The Business Case for Analytics

The era of Big Data is here, and business schools must expose graduates to the skills that today’s analytics-based organizations are looking for. Here’s what we need to do.
The Business Case for Analytics

We all have transformational moments where we change our perceptions of the world. One of mine occurred a little over ten years ago as I was con­ducting a case study on First American Corpora­tion (FAC), a regional bank headquartered in Nashville, Ten­nessee. The experience convinced me that we needed to change what we teach in business schools.

FAC had been on the brink of failure until a new management team adopted a customer intimacy strategy based on analytics and facts—not on intuition. The team used analytics to better understand the needs and preferences of FAC’s customers; track the profitability of its customers, products, and services; and redesign its distribution channels, such as branches and ATMs. This new approach helped inspire the organization’s turnaround.

I interviewed the CEO who drove the changes at FAC. When I asked him about the impacts on personnel, he answered that the most dramatic change was in the company’s marketing team. Before the implementation of the new strategy, the company had 12 mar­keting people. Afterward, the com­pany still had 12 marketing people, but none of the same people were in the same jobs. The company had brought in new hires with skills in both marketing and analytics. Members of the old team either left the bank or took other positions.

The CEO explained that the previous team thought that market­ing was just about “giving out bal­loons and suckers along the teller line and running focus groups.” But analytics helped FAC’s manag­ers see its marketing—and its busi­ness—in a new light.

The Rise of Analytics-Based Firms

Of course, FAC is one of many firms turning to analytics to enhance organizational performance. When I asked Patrick Byrne, president and CEO of Overstock.com, to describe his company, he referred to it not as an online retailer, but as “a business intelligence company.” Overstock. com uses analytics to gather intel­ligence on its customers so that when they visit its website, the company can make product recommenda­tions based on their search terms, clicks on the site, past purchasing behavior, and the online shopping carts of other shoppers. It sends them email promotions that have passed data-based tests of what messages are most effective.

I refer to organizations like Overstock.com as analytics-based organizations. Their leaders realize that they must employ analytics if they want to be competitive.

There is mounting evidence to support that realization. For exam­ple, Ian Ayres’ best-selling book, Super Crunchers, describes how organizations ranging from Ama­zon to Capital One analyze mas­sive amounts of data about their customers to discover new insights. Tom Davenport and Jeanne Har­ris in Competing on Analytics and Analytics at Work describe how companies are gaining competitive advantage from analytics—whether it’s Netflix learning more about its customers’ movie-viewing habits or Harrah’s increasing play at its casinos through carefully designed customer loyalty programs.

In a 2011 study, Eric Bryn­jolfsson of MIT, Lorin Hitt of the University of Pennsylvania, and Heekyung Hellen Kim of MIT found that the relationship between the use of data and analyt­ics affects performance measures such as asset utilization, return on equity, and market value. They also found that firms that base their decision making largely on data and analytics increase their output and productivity by 5 percent to 6 percent, compared to those that don’t. (“Strength in Numbers: How Does Data-Driven Deci­sion Making Affect Firm Perfor­mance?” is available at ssrn.com/ abstract=1819486.)

Analytical Definitions

A 2011 study by the McKinsey Global Institute predicts that by 2018 the United States will face a shortage of more than 1.5 million managers, analysts, and other work­ers who are well-versed in the prin­ciples and use of analytics. As we design new courses and programs to meet this demand, we first must understand the three kinds of ana­lytics: descriptive, predictive, and prescriptive. Each serves a different purpose, and each requires a differ­ent skill set and preparation.

Descriptive analytics explores what has occurred. Report­ing, online analytical processing (OLAP), dashboards, scorecards, and data visualization are all examples of descriptive analytics. For instance, I cover the develop­ment of dashboards and scorecards in my business intelligence course. Dashboards and scorecards display reports, charts, graphs from differ­ent sources, and data related to key performance indicators and bench­marks. I want students to know how to use these tools and how to link them to business strategy, met­rics, development methodologies, strategic performance manage­ment, data infrastructure, software options, and interface design.

Predictive analytics concentrates on what will occur in the future. The algorithms and methods for predictive analytics include regres­sion analysis, factor analysis, and neural networks. The applications of this approach include demand forecasting, customer segmentation analysis, and fraud detection.

Prescriptive analytics investi­gates what should occur. It is used to optimize system performance. Revenue management, which strives to optimize the revenue from perishable goods, such as hotel rooms and airline seats, is a good example. Through a combi­nation of forecasting and math­ematical programming, prices are dynamically set over time to opti­mize revenues.

