Building a Data Science Team Innovatively 22 Apr 2015

Building a Data Science Team Innovatively

In an ultra-competitive business world, shortages of any kind are often used as proxies of how important a service, skill or commodity is in “real life” situations. A recent study by Gartner revealed the likelihood of a talent shortage involving data scientists — demand for qualified individuals in this specialized profession is increasing at a rate that is three times faster than demand for business intelligence analysts and statisticians. Through the year 2020, data scientist demand is projected to exceed supply by more than 100,000.

What should businesses do with this kind of information? In this article, Research Optimus shares some tips about building a top-tier data science team.

Who Are Data Scientists?

What do data scientists do? In many cases, a data scientist is on a team of big data experts — key tasks typically include writing code and working with the rest of the team to create websites, data dashboards, modules, packages and pipelines. This entails understanding business challenges, formulating actionable insights and communicating findings to business executives and managers. Like many individuals in the modern world of science and commerce, data scientists are required to wear many hats:

  • Scientist – Testing theories by running experiments.
  • Manager – Training and guiding quantitative professionals.
  • Engineer – Acquiring and managing data plus maintaining systems.
  • Programmer – Visualizing data and developing statistical models.

One reason for the growing difficulty of filling data scientist jobs is the shortage of qualified applicants that possess all of the specialized skill sets just noted.

What Comprises a Data Science Team?

Three types of experts are required for a “first class” data science team. Typically, business analysts work with front-end tools. Data engineers capture, store and process data. Machine-learning experts build data models that include quality checks for accuracy and lack of bias.

Qualities and Qualifications of Data Scientists Team

Competitive business analysis and data analytics for business should include an effective data science team – a “great” team will usually possess four distinctive qualities:

  1. Curiosity – Constantly asking (and answering) “how” and “why” questions.
  2. The Art and Science of Communication – The ability to translate problems into workable solutions often hinges on listening and communicating effectively.
  3. Professional Grit – While this might seem like a “non-technical” attribute, it is an essential quality that refers to a long-term motivation and passion for overcoming new and unknown challenges.
  4. Creativity – Seeking new and multiple paths for problem-solving and communicating.

Challenges in Creating Strong Data Science Teams

As noted in the opening paragraph of this article, a shortage of talent is a primary constraint when organizations are striving to build a world-class data science team. Some companies are trying to address the data scientist talent shortage by hiring managers and analysts with less-specialized skills. However, McKinsey reports that this might not work well in practice — shortages in the United States are projected to be as high as 1.5 million analysts and managers with skills for making “Big Data” decisions.

With the seemingly unavoidable talent shortages facing them, businesses can be tempted to “compromise” specialized skill requirements when filling data scientist jobs. This will rarely prove to be a winning strategy and should be avoided at all costs.

How to Build a Strong Team of Data Scientists

The advantage of using a team concept to fulfil data science needs is that it is difficult to find the “superhuman” skills in one individual – and this will become increasingly difficult in coming years. Organizations as diverse as PayPal and American International Group (AIG) have successfully assembled data scientist teams. For example, AIG uses a team of engineers, system architects, project managers, business analysts and statisticians for their Property and Casualty business. For smaller companies, the team size will vary depending on project complexity and duration.

Outsourcing as a Practical Data Science Solution

Companies of all sizes should constantly be searching for new ways to cope with data management and “Big Data” – these needs often range from competitive business analysis to dashboard preparation and trend analysis. Even if companies have not previously relied upon outsourcing as a business strategy, the shortage of critical labor skills such as those needed for data scientists provides a compelling and practical reason to think about using data science experts such as Research Optimus to help solve what appears to be an “unsolvable” problem.

Please share your data science experiences by leaving a comment below and using the social media icons to communicate with your circle of friends and business colleagues.

– Research Optimus

Related Posts