Analytics Is a Team Sport

By: John Tardy

One of the things I enjoy about our work in analytics is the variety of business problems we solve. While the specifics of the problem vary, our overall objective is always to derive value from the data. Putting the client’s needs first means that we adjust our approach to best fit the situation rather than fixate on a specific tool or pre-selected solution. This variety makes the work exciting, but it also creates challenges in staffing the appropriate skills. Over the years, I have come to refer to our solution to this challenge with the phrase “Analytics is a team sport.”

While there are common foundations, no one person will have all of the skills that could be needed to effectively solve all of the possible needs we may face. One phase of the project may need a heavy dose of data engineering, requiring heavy SQL and data preparations. Another may require an eye for data visualization and development of interactive dashboards. Statistical analysis is its own specialty with a mathematics base. Finally, predictive models involve their own discipline and expertise.

Sometimes a project will focus primarily on one of these areas, but other engagements will shift from one specialty to another over the course of work. Our consultants have a broad range of experience and skill sets, which means that they are much better positioned for success. Still, there is no “unicorn” that has expertise across all these areas.

We address these challenges through communication and collaboration.  “Analytics is a team sport” means that you do not have to solve it all yourself. No one stands alone. We maintain a team communication channel across mobile and desktop that allows any of our analysts to connect to the rest of the team as a sounding board. Sometimes that involves a specific technical question, or “has anyone seen this before.” In other cases where a different expertise is required for a longer period, we will shift the allocation of resources on the project to best fit the client’s needs.

A key driver for data project failure rates that are quoted at 80% and above is that all of this is easier said than done. I have seen many situations where a company will hire in a specific skill set and then expect that they are going to address the entirety of the business problem. This miss between expectations and true capability raises the second part of this challenge. The coordination role itself requires an understanding and expertise in the range of problem domains and potential solutions in order to effectively manage the team. While our team has the ability to address these challenges holistically, we also have the capability of partnering with a client team to augment the skill sets and/or help them develop the capability.

In the end, I love to solve problems, and I am grateful to work in such a dynamic and exciting area and have the opportunity to be part of a talented and cohesive team.