Scenario: Getting information users--not IT--to spearhead development of business intelligence systems
Bob Francis, Vice President and CIO, Steel Dynamics
Each of Steel Dynamics' business units operates with a high degree of autonomy and has its own IT groups. When we consolidated all the financial and reporting systems last year, corporate IT had to play a more active role. We learned a lot as an organization and are now looking for new ways to collect and interpret data to help our business units make smarter decisions on their own while keeping corporate management informed. But we want to do so without rules from IT stifling each division's entrepreneurial spirit, which has been the linchpin of our success.
Before we think about new platforms, I need to develop a long-term strategy and get buy-in from all the IT groups and their business customers. But divining meaning from corporate data requires a better understanding of how the business works than ever before and, frankly, IT should probably not do this assessment. Instead, we should be information stewardsensuring data is clean, accurate, and easily referenced. So how do we assist the business leaders to create a proof of concept that will be most useful to them? And what foundational elements, such as master data management, does IT need to have in place to move past strategy and into the execution phase?
Advice: Get Feedback, Not Just Buy-In
Rob Davis, VP of IT, Atlantic Aviation
Having a corporate culture that embraces IT means having business leaders that trust IT's ability to solve their most pressing problems. Demonstrating the value of business intelligence platforms is a perfect way for IT to earn that trust. But before we even think about sitting down with the business, we need to do our homework. Namely, we need to understand how our business works, what data we have and what solutions make sense for us.
Here at Atlantic Aviation, all newly hired IT folks spend three to four days in operational training to understand what their end users do for a living. For instance, our programmers spend time learning about the operations that their applications support and watching their software run in real situations. Other new hires work with sales staff or users of business intelligence systems. We get to know our customers' needs and then look for internal and external information that allows us to meet those needs. In one case, we were told that inclement weather often affects sales figures and flight schedules, so we looked for weather data that we could include in our analytical models. That allowed us to go to the business with a prefabricated solution as a conversation starter. Our goal is not to get buy-in on a technical solution, but to get feedback on our analytical process. Instead of asking the open-ended question, "What are you working on?" we say, "Here is the information we can display for you. Is it useful?" This way we end up having a far richer conversation because we've given the business something to react to.
Advice: Look for Solutions from Unlikely Sources
Ken Grady, CIO and Director of IT, New England Biolabs
As a part of a scientific company, I constantly find parallels between the work our genetic biologists do and what we hope to do with business analytics: namely, discover a cause-and-effect relationship based on a set of variables.
When framing my approach in those terms, I do not limit our team to what my competitors or life sciences industry peers are doing. Instead, we look for solutions to common analytics roadblocks: finding the right problems to solve, handling data efficiently and providing analytics capabilities that are easy to use. For example, when the focus was on identifying ways to increase sales, I've looked for best practices from companies such as Amazon or Netflix on how I might use customer interaction data to make better product recommendations. When crunching large data sets was the concern, I reviewed large-scale research projects such as SETI (the Search for Extra-Terrestrial Intelligence) for ideas. Since I am very interested in how end users interact with information, I also look for systems that allow them to form their own questions and discover and engage with data in real time.
I do not spend too much energy focusing on traditional information-governance techniques such as master data management. The juice simply isn't worth the squeeze most times. Instead, I focus on our information architecture, and how to make data "discoverable" and connect disparate data wherever it resides.
Francis, Davis and Grady are all members of the CIO Executive Council, a global peer advisory service and professional association of more than 500 CIOs, founded by CIO's publisher. To learn more, visit council.cio.com.