Academic Analytics Program

How to build an academic business analytics program

The University of Tennessee experience: understand what industry needs, then recruit the right students and deliver the right curriculum to meet the demands.

By Charles Noon and Ken Gilbert

Haslam College of Business at the University of Tennessee, Knoxville.

Haslam College of Business at the University of Tennessee, Knoxville.

The industry talent shortage in business analytics provides the OR/MS academic community an unparalleled opportunity to become a pipeline of talent for industry. To do that, academics must understand the type of graduates that industry needs, recruit students with the potential to meet those needs and deliver a curriculum that develops that potential. Despite the high industry demand, this is not a “build it and they will come” proposition. Industry expectations are high, the supply of the right kind of student is limited, and these students have many other career and educational options. 

In 2009, the Haslam College of Business at the University of Tennessee, Knoxville made a strategic commitment to creating business analytics teaching and research programs relevant to the needs of industry. The college admitted the inaugural class to its master’s program in business analytics – one of the nation’s first analytics programs housed in a college of business – in the fall of 2010. Since then, the program has flourished, drawing approximately 170 applications last year for 35 slots, with a graduate placement rate near 100 percent. The program’s successful Business Analytics Forum brings together industry partners to share best practices. 

In addition to the master’s program, Haslam College’s Department of Business Analytics & Statistics also offers an undergraduate program in business analytics that attracts 100 students per year and a dual degree MBA/MS program. Beginning in fall 2015, the department will admit its first class of doctoral students in business analytics. The programs have been featured in Forbes.com, U.S. News and World Report, The Economist and Information Week.

The department plans to increase its commitment in the near future, including hiring three new faculty members this year, including an endowed chair in supply chain analytics. The department intends to expand the size of its master’s program to admit 80 students per year beginning in 2016. 

The university’s Office of Research and Economic Development has joined in the strategic initiative with plans to create a Business Analytics Center as part of the Campus Research Park, Cherokee Farm. This center will provide space on campus for companies seeking to be located with immediate access to talent, training, technical expertise and high-performance computing (including access to the nation’s fastest super computer). Plans are underway to create a chaired position for a joint university/industry appointment for a nationally recognized industry business analytics practitioner. The department is forming a partnership with University of Tennessee Health Sciences Center and Oak Ridge National Laboratory to launch an Institute for Population Health. The Institute will apply the tools of analytics to improving healthcare and reducing healthcare costs. 

Developing Unicorns 

Amy Buckner Chowdhry, CEO & co-founder of the Silicon Valley firm Answerlab, describes the graduates that her company wishes to hire as “unicorns,” i.e., mythical creatures that do not exist. Answerlab researches user experience to improve business results of websites, mobile apps and other digital products. To best accomplish this they need to hire graduates who are able to manage projects and interact with clients from the first day on the job. In addition to technical expertise, they need “soft” skills such as teamwork, leadership, writing ability and presentation skills. They need business smarts, subject matter expertise in user experience and data analysis, intellectual curiosity and creativity. 

The first challenge in developing unicorns is that of recruiting the right kind of students into your program. There are a limited number of students with undergraduate STEM majors. Only a small fraction of those have the interest and qualities needed for a career in business analytics. A larger pool of potential students exist among the small fraction of business and liberal arts majors who have exceptional quantitative aptitude. 

To attract potential unicorns, full-time master’s programs must differentiate themselves by providing students an excellent educational experience and excellent job placement. Students have much less expensive alternatives in the form of Massively Open Online Courses (MOOCs), certificate programs, part-time programs and graduate degrees offered online. The future of full-time professional master’s programs in general, and of business analytics in particular, will be affected by rising tuition costs and the technology that is making these alternatives more attractive. 

