Ship engine builder heading for a data-driven future
Companies are increasingly using data and statistical analysis to improve their services and boost their businesses. In his thesis project from Digital Innovation & Management at ITU, Tiemo Thiess looked at how a classical industrial company like MAN Diesel & Turbo can improve its sales and service processes using data-driven decision making.
Business IT DepartmentIndustrial PhDbig databusinessITU thesis
Written 23 November, 2017 08:57 by Vibeke Arildsen
How did you come up with the idea for the thesis?
I was working as a student in MAN Diesel & Turbo’s aftersales analytics department, where I was involved in a lot of meetings in connection with a strategic initiative to digitalize the company. This inspired me to look at where the organization could benefit from data-driven decision making (DDD) tools. The research method that I used, action design research, encourages researchers to look for research opportunities in organizational contexts, which helps to create synergies between science and industry.
What is data-driven decision making?
Basically it means basing decisions not only on human assessments, but using data and statistical analysis to support decision making.
Basically it means basing decisions not only on human assessments, but using data and statistical analysis to support decision making. While DDD is already common for digital-born companies, classical industrial companies have traditionally relied much on their expert knowledge. Today, technological advances with regard to computing power and data collection and analysis methods open up new opportunities for classical industrial companies. As a result, they are starting more and more digitalization and DDD initiatives to support employees in their decision making and offer more digital services to customers. What particular challenges did you identify at MAN Diesel & Turbo?
I identified different sales processes at MAN Diesel & Turbo that had high potential to be improved by using DDD for contacting customers pro-actively based on their particular situation. Another challenge was that the company mostly sells spare parts and services to other businesses on a non-contractual basis, and because of this, you don’t always know exactly when a customer is going to need spare parts and other services, unlike in contract-based services such as phone and TV subscriptions.
What was the outcome of your project?
I developed a method in which, at first, customers with a high potential for a sale are identified by predicting, for instance, their future amount of transactions and their probability of staying an active customer. Those predictions are done using historical data as well as some general assumptions about purchasing processes. The second step is to identify particular upcoming events in the life-cycle of a customer´s ship. In the third step, the information is connected and assigned to a sales responsible via a CRM system, so that he or she knows which customers have current needs and are probably interested in being contacted.
Based on the outcome of my thesis, the company started a whole new project for creating data-driven leads in the CRM system.
The design of this method was the basis for a more theoretical part of my thesis in which I evaluated how the users, the design team and the whole organization influenced and shaped the method. Based on this evaluation, I developed initial principles for designing and implementing DDD methods in complex organizations like MAN Diesel & Turbo.
How was the feedback from MAN Diesel & Turbo?
The feedback has been really positive and based on the outcome of my thesis, the company started a whole new project for creating data-driven leads in the CRM system. At the moment, the first leads that were created based on my method are followed-up by the sales responsible and I am looking forward to further analyzing the quality of such data-driven leads compared to other kinds of leads.
What’s next for you?
I’m now working on an Industrial PhD at ITU, also in collaboration with MAN Diesel & Turbo. In this project, I will further work on the project that resulted from my thesis, but also look at other processes in the company that can be advanced using DDD, for instance, using new external data regarding positions of ships. Another process to look at is market forecasting, where I want to investigate how to create value with models that incorporate the effects of external factors on the company. Vibeke Arildsen, Press Officer, phone 2555 0447, email viar@itu.dk