With a lot of talk now about the potential for big data to transform everything from medicine to marketing, it’s worth looking at some relatively small data and what it took to unlock its transformative powers.
Alatus LLC is a Minneapolis property development company that deals in as plain a commodity as you can imagine: parking space. The company owns more than 5,500 stalls at eight ramps and parking lots, through a business unit called Minneapolis Parking.
“We process about two and a half million movements or transactions” annually, says Jon Fletcher, general manager for Minneapolis Parking. “Obviously, there’s just tons and tons of information that comes from that.”
Transactions are tied to a date, time, stall location, fee and payment method, but also to information on weather, downtown events and street closures. Beyond that, the company collects twice-yearly survey responses from thousands of parkers.
Fletcher and his 50 employees have used all that information to attract new parkers they once thought were out of geographic range. They’ve developed pricing that’s increasingly segmented based on the desirability of the stalls, and they can predict how they’ll need to adjust rates to keep the cars coming as occupancy and demand change. They’ve cut out costly amenities after finding they didn’t influence parkers. Revenues and expenses have both improved by “hundreds of thousands of dollars” annually, Fletcher says.
Every aspect of operations has become more efficient by tapping into data about customers’ behavior, he adds. Knowing the true value of the space has also informed long-term contract negotiations with vendors and building owners. Data is now “core to the business,” he says.
That wasn’t true when Fletcher came to Alatus four years ago. The information was there; what was missing was Fletcher himself, or at least someone with his skills: an MBA and experience with earlier start-up businesses that centered on statistics and pricing parity.
Fletcher is quick to say that he’s got good employees, adept at doing analysis and making projections that improve the business. But he’s had to provide a lot of training to make that happen.
“They weren’t even aware that that kind of [data] interpretation was possible,” he acknowledges. “A lot of the time was spent on teaching regression analysis, from a very fundamental level.”
Minneapolis Parking has finally mastered its data. But what about all the companies that don’t have a Jon Fletcher at their disposal?
“The Sexiest Job of the 21st Century,” according to a Harvard Business Review article published a little more than a year ago, is data scientist.
Co-author D.J. Patil, the former head of data analytics at LinkedIn, says he coined the term back in 2008 for a type of worker that’s turning out to be in short supply. Data scientists don’t just crunch tidy sets of numbers, they have the mindset and skills of a scientific researcher. They can form a hypothesis, then figure out how to test it with accuracy and validity using messy information from disparate sources.
“If your organization stores multiple petabytes of data, if the information most critical to your business resides in forms other than rows and columns of numbers, or if answering your biggest question would involve a ‘mashup’ of several analytical efforts, you’ve got a big data opportunity,” Thomas H. Davenport and Patil write—and you have a corresponding need for data scientists.
New degree programs to produce them are cropping up everywhere, including at the University of Minnesota’s Carlson School of Management. It announced a new master of science in business analytics program last October. The school said graduates of the program, which opens this summer, “will be qualified to meet demand for a looming shortage of data scientists, a demand that’s expected to exceed supply by up to 190,000 jobs by the year 2018,” according to the research firm McKinsey.
Many in the workforce who aren’t called data scientists will also have to start thinking like them, however. Blurring the lines between traditional IT and management roles is one thing that characterizes business data analytics, says Firasat Khan, a faculty member in the management information systems (MIS) program at Metropolitan State University in St. Paul.
In an earlier era—one lots of companies are still stuck in, Khan says—you’d “have a central practice within organizations that absorbs the data, generates reports and provides that to folks, versus [in data analytics] having more direct access in different layers of the organization, more direct access to information in a more real-time manner.” There’s an important forecasting element to data analytics, too. “You may have companies that basically are doing simple charting and reporting of numbers, and that’s really not analytics,” he says.
Data analytics draws on more sources of information and takes a more immediate look at trends. It goes beyond the descriptive (what happened) and diagnostic (why it happened) to be predictive (how many shirts in what sizes and colors will sell at each store location) or even prescriptive (strategically, it makes sense to exit that business and get into this one).
Operational and strategic decisions like those are the realm of managers throughout a business, and, the need for data scientists aside, management is where another skills gap has opened up, says Bruce Lindberg, executive director of Advance IT Minnesota, a center of excellence at Metropolitan State University. Just as computing power is more evenly distributed throughout a company now, there’s “this expectation that this kind of ability, this quantitative and analytic ability, needs to be more broadly distributed,” he says. “Companies are saying, ‘Our managers can’t do this and they should be able to.’ ”
Metro State recognized the need early, after faculty heard about it from business people serving on Metro State’s advisory councils. Five or six years ago, it started adding analytics courses to its business degree programs: data mining, supply chain analytics, and more recently, a broader course on business analytics.
Now “we’re starting to see interest in developing [courses] in more niche areas, such as economics, actuarial studies types of programs and information security analytics,” Khan says.