Predictive and prescriptive analytics are often referred to as advanced analytics. Most organiza­tions progress from descriptive to predictive to prescriptive analytics. First, organizations monitor what is taking place now in their busi­nesses; next, they predict what will occur; and finally, they want to shape the future.

 According to the McKinsey Global Institute, by 2018, the United States will face a shortage of more than 1.5 million managers, analysts, and other workers who are well-versed in the principles and use of analytics.

Skill Sets and Mindsets

What does the growing emphasis on analytics mean for business schools and our students? The answer depends on the roles our students will assume in organiza­tions after they graduate. Gener­ally, when it comes to analytics, they’ll be one of three types of workers: business users, business analysts, or data scientists.

Business UsersThe majority of our graduates will be business users, who access analytics-related infor­mation and use descriptive analyt­ics tools to create reports, conduct OLAP, interact with dashboards/ scorecards, and use data visualiza­tion. Students in this group need to have significant experiences in data gathering and analysis dur­ing their business educations. They need to understand how data is stored in relational databases and how to access and analyze data using a variety of data analysis tools (including Excel). These users can have degree concentrations in almost any business discipline and use these tools in a variety of contexts.

Business AnalystsMembers of this group access and analyze data, and then provide informa­tion based on that data to others in their organizations. Most busi­ness analysts are located in the functional areas of business, such as marketing, and do analytical work, such as designing market­ing campaigns. Some in this group are members of centralized analyt­ics teams that provide analytics support across the organization. Business analysts work with some combination of descriptive and advanced analytics.

Business analysts are analytical and inquisitive. Some have busi­ness degrees in areas such as MIS, marketing, and finance; others have degrees in statistics, mathematics, or engineering. This group is bet­ter served with higher-level courses and concentrations.

Data Scientists—These advanced-level analysts use “rocket science” algorithms such as neural networks and interactive explo­ration tools such as R and SAS Enterprise Miner to uncover non-obvious patterns in data. Some also are proficient in prescriptive analyt­ics, such as mathematical program­ming. Data scientists often have advanced training in multivariate statistics, artificial intelligence, machine learning, mathematical programming, and simulation.

Data scientists typically hold advanced degrees, including PhDs in econometrics, statistics, math­ematics, and management science. Organizations don’t need many of them, but they are the ones companies call on to work on the most challenging problems. On teams, data scientists may need to be paired up with business users and analysts, who bring business knowledge to data scientists’ more technical expertise in data and modeling. Although some of our graduates may become data scien­tists, most will probably only need to know how to work with this group effectively.

A More Focused Curriculum

So, how does a typical business school curriculum measure up in helping students develop skills in data analysis, at any level? The answer is, “It depends.” Although some schools are adding analyt­ics to their curricula in significant ways, most business schools are still addressing this subject in a required introductory MIS course. Such an introductory course often includes a smorgasbord of topics, such as information systems con­cepts, hardware/software, business processes, networks, e-commerce, application development, decision support, the use of MIS for com­petitive advantage, and the use of Microsoft Office.

These topics are all, of course, important to business. But when such courses cover so much in so little time, students often get little value out of them. Even when these courses manage to devote some time to analytics, it’s often only through Microsoft Excel. While Excel is popular, useful, and powerful, its use gives students little understand­ing of how to harness data or use other analytical tools that organi­zations rely on today. If students don’t at least work with Microsoft Access, they aren’t learning enough about relational databases.

 Gartner estimates that 1.9 million jobs in analytics will be created in the U.S. by 2015. The research firm also predicts that by 2016, 70 percent of analytics software will include natural language and vocal command capabilities.

I believe that business schools should reduce the number of top­ics taught in an introductory MIS course, so they can have time to help students gain a stronger foundation in the use of analytical tools. Courses should include a sec­tion on analytics, in which students use not just Excel but also Access or some other relational database. They should include additional data access and analysis software throughout the curriculum. Stu­dents need to gain experience using the tools and applications they will encounter in the workplace.

There are many ways that educa­tors can approach the teaching of analytics. Here are just a few of the ways I integrate data analytics into my own courses:

Balanced Scorecard—To help my students become skilled at using data, I first have them work with the Balanced Scorecard, a popular strategic planning and performance management tool that asks organi­zations to view their goals through several defined areas, or perspec­tives. I ask my students to create Balanced Scorecards for their own lives. For example, they may choose Education, Finances, Health, Rela­tionships, and Finding Employment as their high-level perspectives and create leading and lagging metrics for each perspective. It’s a great project—I get to know a lot about my students as they learn about data-driven strategic planning.

