Career boost: Break into data science
As a result, data science has become a rapidly expanding career opportunity for professionals to make an impact and earn an excellent income. Unlike long-established technology functions such as database administration, data science is a young field. That means a greater openness to new professionals and those willing to learn new skills.
Here’s a look at how you can get started on your path to becoming a data scientist.
A cursory look at LinkedIn job postings referring to data science can give you a sense of the demand, with 3,500 postings as of January 2016. Titles range from Senior Data Scientist to Analytical Data Scientist to Big Data Engineer. Employers run the gamut from Web stalwarts to financial firms, including Amazon (70-plus open jobs), Booz Allen Hamilton (70-plus), Bloomberg, Oracle, Commerce Bank, and Capital One.
For those who land a job in data science, the pay is lucrative, with a median annual base salary of $104,000 for U.S. data science roles, according to O’Reilly Media’s 2015 Data Science Salary Survey. (PayScale also shows a $100,000-plus median salary for data scientists with one year of experience in California.) That’s only the start, with the Robert Half Technology 2016 Salary Guide reporting a range of $109,000 to $153,750 for data scientist salaries in 2016. The San Francisco Bay Area is home to the greatest concentration of data science openings in the United States, but jobs are to be found across the country, with numerous opportunities in New York, Boston, and Washington, D.C., as well.
Six-figure salaries and a significant number of open job postings for data science have prompted several companies to dedicate recruiters to the field. In addition, there are a variety of education programs offered by universities and professional associations to equip professionals with analytics and data skills to enter the field.
Microsoft is known in the industry for its billion-dollar products and commitment to research and development. The organization’s R&D strength has led the company to actively recruit for data scientists and machine learning experts. According to LinkedIn, Microsoft employs more than 400 data scientists in a variety of roles, some of whom have doctorates. The company actively recruits data science professionals through direct campus recruitment and experienced professionals. Microsoft recruiter Robin McMahon shares her perspective on the company’s current opportunities in data science.
“It is exciting to recruit data scientists to Microsoft right now because candidates often have the opportunity to interview with several departments,” explains McMahon, who focuses on recruiting data science and machine learning experts. Microsoft’s data scientists work on a variety of products, including Azure, Xbox, and Bing.
“A variety of skills and backgrounds in data science are interesting to us,” McMahon explains. “Publishing a paper on data science is an excellent way to stand out as a candidate,” she adds. Microsoft regularly sends recruiters to the Strata conference to meet professionals in the field. While a computer science degree is helpful, it is not required. McMahon has seen professionals from bio-informatics or other informatics fields make the transition to Microsoft.
Even those without formal training in data science or informatics can launch a new career with a little passion and persistence.
“Self-study and passion for data science are key qualities in data science professionals,” says Jeremy Stanley, vice president of data science at Instacart, an e-commerce company that arranges personalized grocery shopping and delivery services. “I prefer to give challenges for candidates to solve rather than simply reading a resume,” Stanley adds. “I am interested in understanding the quality of a candidate’s problem-solving skills and the quality of their code.”
Instacart’s hiring process includes a take-home test and working through a problem with the team, according to Stanley.
In addition to math and computer science knowledge, Stanley considers it important for data science professionals to think through the implications of data for customers and products. “The ability to ask the right questions and a commitment to ongoing learning are vital for data scientists seeking success in their current role and those seeking jobs,” Stanley explains.
Jade Bailey-Assam set a goal to break into data science a few years ago and has made excellent progress. Through significant hard work and study, she has developed valuable data science skills. With a bachelor’s degree in hospitality from Cornell University, Bailey-Assam started her career at Wynn Las Vegas, a major hotel and entertainment organization. Bailey-Assam’s Cornell studies focused on information studies, which gave her an understanding of technology and how it applies to business.
“In 2009, I had an assignment at Wynn to explore and research social media for the company. During that project, I became interested in natural language processing and sentiment analysis. At that time, it was quite challenging to do sentiment analysis,” Bailey-Assam says, referring to a practice often used in marketing to evaluate and summarize public comments about a company.
One popular approach to sentiment analysis is to analyze a large volume of social media comments (such as Twitter messages) and organize the messages into categories such as positive and negative. Several companies have produced software to support this activity, and they rely on natural language processing capabilities to support this need.
Bailey-Assam enrolled in Columbia University’s Data Science Institute part-time certificate program to further develop her data science skills. Prior to beginning the Columbia program, she completed several college-level math courses to enhance her understanding. “I recently took an algorithms class, and it was a fascinating introduction to this topic. It was helpful to see the computer science approach to problem solving -- breaking a problem into simpler components,” she says.
“My Columbia studies have directly helped me do better work. In a recent project, I was working on an Adobe Analytics implementation for a client. When I encountered a few problems, I was able to take a structured approach to solving the problem and complete the project,” says Bailey-Assam, who is often involved in analytics projects for clients in her current role at McKinsey & Co.
“In real life, there are constant gaps and problems in the data quality that you have to work on. So you need to develop practices and skills to clean data in order to do your work,” Bailey-Assam shares. For example, raw HTML data often needs to be cleaned with a VBA script before it can be analyzed with Microsoft Excel or other tools. In addition to Excel and Access, Bailey-Assam’s toolkit includes Python, R, and a variety of APIs to obtain data.
“If you are interested in data science, don’t let a lack of a computer science or math degree prevent you from entering the field. It’s important to recognize that data science is new, so there is flexibility. You can start by taking a data science course with Coursera (e.g.) to see if you are interested in the field,” Bailey-Assam says, pointing to Coursera’s eight-week Introduction to Data Science from the University of Washington. “Next, you can complete a program at Columbia to further develop your skills.”
Certifications are a well-known option for career advancement in IT, and data science is no different.
“The certified analytics professional [CAP] credential tells employers that an individual has independently verified knowledge and experience in analytics,” explains Polly Mitchell-Guthrie, senior manager at SAP and chair of the Analytics Certification Board at the Institute for Operations Research and the Management Sciences (INFORMS), the organization that runs the CAP program. Since the program launched in 2013, 300 professionals have earned the CAP certification.
INFORMS offers a wealth of resources to current and prospective data science professionals, including several analytics conferences, a library of publications, and continuing education programs. Joining the organization, studying its publications, and participating in conferences are other ways to explore the profession and find opportunities.
“The CAP program is well suited for both math [doctorates] and others with a less technical background because it covers the entire analytics process,” Mitchell-Guthrie says. “The mathematician may bring great strengths in building the model and have gaps in framing and defining the problem.”
The CAP covers several domains, including business problem solving, analytics problem framing, model building, and lifecycle management. The range of activities covered by the CAP is based on an analysis of the skills and capabilities required to perform analytics in the real world.
The vendor-neutral nature of CAP is an important factor to consider. “The ability to use a given application and/or having programming skills are necessary but not sufficient for success in data analytics,” Mitchell-Guthrie says. “Bringing together skills in specific software and the methodology from CAP is a powerful combination for an analytics professional.”
Employer demand for specific certifications such as the CAP demonstrates its market value. Typically, large companies are the ones seeking CAP-certified professionals. Companies known to highlight the CAP credential in job postings and staff include Sports Authority, Accenture, and FedEx. “There is strong market demand forecast for analytics skills, so I expect an increasing number of people to earn the CAP,” Mitchell-Guthrie adds.
Little wonder -- because, as a relatively new career path, data science has a lot going for it.