ABSTRACT: An Analytics Center of Excellence empowers business teams to meet their own information needs by changing the role of IT from developer to facilitator.

Most data & analytics leaders say they need an Analytics Center of Excellence (ACoE), but few know what that is or why they need it. And those who have an ACoE disagree about how to shape and implement it. 

Truth be told, an Analytics Center of Excellence can take many shapes and forms. That’s because every organization is unique: it has a distinct combination of business domains, units, and departments with different information needs and data & analytic skills. As such, every Analytics Center of Excellence must be designed and shaped to match the singular contours of an organization.

What is it? 

An Analytics Center of Excellence is a centralized resource of data & analytics experts whose primary objective is to help business teams and individuals make better use of data & analytics. It empowers business teams and individuals meet their own information needs without relying on or waiting for the IT department to do everything for them. It usually consists of three disciplines: business intelligence, analytics, and data science. It works together with the enterprise data team and local (or domain) development teams to optimize the use of data and analytics throughout the organization. (See figure 1.) 

Figure 1. Enterprise Data & Analytics Ecosystem

Background. Centers of Excellence became popular more than a decade ago to address the persistent problem of IT bottlenecks. Rather than try to make IT more agile—which proved stubbornly difficult to do—organizations created Centers of Excellence to liberate the business from IT. To do this, IT’s role changed from “developer” to “facilitator”. In this role, IT coaches, supports, and trains the business to build their own solutions rather than build solutions for them. 

Reality. In data & analytics, that’s not always possible. The reality is that only a few business teams want to take on IT’s role, usually finance, marketing, operations, and sales. The rest may hire a data analyst or two and wait for IT to meet their more complex needs or do without. Thus, IT’s role is more complicated than ever: it needs to be both a facilitator and a developer. Moreover, it needs to know which role makes sense for each business domain it supports—some have skilled data experts and analysts, while others have none. This makes it imperative to hire an experienced senior leader to oversee the entire enterprise data & analytics ecosystem.

Governance. The head of enterprise data & analytics oversees the Analytics Center of Excellence, although this could be a dotted line relationship if the Center sits outside of IT. The ACoE is assisted by a Data Governance Council and an Analytics Council, which are cross-functional committees of stakeholders who review and approve standards and strategies that govern data and analytics assets for the organization. Finally, the ACoE facilitates Communities of Practice of data and analytics experts throughout the organization.

The Analytics Center of Excellence is the face of IT to the business, even if it sits outside of IT.

The Analytics Center of Excellence is the face of IT to the business, even if it sits outside of IT. As such, it also plays a dual role. In its purest incarnation, it provides standards, best practices, and coaching to self-sufficient business domains that build their own solutions and generate their own insights (i.e., decentralized model). In other cases, the ACoE in conjunction with the enterprise data team develops local applications for business domains that don’t want to perform this work themselves (i.e., centralized model.) In most organizations, it plays both roles simultaneously since business teams have a range of analytics needs and skills (i.e., federated model.) (See figure 2.) 

Figure 2. Roles of an Analytics Center of Excellence

Three disciplines 

An Analytics Center of Excellence is usually comprised of experts in three disciplines: business intelligence (BI) experts who build reports and dashboards; data analysts who answer ad hoc questions using descriptive, diagnostic, predictive, and prescriptive analytics; and data science experts who build machine learning models (predictive and prescriptive). These experts can be placed on three separate teams, or they can be combined into one team, depending on business needs, available skills, size of the organization, and location of the talent (i.e., in business units or an enterprise team). 

Three in One. There is considerable overlap among these disciplines. All three require individuals to understand business requirements, source and manipulate data, analyze the results, and visualize and publish their findings. The differences are nuanced, if they exist at all: generally, BI developers work in scrum teams on approved projects that run against a data warehouse or data mart; data analysts answer ad hoc business questions using whatever data they can find and tools ranging from Excel and SQL to Python and R; and data scientists work on both ad hoc questions and formal projects using data science tools, ML/AI practices, and wide-flat tables of data that they create themselves, often with the help of data engineers. 

Summary

An Analytics Center of Excellence is designed to both liberate and empower business departments to meet their own information needs. It does this by changing the role of IT from developer to facilitator. Its mission is to educate, train, and coach the business how to generate insights and analytic solutions using data. This is a big change.

It also builds enterprise or cross-functional analytic applications since no department or business unit will assume this responsibility.

The reality is that IT needs be both a facilitator and a developer, making the task even harder. Through an Analytics Center of Excellence, IT empowers business domains to generate insights and build local solutions. But the ACoE also performs these tasks for business domains that don’t have the interest or capacity to do them on their own. 

Our next blog in this series will explore the functions of an Analytics Center of Excellence and how those functions are delivered across the three analytic disciplines: business intelligence, analytics, and data science.