Universities push to deliver well-rounded analytics talent
Short Takes...on Analytics
A blog by James Guszcza, US chief data scientist for Deloitte Consulting LLP
Is there a data scientist in your life? If so, you know how talented these individuals can be. Even those who went to school before analytics was “a thing” have helped companies in a range of industries use technology and data to improve their decision making and business processes.
While data science has come a long way without formal, college-level analytics programs, universities are beginning to up their game, creating programs to train new generations of data scientists and quantitative analysts with the relevant skills.
As the tools, methods, and the larger data economy rapidly evolve, universities will face the challenge of meeting ever-growing expectations. Most businesses want high-quality analytical talent, not button pushers. How can universities offer the right courses and instruction to turn the right skillsets loose in the marketplace?
A head of STEAM – characterizing analytics talent
Traditional STEM – science, technology, engineering, and mathematics – courses have long provided the foundation of curricula that prepare students for technology-related careers. Today, a growing number of academics and business people are effectively asking that “A” - for “Analytics” - be added so that STEM becomes STEAM. STEAM capabilities encompass the multidisciplinary base of IT, math/statistics, and computer science skills needed to work with organizations’ growing volumes and varieties of data. Relevant skills include:
- Statistics and machine learning methods
- Statistical and mainstream programming languages
- Data visualization tools and methods
- Distributed file system tools, such as Hadoop
- Nontraditional database tools
- Statistical and natural language processing tools
And while data scientists need solid technical backgrounds, they also need the business acumen to turn data into practical insights, reliable predictions, and useful recommendations. Science and technology alone simply won’t cut it. Effective data scientists must have the ability to design technical solutions that address business questions; and then communicate their findings back to business users and decision makers in a way the end-users understand. That's no simple task. For universities, this means that data science and analytics curricula should encompass not only scientific and technological topics but business and communications content as well. It might be useful to think in terms of three C’s: communication, critical thinking, and creativity. These traditional skills, best inculcated by liberal arts and science curricula, are relevant in the age of designing and working with rapidly evolving “smart technologies”.
Well-rounded data scientists therefore possess a combination of business, technical, and communication skills. However it is equally true that different data scientists possess the various skills in different proportions. Some are programming or data management whizzes; some are business-savvy walking encyclopedias of their domains; some have deep knowledge of statistics or machine learning methods. It is typically unrealistic to aim for teams of ideally well-balanced data scientists. Many sophisticated organizations aim instead to build balanced teams of collaboratively-minded data scientists who are able to pragmatically complement each other’s strengths and counterbalance each other’s weaknesses. . In addition, analytically-minded businesses assess which skills they will need internally in the long-term, and which skills are best brought in from outside the business for special projects. In this way, organizations can tap into the needed range of skills and abilities in a cost-effective way.
Perhaps the most important qualification of up-and-coming analytics graduates isn’t their STEM or STEAM know-how, but instead an often-overlooked quality that is difficult to measure: Curiosity. An International Institute for Analytics study of more than 300 analytics professionals found that curiosity was the number one characteristic of data scientists and analytics professionals. Curiosity – the fourth “C” – is what often motivates one to learn about new domains, pick up new tools and programming languages, master new analytical techniques, and engage in type of associative thinking that leads to innovation. It is an abiding trait that serves data scientists – and their employers – well in rapidly changing fields. Both in academia and in business, more can and should be done to identify, nurture, and leverage the innate curiosity of scientific and analytically-minded individuals to solve complex problems.
Beyond technology: Appealing to quants
Once the universities turn out analytics graduates, how can companies effectively recruit them? Some considerations include:
- Post interesting brief write-ups of intriguing analytics projects the company has completed to appeal to candidates' curiosity and garner their attention.
- Feature advanced technologies in your recruiting ads to catch the attention of candidates who want to cultivate their skills and get their hands on state of the art tools.
- If your business has an older infrastructure, consider creating specialized groups where data scientists can build new systems and applications that integrate with legacy systems.
- Connect analytics professionals on your staff with marketing so that quantitative insights inform your marketing efforts - and vice versa.
It is important to convey to prospective analytics talent that you take their contributions seriously. Data scientists are attracted to companies that value their ideas and apply their work to real-world customer challenges.
The well-rounded graduate
For universities, the goal for analytics education should have less to do with churning out data science superheroes and more with nurturing well-rounded, business-oriented, and scientifically-minded problem-solvers. Such people are prepared to continually learn throughout their careers and are able to design, perform, and effectively communicate data-rich and analytically informed solutions to business problems. That may sound like a tall order, but it’s a much more important – and achievable – goal than graduating quants who are interested in data for data’s sake.