At some point, every digital marketing team has to think through how it does its analytics work. Whether you are in an agency working with clients or in-house directly managing campaigns, analytics eventually reaches a point where it can’t be done by just anyone. The time and analyses needed are too big to be a side project. It’s at this point that you will need to think through where the future lies when it comes to analytics in your team, agency, or company.

Much like assessing your PPC data and tooling needs, thinking through how best to build an analytics team involves asking lots of questions and thinking through a structure that suits your needs. Most important is understanding how an analytics team can be built to scale with your needs. Without careful consideration, these teams are often put together based on short term needs at the expense of a long term vision.

Having the ability to scale means that an analytics team can grow and adapt to a changing environment with the least disruption. They can bring on more people, absorb new processes, and improve over time. Lacking the ability to scale means delays and disruptions for much-needed analyses.

Should You Have A Centralized or Decentralized Analytics Team?

Deciding whether your analytics team is going to be centralized or decentralized dictates the future of how you manage everything your analytics team does. From tooling to hiring to management, how your team is organized creates the foundation for the rest of the strategy.

Let’s define centralized and decentralized and look at the pros/cons of each.

A Centralized Team

A centralized team is a single team serving the entire company. It likely has a more generalized analysis function as its scope of tasks and responsibilities is much larger. Responsibilities could include building reporting for the executive team or analyzing the decline in sign-ups for a product. It might have subject matter experts, but in general, it has a wider breadth of knowledge at the expense of deeper skills on a specific topic or problem. These teams are often “farther” from the work and thus require more lead time and explanation to get up to speed on each project. 


  • The simplicity of one team serving the company
  • Easier to create and maintain standardized processes 
  • Easier to manage tools and costs


  • Analysts are not as close to the work or client needs and may need more guidance
  • Less responsive, slower to act due to a large variety of tasks

A Decentralized Team

A decentralized or embedded team is one that has a much more limited scope of work that it does. It might serve a single large client or a single product line within a company. Because of this, it has a much deeper knowledge of its specific scope and is very responsive to that need. A decentralized team will not be responsive to needs across multiple teams or the whole company. This structure may lead to multiple analytics teams with a narrow focus, resulting in increased management needs.


  • Closer to the work that needs to be done
  • More knowledgeable of the needs of the team or client
  • More responsive and efficient with their limited scope


  • Silos are likely to develop with skill gaps across teams
  • Higher chance of ad hoc tooling and processes that don’t scale
  • More, but smaller teams can lead to increased management problems


Having the right structure for your analytics team can pay dividends in the future. The right set up will scale, handle your unique needs, and provide value. Take the time to assess your needs, examine bottlenecks, and figure out whether a centralized or decentralized (or even a hybrid) might work best for you.