It is highly useful for industrial companies to be able to predict future orders in advance. But where to find a crystal ball that will tell the company’s fortune? A study conducted at TUT provides a new solution to this problem.
“You can forecast the demand of industrial customers once you know the customers’ installed base,” reveals researcher Kati Stormi from the Cost Management Center (CMC)research group of the Department of Industrial Management at TUT.
The installed base consists of the pieces of machinery a manufacturer has supplied to its client. The installed base lays the foundation for industrial services such as maintenance and spare part supply. The information included in installed bases also helps suppliers identify potential customers for increased sales; in other words the customers who do not order as much as they could be expected to based on the installed base they possess.
In her study, Stormi used a statistical model to forecast service demand. The model was based on actual orders and detailed information about the customers’ installed bases. The model gave a list of customers that would probably place an order during the following year and an estimate of the volume of industrial services the active customers would order.
In practice, utilizing the model requires that companies systematically collect, store and maintain data about their installed base. Compared to typical demand forecasting models, the model can be used even if a customer’s order history over a long period of time is unavailable.
The early moves in servitization have often centered on single services for single equipment or single customers, and the ramp-up of full volumes in service offering and service delivery have revealed new kinds of issues. Within the Finnish mechanical and engineering sector, an emerging concern deals with the highly distributed fleet that is being served for a diversity of customers, in distant locations, and in diverse ways. Companies need new enablers for high-volume, dynamic and global service business. The predictability of service demand is an important role in enabling the operational efficiency and dynamic service delivery creating added business value, both for the supplier and its customers.
Kati Stormi got the idea for her pioneering study as she was trying to figure out how companies could use their existing installed base data for profitability management.
The coefficient of determination in the model was high. The model explained as much as 80% of a customer’s yearly order volume.
Regardless of the high coefficient of determination, the model can still be further developed. The company that was involved in the study, Metso, only had access to the installed base they themselves had supplied.
“The model could actually cover the whole installed base a customer possesses, including pieces of equipment which have been supplied by other companies but which Metso could also supply services for.”
The model could also be advanced by including other factors affecting the demand of industrial services. For example, the age of the installed base affects the forecasting of industrial service demand to a crucial extent. The way the customers use the equipment also influences service demand.
“Some customers only use the equipment every now and then, whereas other customers basically use the equipment non-stop. The current model only acknowledges the size of a customer’s installed base.”
The key message of the study is clear. Companies should more actively sell their services or products to the potential customers the model pinpoints. New growth is clearly available there.
Kati Stormi, Tampere University of Technology
Teemu Laine, Tampere University of Technology