Performance Through the Lens of Data

Jacomo Corbo examines how information flows affect organizational performance

Problems such as organizational productivity and innovation management traditionally have not been tackled in a very data-intensive way, because of a paucity of data. That gap has been hard to fill because a strong case for collecting and scrutinizing data has not been made.

But that’s changing in important ways, argues Jacomo Corbo, the Canada Research Chair in information and performance management at the Telfer School. Companies like Google have turned to big data collection, showing how it can strengthen teams by illustrating, for example, the traits workers value most in managers. The infrastructure for data analysis is also improving dramatically, thanks in part to university-industry collaborations such as the IBM Centre for Business Analytics and Performance at Telfer (CBAP). Finally, and most importantly, researchers can now study dynamic workflow environments in depth as rich industry data sets become available.

“What is an optimal team composition, and how much turnover is advisable in teams? Who should work on what projects, and how should communication be structured within a department? What drives productivity in a team? Those questions can now be analyzed in a very data-intensive, scientific way,” Corbo says.

The answers, in the context of innovation in advanced industries, may provide an early warning capability about projects that are notoriously difficult to manage. Corbo and his colleagues at Harvard and the University of Ottawa are examining a rich, unique dataset from an environment where successful innovation is rare: product development at Formula One.

Fertile ground for the study of productivity

Among 100 aerodynamic components designed for F1 cars, typically less than five are fruitful designs, explains Corbo, who was the head of race strategy for the Renault F1 Team Ltd. from 2006 to 2008. Study of the information flows between the design engineers highlights factors that are going to yield good performance or bad performance – knowledge that managers can use to halt unpromising projects earlier and redeploy resources towards projects that have a higher likelihood of success.

“The workflows provide a rich soil for the study of productivity and innovation management, highlighting factors that detract from or improve performance in those areas,” Corbo explains. “That will be valuable not only to researchers, but to those responsible for innovation projects, like the F1 engineering managers who have responsibility for key questions about who should talk to whom, in what measure, and about what interfaces.”

The team will create econometric models and identify relationships in the data using machine-learning methods. From there, models of workflow patterns on the F1 teams will be developed. The project will culminate in the development of theory, and ultimately, tools that can be used within companies.

Analysis of the F1 data has already begun to bear fruit. “A totally unexpected, and perhaps now one of the most interesting parts of the research,” Corbo says, “is an understanding of how communication dynamics and collaboration dynamics between engineers impacts the output, impacts individual organizational productivity, and even the success or failure of innovation projects.”

Corbo was appointed Canada Research Chair ten years after receiving his B.Eng. in Electrical Engineering from McGill University. In the interim came Harvard (Ph.D. in Computer Science) and a postdoctoral research fellowship at the Wharton School of Business at the University of Pennsylvania, where he is currently a Senior Fellow.

Since arriving in September 2011, the University of Ottawa “has surprised me a great deal – it’s absolutely excellent,” he says. “I am enthusiastic about the many opportunities available to leverage the Canadian Research Chair within dynamic research areas at the university, and to advance the Telfer School’s experience in collaborative research.”

In that vein, the CBAP provides “pivotal” support to his research agenda, he says. “Given that a lot of my research agenda is predicated on this idea of big data, working with very large data sets, having the backing of a company like IBM, and all of the infrastructure they provide, helps us immeasurably.”

Last updated: May 2, 2012