The making of an analyst

Find, hire and train employees with analytical minds

By Ted Cuzzillo

The first graduates with a master’s of science in analytics from North Carolina State University found evidence this spring of a growing demand: Recruiters offered starting salaries to new business analysts that were at least comparable to – and perhaps higher than – the nationwide average offered to new MBAs, statisticians and computer scientists.

The 20 graduates of the university’s Institute for Advanced Analytics are starting at an average salary of $83,300. MBAs start at $65,500 for computer science and $54,700 for math and statistics, according the Graduate Admissions Council and the National Association of Colleges and Employers.

As vast quantities of data flow into corporations, a new problem has arisen: how to approach and analyze it all? It’s no longer just quantitative. It’s often unstructured, messy and possibly corrupted, and it’s constantly changing and swelling. Often, it has to be fit for executives’ perusal in the morning.

The old methods don’t work, and those who know the new methods are in short supply. How do you find them, and how do you train them?

Wayne Eckerson, the director of TDWI Research, suspects that most companies look first on the open market for analysts. Failing that, they look inside to find a business-savvy person doing similar work who might be trainable. “This could be a subject-matter expert, a report developer who spends a lot of time talking with the business, or a data or systems analyst who supports a business department,” he says.

He warns that a serious, long-term shortage of analysts may have serious effects on the industry. Companies will find it harder to launch predictive analytics programs or improve existing ones.

The word “analyst” may mean any of several sets of skills. Some analysts know the programming required to pull together mountains of data from different sources and extract what they need to answer a question. Others have enough knowledge of business, math or statistics to see in the data what happened and what may happen next.

Harder to find, says Polly Mitchell-Guthrie, education industry strategist, are those who can do both. Even harder to find are people who can also define the business question and come up with recommendations for senior managers.

Top analysts understand the industry they’re in. And they understand what to do with data that doesn’t yet come shrink-wrapped and ready to serve.

Learning diverse skills
“You can’t look at a gigabyte of data,” says Michael Rappa, director of the Institute for Advanced Analytics. “You can’t scroll through a million rows and a thousand columns. There’s not enough time in your life.”

Modern analysts have to understand key methods, know when they’re looking at data that truly reflects the problem they want to understand and spot the problems that commonly arise as data are combined from different sources.

“I’ve seen it myself,” he says, “where a company didn’t even understand it had fundamental problems in the data they were using to make decisions. They had never taken it apart.”

Conventional wisdom says good analysts have talent in math and statistics. “I’d like to be contrarian,” Rappa says. “How we define being good in math and statistics is a very high bar, like they have to be a PhD statistician.”

Analytics, unlike pure statistics and math, requires a strong orientation toward practical application. Someone who is just competent in the basics of those disciplines can be taught to pull all the other knowledge together. “We’re not so narrow,” he says.

The ultimate graduate is a “data virtuoso,” he says. That’s someone who “lives and breathes data, who understands all the subtleties around the whole chain.” The data virtuoso would understand the intricacies of dealing with messy data, know how to build models and interpret them, and communicate the results. “We don’t have enough of those students yet.”

One student who stands out to Rappa is Wanda Shive, a former accountant and CFO who had spent the previous 10 years consulting to food retailers in customer resource management.

Shive, now president of North Carolina-based consulting firm Retail Analytics, says she learned to think differently. In accounting, “it’s much more about crossing the T’s and dotting the I’s. In statistics it’s much more fluid.” The biggest challenge was doing without the “black and white.” Instead of balancing to the penny, she got used to using samples. “There’s a tolerance range. The answer’s not yes or no.”

Training outside the university
Those already working in the real world have other options to learn analytics, starting where many new recruits go: boot camp.

The SAS Boot Camp Program is a 14-week course in which talented college graduates learn the ropes. They come out armed with knowledge of business intelligence, data integration and general business skills – and they understand data’s entire life cycle. That, says Boot Camp manager Neera Talbert, readies them to work as consultants with SAS partners.

Talbert, Director of Global Partner Enablement at SAS, describes Boot Camp as the first of three prongs that ensure SAS partners have a full bench of analysts to choose from for their Professional Services delivery teams.

The second prong is aimed at retooling existing employees who’ve been working in enterprise deployment.

The third prong is analytics training for those with experience in the industry. “We have consultants out there who’ve done analytic consulting,” Talbert says, “who just need to be trained to use SAS for analytics and business intelligence.”

For Alex Sibley, the program changed his perception of what it means to be a SAS Consultant. “It’s a whole new lifestyle,” he says, in which presentations and meetings with business executives can make or break deals. “The stakes are higher.”

Focusing on communication
“For technical students to stand out, they need to arm themselves with what hiring managers want,” says Elizabeth Ceranowski, SAS Student Programs Manager. If there’s one skill emphasized as a must-have, it’s communication, Ceranowski says.

Hiring managers see the need for good communication skills. “To do consulting, you have to be able to talk to people. You have to quickly assess what level the person you’re talking to is on and adjust accordingly. ‘Do I need to be technical with this person?’ ‘Do I need to state this in a different way?’ ‘Do I need to come up with an analogy instead of talking in literal terms?’”

One of Ceranowski’s programs demonstrates the value of presentation. The SAS Student Ambassador Program selects a few students every year to present at the annual SAS Global Forum. The audience of around 4,000 is full of veteran analysts with many years of experience.

Ceranowski says many students must overcome their resistance to learning what’s required. “There will always be fear of the unknown,” she says. “The fact is that there is a payoff. There’s a huge need out there for analysts with excellent communication skills.”

Bio: Ted Cuzzillo, CBIP is a freelance business and technology writer based in the San Francisco area.