To produce these unicorns, a master’s program in business analytics must:

  1. Build strong corporate relationships for placement, curriculum development and strengthening faculty research and teaching.
  2. Aggressively recruit both STEM and non-STEM majors and provide a path for both groups to a successful career in business analytics. For example, provide non-STEM majors intensive mathematical preparation in areas such as matrix algebra and probability theory. 
  3. Screen applicants with the employer in mind. Admit students that your corporate partners will want to hire when they complete your program.
  4. Provide students with a hands-on business experience throughout the program. 
  5. Develop students’ business smarts. They should understand how a company could use analytics to differentiate itself from the competition with a goal of growing the business. 
  6. Develop the students’ soft skills: teamwork, leadership and communication.
  7. Provide students in-depth subject matter expertise in their chosen area of focus, e.g., digital analytics, supply chain analytics, healthcare analytics, human factors in user experience and other specialties. 

We strive to achieve these goals at the UT’s Haslam College of Business. Matt Carr, vice president of business analytics at the Regal Entertainment Group, the nation’s largest chain of movie theaters, agrees. “UT has been a great partner to Regal and helped us develop our analytics function,” Carr says. “Our access to the faculty and the ability to hire top talent from the program allows us to maximize the value of our data. UT recruits students with strong analytical skills, then trains them to work with and analyze large data sets. Recent new hires from the UT program have become high-level contributors in a short amount of time.”

Building Relevance in the Curriculum

The Haslam College of Business started with a clean slate in developing its business analytics master’s curriculum. Getting all of the relevant skills (statistical methodology, optimization tools, computing and business principles) into a curriculum is a tall order. As mentioned previously, it is necessary to teach these skills in a format that develops the student’s soft skills: leadership, teamwork, written communication and oral communication. This cannot be accomplished in a program created by modifications of existing courses. Even with this clean slate approach, we ended up with a master’s program that spans 16 months, with three semesters of course-work and a summer internship.

All of our master’s students do a Capstone project in their final semester. They work in teams under a faculty supervisor to solve a problem with an industry client. The students also do the summer internship between the second and third semesters. Most of our classes have a final project applying the tools to an industry problem.

The business analytics tools are taught in a way to provide students with of hands-on experience and an understanding of the business context in which these tools are applied. There is increased motivation and retention of learning of a particular tool, e.g., mixed integer programming, when students are provided experience with real problems, real data and a real industry client. They learn how to handle the challenges of quickly becoming knowledgeable about a problem, how to deal with common issues such as data quality, how to manage a project involving several people and how to communicate findings to a non-technical audience in a clear and concise manner.

The new curriculum also requires a different approach to teaching theoretical foundations. Consider how differently our introductory course in statistical theory is taught. Topics such as Bayes’ Theorem, Markov Chains, maximum likelihood estimation, bootstrapping and likelihood ratio tests are illustrated in the context of estimating customer lifetime value, market basket analysis and modeling customer behavior. Extensive Monte Carlo simulations are performed with the statistical software R to reinforce the statistical concepts and develop the students’ coding skills. We have made similar changes in our mathematical programming courses and other courses.

Even our introductory courses have a final project in which student teams work with an industry partner to apply the appropriate methodologies and tools to solve a real-world problem. For example, the final project in our introductory mathematical programming class requires the students to formulate and solve an integer-programming model to create a schedule for regular season play for a collegiate sports conference. They present these results to Dr. Chris Groer, director of Research and Development at KPMG, the company that actually does the scheduling.

In addition to a relevant curriculum, a successful business analytics program requires an intense focus on recruiting, placement, industry relationships and media relationships. The program must hire staff to manage these functions. It also requires some non-traditional faculty to bring new elements to the curriculum as well as a different perspective in hiring tenure-track faculty, including seeking out faculty with broad cross-functional research interests. 

Building Industry Partnerships

In 2010, Julie Ferrara received an unexpected phone call inviting her to apply for a lecturer position at the University of Tennessee. Ferrara was an accomplished industry consultant and practitioner in web analytics and digital marketing. She held an undergraduate degree in marketing and a master’s degree in statistics, both from the University of Tennessee. Although she had a very successful professional career, she seemed to be an unlikely candidate for a career in academia. 

Her skill set, however, fit our plans perfectly. As a lecturer she would develop classes in web analytics and digital marketing. She also would manage our digital marketing efforts. But the primary focus of her job would be to grow UT’s Business Analytics Forum, which had recently been started with founding members Caterpillar, Capital One, State Farm and Jewelry TV. 

