Building your Data Science Team: Business Engineering Unit

Every day, we create 2.5 Quintillion bytes of data – so much that 90% of the data in the world has been created in last 2 years alone. Every organization is now accumulating Terabytes and Petabytes of data coming from various devices – machines, mobile, user, web logs and cookies, social media, application servers and transactional logs, etc. Organizations are rushing to store this wealth of information fearing missed opportunities. The challenge is not in storing this information but able to find usage of this data to its competitive advantage.

Leading organizations are rapidly reorganizing themselves and building the Data Science Team. This is very similar pattern that you may have noticed in early to mid 90’s wherein organization started with Business Intelligence team rebranding MIS or DSS teams. The key to competitive advantage is not by storing all and most of the data but by deriving value and insight from it and be able to tie it with business plan that can drive tangible business outcomes. This new function that is evolving to be a Data Science team primarily needs four tiered layer of experts: Data Science Champion, Business/Product Managers, Analytical Data Modeler and Big Data Engineer. To make the Data Science team successful, it’s key that they operate under a Data Science Champion such as Chief Data Officer or Chief Data Scientist and not under a traditional IT organization.

Let us define what each of these layers mean to an organization.

§         Data Science Champion: This is an Executive level sponsor such as Chief Data Officer or Chief Data Scientist. They lay out the vision and lead the mission for the team. This champion is a domain expert in a data science field. A large organization may have several Chief Data Scientist roles specific to LOB working for a Chief Data Officer.

§         Business/Product Managers: They are the product managers from the business that closely partners with the CDO or Chief Data Scientist

§         Analytical Data Modeler: These are the advanced mathematician and statisticians who can apply their computer skills to create complex analytical models largely contributing to the mission

§         Big Data Engineer: These are the computer science engineers who apply the sophisticated engineering skills to process large volumes of datasets using big data tools.

It’s hard to find a Data Scientist who will have all the above 4 skills – leadership, business knowledge, analytical experience, big data processing skills. A successful Data Science Team is partnership of above defined four distinct roles. When building your team, it’s important to focus on these key data science skills: analytical and curiosity mind, creativity, domain knowledge, advanced math skills including a solid background in calculus, geometry, linear algebra, and statistics and a computer science background. Leadership, Analytical and Computer Science skills are important, particularly for the first members chosen for your data science team.

A new concept of Business Engineering Unit under the leadership of CDO/Chief Data Scientist is slowly evolving that will house the Data Science team. This has been seen in some of the large retailers, telecom, and media/entertainment companies. Data Science deals with identifying the real problem or a business opportunity. Unlike IT as a support organization in many companies, the Business Engineering Unit needs to be a profit center that will contribute to the company’s topline or bottomline. To prove the value from Data Science, the Champion needs to initially focus on the hardest business problems within an organization or unique business opportunity that have the highest return for key stakeholders. Organization boundaries and internal political environment are often the biggest challenges facing a data science team.

Companies that have made investments in big-data computing will reap extraordinary near-term and long-term benefits. Data Science is perhaps the biggest innovation in computing in the last decade.  The benefits from data science have already been proven in some industry sectors; the challenge is to extend the technology and to apply it more widely and in all facets of interaction between humans and machines.