A career in decision sciences/analytics continues to be one of the sexiest jobs of the 21st century, but the supply of analytics talent threatens to limit the promise of decision sciences. A report by McKinsey and Company estimates a shortfall of 140,000 to 190,000 data scientists and 1.5 million managers who have the skills needed to use the insights to drive decisions. And Gartner predicts that by 2015, big data will create 4.4 million jobs globally. Data scientists are in short supply, but the dearth of decision scientists – the rare breed that combines the interdisciplinary prowess of math, business, technology, behavioral sciences, and design thinking – is even more alarming. For this reason, there needs to be an increased emphasis on recruiting and training as opposed to relying on acquisition.
Here are five strategies to help recruit and train the analytics maven known today as a “decision scientist.”
1. Ignore the Top-Tier Myth
The idea that the right analytics talent can be found only in top-tier institutions is a myth. The fact is that there isn’t enough talent. The focus needs to shift from attracting and acquiring talent to creating and nurturing talent. For organizations, this means broadening the hiring universe, moving beyond top-tier institutions and finding talented people who can be groomed into decision sciences professionals.
People who can become quality decision scientists showcase many skills, including a quantitative bend of mind with creative problem solving skills; the ability to communicate, synthesize, and articulate learning (new skills, new domains, etc.); and the ability to work in a group setting with people from diverse backgrounds.
Organizations must rethink their hiring strategies to accommodate testing for the aforementioned skills. And by not limiting their searches to top-tier institutions, organizations can spend their hiring budget on mid-tier or local state schools. This will allow visits to greater number of campuses, yielding a greater return on investment as the effort in training newly acquired talent with what can be considered “transferable or teachable skills” is about the same.
2. Test for Curiosity and a Learning Mindset
“I have no special talent. I am only passionately curious.” — Albert Einstein
Knowledge is no longer the prime asset for organizations – learning is. Organizations today compete on their ability to learn, but tomorrow they might compete on the rate of change of learning. With change being the only constant, learning outplays knowledge, and a student is more effective when he/she learns and adapts faster. Professionals should have the openness to learn from first principles, and one way to test this is by gauging curiosity quotient.
This would test not how much a candidate knows, but how much he/she has demonstrated learning, inquisitiveness, and curiosity, both in their academic and extracurricular activities. It will help the recruiter to understand if the candidate has pursued “out of the box” approaches to solve a challenge or if candidates have gone out of their regular course work and pursued another topic.
By and large, organizations must experiment and introduce innovative recruiting strategies to rightly assess skills at the desired level of employment. Recruiters should test candidates’ tendency to embrace new challenges, and their willingness to learn from criticism through problem and puzzle solving.
Once a candidate is in the system, focus should be on imparting continuous and holistic training, including consulting skills, structured thinking, basic and advanced statistics and analytical techniques, design thinking, understanding of industry verticals, and business functions such as marketing, risk, and supply chain.
3. Appreciate the Inter-Disciplinary Perspective
Most companies look for skills that are very specific to the actual job description at hand, but decision scientists need to be multi-faceted and not just number crunchers. The traditional approach of viewing data analytics skillsets in silos must change, and focus should shift from finding experts in one discipline to creating professionals with multi-disciplinary skills. Candidates must possess not only strong quantitative and technical skills, but also business understanding, the ability to abstract and synthesize, skills to balance hypothesis and discovery-driven problem solving, as well as communication skills. The former are required for efficient problem solving, and the latter are just as important to transform analytical solutions to business solutions and vice versa.
As a decision scientist moves up the ladder, the required skillset changes. Client relationships, people management, and delivery management across global teams become essential. Analytics companies should seek the right skill set, mindset, and tool set coupled with quick learning and the ability to adapt to a rapidly evolving and growing environment. They must devise recruitment strategies and interview series to assess skills at different level of employment, and find candidates best suited for them.
4. Teach the Art of Asking Questions
Problems are singular, answers may not be. One question can have many answers. That’s why emphasis must be placed on asking and probing. For example, focus more on the Why as opposed to the What so the problem becomes clear and appropriate solutions can be devised.
Organizations must assess employees’ backgrounds and devise a training plan that would make them well rounded, while focusing on questioning. An appreciation for different types of questions that can be asked, and when to use each one, is critical. Organizations should build a curriculum that underscores asking questions and using that element in day-to-day problem solving, where question definition and re-definition can happen.
For candidates, asking the right questions and having a point of view or hypothesis is equally important. The point of view does not have to be right every time, but each point of view or hypothesis should be tested, and that candidates must be able to let facts rule while testing.
If the point of view is proven right mathematically, then one makes a measurement. If it is proven wrong, then one makes a discovery. It’s a win-win situation.
5. Learn to Rapidly Disaggregate and Aggregate Problems
The life cycle of problems includes disaggregation, aggregation at multi-levels, and communication.
The problem-solving strengths of decision scientists create a natural inclination to disaggregate business problems into numerous analytical parts. Most decision scientists approach this stage well. However, aggregation is a muscle that needs to be built. This needs to be done at multiple levels. At a base level, it involves building hypotheses for each problem and deriving connections between hypotheses. At the next level, it involves translating connected hypotheses into messages for the immediate problem. Finally, the messages for each problem need to be woven into a connected narrative that explains the impact and answers the question, “So what?”
In the end, candidates should be able not only to disaggregate problems but also to connect the dots and aggregate the focused analytics questions to provide insight and catalyze impact.
Rajat Mishra is a seasoned business leader with 15+ years of experience in management consulting (McKinsey), technology (Microsoft), and business intelligence (Google). At Mu Sigma, Rajat is leading transformative change in decision-making processes across various clients in the San Francisco Bay Area. Rajat holds an MBA from the Wharton School of Business at University of Pennsylvania, where he was on the Director’s List, and a Bachelors in Technology from IIT Delhi, where he was on the Dean’s list.
The following authors contributed to this article: Rajnita Kamath (Engagement Manager, Mu Sigma), Kaushik Manchala (Associate, Mu Sigma), Mateo Gonzalez (Junior Associate, Mu Sigma), Syed Ahmed Ali (Junior Associate, Mu Sigma)
Subscribe to Data Informed for the latest information and news on big data and analytics for the enterprise.