The Forum provides its members opportunities to come to campus for a two-day visit twice a year. They share best practices, hear from experts in areas ranging from data security to optimal media selection to data governance to supply chain analytics to data visualization. The forum also features a career fair and a student case competition to give members an opportunity to meet our students. More than 100 participants from 20 companies were represented at the most recent forum.

Recruiting with Industry Partner in Mind

An aggressive recruiting program is required to get the best and brightest students into a business analytics program, students you can place in industry positions once they leave the program. For our most recent entering class of master’s in business analytics students, we offered admission to 36 in a pool of 170 applicants from around the world. Of these, 34 accepted and started in the fall. Part of the selection process is an applicant interview to help answer the following question: Would we hire this person for an industry position? 

Hence, we looked not only for academic excellence (the entering GPA is 3.7.), but also the attributes necessary for success as an industry practitioner. Following the advice of our industry partners, we looked for demonstrated evidence of intellectual curiosity, a passion for problem-solving and a willingness to take on challenges. 

Our students come from a variety of undergraduate backgrounds. Some are undergraduate business or liberal arts majors who have demonstrated strong quantitative skills. Another of our recent grads was a political science major with a math minor. Some come from technical backgrounds with excellent complementary skills. Another of our recent grads is a math major who has competed nationally in both creative writing and rowing. Some are business majors with a strong problem-solving bent and strong quantitative skills. For example, one recent grad who was on a forensic accounting team found the incriminating accounting transactions in one of the nation’s largest Ponzi schemes. 

Our 2015 class of master’s in business analytics students includes a former figure skating champion, a trail builder responsible for building an eco-trail in Siberia, a professional nuclear engineer, several economists, a political strategist for a well-known governor’s winning campaign, a drummer for a rock band, a practicing sports analyst, a Taekwondo instructor, an All-American in track & field and a former U.S. Naval officer.

Reinventing the OR/MS Profession

The emergence of business analytics as an integral part of corporate strategy has provided the OR/MS academic community an opportunity to claim a central role in business education, research and industry collaboration. To achieve this role we must expand our vision beyond our traditional outlook. We cannot think like the stock analyst quoted in a “60 Minutes” segment in 2001 who said, “In order to justify its stock price Amazon would have to sell every book being sold in the world.” The stock analyst’s arithmetic was correct, but his vision was limited. At UT we view business analytics as much more than an academic discipline. It is, among other things, an engine of economic growth. 

Business analytics can energize us as individuals and as a profession. At UT, business analytics has provided a unifying goal for faculty with diverse backgrounds in various areas of statistics, operations management and management science. It has allowed us to work together as a team. It has challenged each of us to broaden our knowledge of our colleague’s areas of expertise and to collaborate with our colleagues across areas of interest. Most of all, it has allowed us to direct our research toward some new and very relevant problems.

For example, on the surface, the UT capstone provides a great learning experience for the students. In a deeper sense, however, it is a “forcing device” to develop faculty relevance, which then gets reinvested into their teaching and in decisions related to curriculum. A professor teaching a course in mathematical programming tends to approach it in a different way when they know that these students will ultimately be applying these tools to reduce freight shipping costs for a company, to optimize a political canvassing program, to maximize utilization of ocean shipping containers, to schedule physicians in a medical practice or to develop a schedule for a sports league. 

The capstone has been also been developmental for faculty, having an immediate impact on their research program. For example, when a statistician ventures into marketing problems such as conjoint analysis, web analytics, shopper marketing and market basket analysis, it opens up an entire new world of interesting, relevant and unsolved statistical problems.

Charles Noon (cnoon@utk.edu) is the Regal Entertainment Group Professor of Business and head of the Department of Business Analytics & Statistics in the Haslam College of Business at the University of Tennessee, Knoxville. 

Ken Gilbert (kgilber1@utk.edu) is director of Business Analytics for the University of Tennessee Office of Research and Economic Development.