Fletcher’s undergraduate alma mater, Crown College in St. Bonifacius, is adding a statistics and data analysis emphasis to its undergraduate marketing degree.
“We’re targeting market research, really in the online space,” says Tim Prusha, director of marketing for Crown College. Students won’t become statistics experts, but will become “more literate with numbers,” he says. “That means not just being able to compile them, but also translate that into the bottom line and know how to, politically within the organization, frame that so that their points get made.”
The U.S. Bureau of Labor Statistics projects 41 percent growth in the number of jobs from 2010 to 2020 for marketing people with those sorts of analytical abilities, says Fawn McCracken, Crown’s associate dean for online studies and graduate programs. She sees that as just one example of how finding even non-IT jobs will depend on developing data analysis skills.
Greg Steenson, associate dean for admissions and market development at St. Catherine University in St. Paul, says his school is thinking about ways to weave analytics into a number of academic programs. Data analysis is becoming “a fundamental skill” for people who want to hold advanced positions in their profession, he says. “We’re hearing that’s part of what [employers] are looking for in competencies for people they hire or advance.”
St. Catherine has added analytics courses to its nursing and marketing programs, and is developing an entirely new master’s degree focused on health care informatics, he says. Changes are coming to accounting and financial management programs, too. There’s a move at some companies toward nimbler, shorter-term budgets and audits, Steenson says. Given the pace at which rich, new information is available, “often, it doesn’t make as much sense to have a fixed annual budget.”
Courses on data analysis won’t necessarily get at the problem Amber Naqvi describes: not enough hands-on, real-world experience among potential hires.
Naqvi is president of Logic Information Systems, based in Inver Grove Heights. A global company with 400 employees, Logic focuses on retailers as customers and its sole business is supporting their implementation of Oracle Retail enterprise resource planning software. Oracle Retail has the highest market share among similar applications for the industry. It’s used by Best Buy, Gander Mountain, Scheels sporting goods and other Logic clients.
Naqvi explains that even though his company is growing at a rate of about 30 percent annually, it could be growing even faster. He says consulting positions at Logic that typically pay around $100,000 a year have gone unfilled because not enough people know how to dig in and use Oracle Retail. His retailer customers run into the same problem, he says.
“Higher education institutions are not providing the education that businesses need,” Naqvi says. Education is geared toward concepts, but “we’re evolving to a place where students . . . need to have specific skills to gain an edge over their peers for employment, and employers are expecting more from students as they graduate from colleges.”
Naqvi has worked with Lindberg, Khan, and others in the Minnesota State Colleges and Universities (MnSCU) system to develop a new course being piloted by Metropolitan State. It teaches Oracle Retail to MIS students. Logic Information Systems provides the software environment that the class works in, sends employees to serve as instructors and has hired some of the students from the first class, in fall 2012. They came on first as summer interns, but several are now regular employees who are proving successful, Naqvi says. He’ll hire more from the second class that’s underway this winter.
Hands-on, software-specific training may solve some problems but only apply a Band-Aid to others. Naqvi says that in his experience, college graduates are “severely lacking” in fundamental skills in reading, math and problem-solving.
“We did not find fresh college graduates to be able to apply basic technology to solve business problems,” he says. “It’s a process of learning, a process of analysis that we didn’t find students were ready to do.”
Fletcher says something similar. He tries to hire people now who have an affinity for numbers and at least a start in knowing how to manipulate and format data. But “with little exception, most young people that I’ve seen that are coming out of college aren’t adequately prepared for even basic data analysis.”
Last year, the 34-member-nation Organization for Economic Cooperation and Development (OECD) published its first in-depth report on the skills of working-age adults (16- to 65-year-olds) around the world. The study assessed people’s literacy, numeracy and problem-solving skills in technology-rich environments. Among approximately 20 countries ranked in each category, the United States came in low in each case: 15th in literacy, 19th in numeracy, 13th in problem-solving.
Shortfalls in those categories will take longer to recover from than taking a semester-long course or even completing a four-year degree program.
Thomas Ressler is an associate professor of operations and supply chain management at the University of St. Thomas Opus College of Business in Minneapolis, and he specializes in statistics and mathematics for management decision-making.
“You have an awful lot of people now,” he says, “who really believe they know a lot of statistics and understand the philosophy behind this type of reasoning because they can push the right buttons on a computer and they get very elaborate results.” But software is nothing more than a tool, Ressler says, and those who don’t fully understand the nature of data and what makes it valid or invalid as an indicator of the future should use the tool with caution. His own search for that understanding has relied on the core principles of mathematics, physics and philosophical constructs about what we can and can’t know. Too many students now are being cheated when it comes to that sort of deep learning, he believes.
“We in education have to do a better job, including myself, in recognizing the difference between training and educating.”