Data Analytics - Centralized or Decentralized?

Kim Wienzierl, Director of Data, Radio Systems Corporation

Kim Wienzierl, Director of Data, Radio Systems Corporation

How should your Data Analytics functions be organizationally aligned in your company? That is a very good question, and like many things, the answer is “it depends.”

Let us start with a definition of what a Data Analytics function does. This is the team that converts data into information, knowledge, and wisdom to optimize business performance, maximize profit or make more strategically sound decisions. They typically write reports and dashboards for operational use or one-off analytics. Team members may have job titles such as Data Analyst, Business Intelligence (BI) Analyst, or BI Specialist. Depending on your company, the data may not be fully curated, which means the Data Analytics team members fulfill additional roles such as Data Engineer, Business Analyst, and Quality Assurance Analyst to prepare the data to be converted into information, knowledge, and wisdom.

The question remains — should the Data Analytics function be centrally working out of an enterprise-wide analytics team, or should the function be decentralized with Data Analytic functions embedded in the business departments like Finance, Sales, Supply Chain, Human Resources, Marketing, or Engineering?

The answer is both. If your company happens to be doing little to no analytics, it is a good idea to start with a centralized team to develop the technology platform, curated data sets, data catalog, and standard processes for doing data analytics. Today, it is rare to find a company starting with no analytics at all. Chances are your company already has some data functions underway. With that in mind, it is best to have a centralized team to function as the center of expertise while also embedding Data Analysts in key business areas where there is great value in converting data into wisdom. This does not mean that every business department needs its own Data Analysts. Some departments would fare well using the centralized team when needed. These are typically internal service-oriented teams such as Human Resources or Facility Management.

The pros to having a centralized Data Analytics team are that the curated data sets become stronger and more reusable across a shorter period of time because the teammates are working closely together, and they are able to cover a broad area of topics. The pros to having a decentralized Data Analytics team are that the expertise in each business area grows deep very quickly, and the work is not being prioritized against the rest of the company, so insight is gained rapidly.

Both centralized and decentralized data analytics can work well, but the key is to determine when and where to place the functions at your company.

A con to having a centralized Data Analytics team is the difficulty of prioritizing the analytics work since it is usually urgent to every department. It may also be challenging to specialize in a specific business department if the requests are coming from a multitude of areas. A con to having a decentralized Data Analytics team is that they may push much of the data engineering work back to the centralized team. There is a chance that they also may become siloed with their knowledge. These team members could get assigned other departmental work deemed an emergency and never get the chance to dive into the analytics work. Also, business leaders may not know how to adequately train, in cent, career path and reward someone embedded in the business in a Data Analyst role.

To summarize, having both centralized and decentralized data analytics can work well, but the key is to determine when and where to place the functions at your company.