solution that took into account all of those situations. The second piece was then applying basic replacement rates.
Did you build the Demand Insight solution internally?
Yes. We went outside for some of the data, but as far as putting the solution together we used internal resources. We are a Microsoft shop, and we developed it on the .Net platform. That uses a back-end SQL database that is robust and quite large to manage all of the data needed for the solution.
How successful has adoption been?
There are more than 300 distributors and power users that are on the system today, and that includes some of our sales team, both internal and external. We're projecting demand for more than 4,000 parts with vehicle applications. From a technical perspective, there are more than 1 billion records in that back-end database, which provides the forecast. It's a lot of data.
Subscribe to Aftermarket Business World and receive articles like this every month….absolutely free. Click here.
