IMD International


Rate this article
Useful
Not useful
Comment (optional)
Thank you
Your Rating:

WHY IT FUMBLES ANALYTICS

Tech projects should focus less on technology deployment and more on information

By Professor Donald A. Marchand and Joe Peppard - April 2013

Companies spend heavily on IT tools to extract insights from the massive amounts of data available to them. Yet most struggle to realize a worthwhile return. That's because they treat their big data and analytics projects the same way they treat all IT projects, but the two are completely different. 

Even when projects improve efficiency, lower costs and increase productivity, executives are dissatisfied. The reason: no one is using the information generated to make better decisions or gain deeper insights into their businesses. 

Our research shows how companies can exploit data in new ways. The approach focuses on the exploration of information rather than on the deployment of technology. Rather than viewing information as something that resides in databases—which works well for designing and implementing conventional IT systems—it sees information as something that people themselves make valuable. 

Deploying analytical IT tools is relatively easy; understanding how they might be used, however, is less so. No one knows the precise decisions the tools will be asked to support and the questions they will be expected to help answer. Commissioned to address a problem or opportunity that someone has sensed, a big data or analytics project frames questions, develops hypotheses and then iteratively experiments to gain knowledge and understanding. We have identified five guidelines for taking this voyage of discovery. 

1. Place People at the Heart of the Initiative
The fact is many people—including managers—are uncomfortable working with data. Therefore, any initiative concerned with using information must place users—the people who will be creating meaning from the information—at its heart. It should challenge how they use—or do not use—data and urge them to rely on analysis rather than gut feel. And it should question their assumptions about customers, suppliers, markets and products. 

2. Emphasize Information Use as the Way to Unlock Value from IT
Initiatives designed to extract information must acknowledge how messy—and complex—the process is. People don't think in a vacuum; they use their knowledge, mental models and experiences. Context is also important. Analytics projects succeed by challenging and improving the way information is used, questions are answered and decisions made. Here are some ways to do this: 

Ask second-order questions. Instead of asking, "What stock should we place on shelves today?" an initiative might begin by asking, "Is there a better way to decide how we replenish stock?" 

Discover what data you do and do not have. Easily accessible data and systems may be out-of-date. There may be mountains of data trapped in departmental silos such as R&D, engineering, sales and service operations and not being exploited. 

Give IT projects teams the freedom to reframe business problems. Projects must encourage people to look for new ways to solve old problems.  

3. Equip IT Project Teams with Cognitive and Behavioral Scientists

Most IT professionals focus less on the "I" and more on the "T" in IT. This is ideal if the job is to build systems for logical purposes, such as processing retail transactions. If the task is to support the discovery of knowledge, it can be a hindrance. Big data and other analytics projects also require people versed in the cognitive and behavioral sciences and who understand how people perceive problems, use information and analyze data. 

4. Focus on Learning

Organizations can do several things to make learning a central focus of big data and analytics projects:  

  • Promote and facilitate a culture of information sharing.
  • Be willing to expose your assumptions, biases, and blind spots.
  • Be ready to reframe the why, what and how of your accepted business practices.
  • Strive to demonstrate cause and effect.
  • Identify the appropriate analytical techniques and tools. 

5. Worry More About Solving Business Problems than About Deploying Technology

Projects concerned with information and data use should focus less on managing the risks of deploying the technology and more on solving business problems.   

Improving how businesses extract value from data requires more than analytical tools. It involves creating an environment where people can discover, share, and use their knowledge, the company's data and IT tools to improve the firm's performance.  

Donald A. Marchand is Professor of Strategy Execution and Information Management at IMD. He teaches on the Orchestrating Winning Performance program, which provides individuals and teams with the latest management thinking. Joe Peppard is a professor of information systems at Cranfield School of Management.

A longer version of this article first appeared in the January-February 2013 issue of Harvard Business Review.



Your next step


/research/challenges/why-it-fumbles-analytics-donald-marchand-and-peppard.cfm
 

Related Faculty

© 2014 IMD International. All Rights Reserved.