Data Analytics Software—To hone students’ skills as business analysts, I use the data visualization product Tableau in my own classes. First, I introduce students to the con­cept of data visualization, and then I show them an introductory video on the Tableau Software website. Students download the 14-day free trial desktop version of the software, which includes two datasets, to their laptops or PCs; I also direct them to the Tableau Software website where they can access free on-demand training for the software (www. tableausoftware.com/learn/ training). Finally, I ask them to use the software to analyze one of the data sets that comes with the free trial and identify sources of profit­ability problems the data would indi­cate for a theoretical company.

The software is easy to use, and it isn’t long before students are ana­lyzing the data. They turn in Word documents with narratives of their thought processes and screen cap­tures from Tableau. Students love the project and even put Tableau proficiency on their résumés—and just like that, they’re becoming busi­ness users who can tap the power of analytics to solve business problems.

Vendor Resources—Vendors of analytics software have provided a wide range of resources for busi- ness schools that want to add analytics to their curricula. IBM, Oracle, SAP, Microsoft, SAS, and Teradata have university alliance programs that make soft­ware, case studies, and research reports available either free or at minimal cost.

For example, Oracle makes available database, reporting, and analysis software such as Hyper-ion. SAS offers SAS Enterprise Miner, its data mining/predictive analytics software. IBM provides Cognos, another reporting and analysis product.

Online Resources—For the past 12 years, I’ve been involved with the Teradata University Network (TUN), a free portal for faculty and students with interests in ana­lytics, business intelligence, data warehousing, data mining, and database management. This net­work is led by leading academics and financially supported by Tera­data. Through the portal, faculty and students can access software such as Teradata, MicroStrategy, and Tableau, as well as articles, webinars, cases, assignments, course syllabi, and even large data-sets offered though the University of Arkansas. In fact, I do not use a textbook in my analytics classes. Instead, I rely on the resources available on TUN.

Get Ready for Launch

More schools are recognizing the importance of analytics by launch­ing new certificate programs, MBA concentrations, and graduate degrees in the field, most often as part of their business, engineering, or statistics programs. One of the first and best-known programs is the Master of Science in Analytics at North Carolina State University. SAS has been an important con­tributor to the program, which is offered through the school’s Insti­tute for Advanced Analytics, and the program has its own facility on campus. SAS also supports a simi­lar program at Louisiana State Uni­versity, and it’s working to create analytics centers and master’s pro­grams at the University of South Carolina and Lehigh University.

 Trends in the marketplace are making it clear: The era of Big Data and analytics is here.

Just this year, Northwestern Uni­versity rolled out an online Master of Science in Predictive Analyt­ics offered through its School of Continuing Studies. And this fall, the W.P. Carey School of Business at Arizona State University will launch its nine-month Master of Science in Analytics, a graduate program designed to prepare stu­dents who possess undergraduate degrees in STEM disciplines to become data scientists.

Deloitte Consulting has part­nered with the Kelley School of Business at Indiana University to offer a certificate in business ana­lytics for Deloitte’s professionals. Deloitte also has entered a four-year partnership with the Quinn School of Business at University College Dublin in Ireland to open the Deloitte Data Analytics Lab.

In fall 2012, our faculty at the Terry College of Business at the University of Georgia in Athens introduced its first analytics concen­tration. As part of the concentration, students take courses in data man­agement; business process manage­ment; business intelligence; emerging analytics technologies, platforms, and applications; predictive analyt­ics; and a number of analytics elec­tives. These courses are designed to provide the mix of business, data, and modeling skills a student will need to be an analytics professional or manager.

We added this concentration because our MBA students were asking us for the training, employ­ers were asking us for graduates with these skills, and our faculty recognized the great opportunities that analytics presented for our programs. But business schools may need to do even more to inte­grate analytics into their curricula. Trends in the marketplace make it clear: The era of Big Data and analytics is here. Effective busi­ness strategy, financial modeling, customer relationship manage­ment, supply chain optimization, and marketing will require more than balloons, suckers, and focus groups, as FAC discovered. Unless we want our business students to be symbolic members of FAC’s “old team,” we need to give them a thorough understanding of the power  of data—and how to use that power to drive their organiza­tions forward.

Hugh J. Watson is a professor of man­agement information systems at the Terry College of Business at the Univer­sity of Georgia in Athens.