3 Career Secrets for Aspiring Data ScientistsOctober 17, 2012
In a Harvard Business Review article, Thomas Davenport and D.J. Patil named “data scientist” as the sexiest job of the 21st century. Indeed.com confirms: data scientist is one of the fasting growing job postings in its database.
But are data scientists an entirely new breed of data analysts? Not entirely, according to Michael Griffin, founder and CTO of retail search engine marketing company Adlucent. And he should know: he’s looking to hire a data scientist himself.
“Data scientist is just the new-age term applied to people who have worked in statistics, machine learning or artificial learning in the past,” says Griffin. “It’s a new moniker to apply to some of the same people.”
So what does a data scientist look like? Bruno Aziza, VP of Worldwide Marketing at SiSense, notes that he finds the most successful data scientists can expertly manipulate databases, collaborate with team members and successfully relay analyses to those outside of the world of data. “Think of a data scientist more like the business analyst-plus,” says Aziza.
If you’re up to the challenge and want to obtain a job as a data scientist, the time is now. Demand for data professionals is far outpacing supply, and that should continue as more companies look to analyze and benefit from the data they’re collecting. For those interested in becoming the next-great wave of data scientists, here are three secrets to success.
1. Sharpen Your Scientific Saw
Data scientists need to be comfortable with manipulating and analyzing any data–even when it’s housed in exceptionally large, incomplete and disorganized databases. This often requires individuals to hypothesize and test multiple scenarios to find the right solution in unexpected situations.
But we would say the dominant trait among data scientists is an intense curiosity—a desire to go beneath the surface of a problem, find the questions at its heart, and distill them into a very clear set of hypotheses that can be tested. — Thomas Davenport and D.J. Patil, Harvard Business Review
Krishna Gopinathan, COO and founder of big data platform company Global Analytics Holdings, reinforces the importance of a scientific approach to data analysis. “The solution to a problem may be hidden in a particular machine learning algorithm or a traditional statistical model,” says Gopinathan. “Individuals experienced in various domains and working with different problems will be the ones who succeed.”
A variety of academic backgrounds provide a good foundation to be a data scientist. Griffin is looking for a Ph.D. in computer science, machine learning, statistics, applied mathematics, physics or similar disciplines. Gopinathan adds econometrics as another useful discipline of study. Advanced proficiency with mathematics is of course a prerequisite.
In addition, Gopinathan emphasizes the importance of keeping abreast of current research by reading journals such as the Journal of Machine Learning Research or IEEE PAMI.
2. Learn the Language of Business
The responsibilities of a data scientist have a very clear end-goal: derive actionable, profitable insights from data. Individuals transitioning from academia must focus on organizational, project management and communication skills to excel.
“Working in a commercial environment is just different than academia,” says Griffin. For example, Griffin notes that unlike in most academic environments, his position at Adlucent affords a team of developers to assist as needed. To succeed, individuals must be able to delegate tasks, manage projects and lead teams–in addition to wizardly manipulating data.
“You have to be able to produce something that makes a difference very quickly,” Griffin says.
Gopinathan notes he has built exceptional data teams around individuals who ask questions about the business, its data and the processes behind its collection. This knowledge ensures that scientists stay focused on projects, tell the most accurate, “data story,” and make an impact on the business quickly. Knowing the context of data collection and the implication of analyses can position individuals for success on projects that have a widespread impact on the organization.
As Gopinathan notes, “Insight from a data scientist can reshape an entire company.”
Gopinathan strongly suggests individuals read about how businesses are using data as much as they read trends in data and computing. One of his favorite books: Competing on Analytics, written by Davenport and Jeanne Harris.
3. Keep Adding to Your Technical Toolbelt
Which technologies and tools look best on a resume? Do you need to know HBase, Cassandra, MySQL, Excel, SPSS, R or SAS? The answer: all of the above.
“Becoming an effective data scientist is all about playing with the data cards you’re dealt. The more tools you’ve mastered, the stronger your play,” says Gopinathan.
While working as the chief data scientist at Facebook, Jeff Hammerbacher described how, on any given day, a scientist team would utilize Python, R and Hadoop, and then have to relay the analyses to colleagues. Additionally, a recent SiSense data professionals study found that 60 percent of respondents use three or more data warehouse and business intelligence interfaces.
Thankfully, there are ample resources on the Web to develop and hone your skills. Big Data University, for example, offers free resources to help data professionals gain proficiency in JAQL, MapReduce, Hive, Pig and others.
It’s also important to gain experience using these skills in the “real world.” Gopinathan advises aspiring data scientists to participate heavily in open-source projects and data contests, such as Kaggle, to practice utilizing technical, scientific and visual skills in real business scenarios.
I’d like to hear other suggestions on how individuals can take their career in data to the next level. What advice would you provide to aspiring data scientists? Please leave your thoughts in the comments section below.
Blog thumbnail image created by Andrew Hazlett.