It might seem slightly counterintuitive that a center of excellence is about the entire enterprise. But it’s true. Becoming insights- or data-driven is a big mandate. The CDO or equivalent has taken on a massive task. And just how does the CDO do it? The answer: certainly not alone. I’m not going to say, “It takes a village.” But, oops, I did. It takes the whole enterprise. With requests for data and insights coming from all corners of the company, data and analytics leaders must help capture needs, assess skills, and build an insights center of excellence to support the insights-to-impact goal. It might be called a “center” but it represents all groups and the broader business goals.
Why a Center of Excellence?
A center of excellence (CoE) isn’t about creating an ivory tower. It’s about building a broader base of knowledge and understanding of data and analytics. It’s about capturing requirements, coordinating initiatives, and galvanizing resources. By creating a CoE, leaders highlight the strategic nature of data and analytics within their organizations.
Data and analytics CoEs help:
- Define and disseminate policies and best practices for data governance, data operations, and program management that streamline processes and facilitate value creation.
- Foster a deeper data and analytics culture by creating a language common to both tech and business teams and aligning incentives around delivering business value.
- Sponsor training to build broader data literacy, enhance data-driven decision-making, improve data and analytic skills, and allocate resources across the organization.
- Establish a sense of community and encourage cross-organizational participation.
- Deliver business value through data and analytics.
According to Forrester Research, 45% of survey respondents at organizations with CDOs have implemented or are expanding insights CoEs, compared to only 32% of those without a CDO.1 Not surprisingly, it is the most analytics-mature companies who are leading the way. But that cuts across industries, regions, and even company size. For example, The Wall Street Journal reports that Morgan Stanley’s CoE includes about 30 experts specializing in data architecture, infrastructure, and governance.2 They act as data advisers to different business and technology divisions within the bank to accelerate its insights-driven transformation.
How Should a CoE Be Structured?
CoEs are not one-size-fits-all but tend to reflect the needs—and the current maturity level—of the company. CoEs are not necessarily a fixed organization or structure but can be thought of as more of an organizing principle. The CoE will have a core team, a hub, with members either reporting directly to the CDO or via a dotted line to ensure coordination. Each of the business units or functional areas of the organization houses a spoke that connects into the hub. Responsibilities will be distributed across the hubs and spokes with some gray areas that vary across organizations.
- The Hub. At the hub of a CoE, data insights leaders drive strategy, establish enterprise-wide data governance, and deliver consultative services including talent recruitment, training, technology acquisition, vendor management, and data and analytic services. The degree of responsibility and involvement in these activities depends on the level of maturity of the rest of the organization. The most common roles lie in coordinating governance policies, skills assessment and development, and other enterprise-wide processes.
- The Spokes. The spokes of a CoE lie within the lines of business, functional teams, or geographical units. Where the degree of analytic maturity is high, the spokes are most closely associated with the actual data and analytics initiatives. They define use cases, output requirements (reports, dashboards, machine learning models, and so on), oversee data execution, develop models, track performance, and measure outcomes.
- Gray areas. The gray areas reflect shared responsibilities depending on spoke-level capabilities. For functional teams or business units without the required skills, the hubs can assign resources to help develop and execute an initiative.
How to Get Started
Establishing a data and analytics center of excellence might seem daunting at first. It’s best to not try to boil the ocean from the start. CDO and leaders initiating the effort need to enlist both top-down and bottom-up support and demonstrate value quickly. To do so they must:
- Establish a steering committee with broad representation to ensure joint vision. Ideally, the CoE will report to a high level in the corporate structure and have the sponsorship of both business and technical leaders. The key here is (1) executive sponsorship and (2) an agreed-upon mandate to coordinate cross-company initiatives and resource allocation. This can be the same steering committee established for enterprise-wide data governance and prioritization.
- Pick an initial project to test the process and demonstrate value. The prioritization process will identify potential lighthouse initiatives. There might be existing projects across different divisions, but the lighthouse CoE project should demonstrate how central resources can advance cross-departmental goals that were previously challenged by issues such as data silos or definitional differences. An ideal candidate would be a project that leveraged data across data domains. However, it’s best not to take on something too big. Choose some low-hanging fruit to start with or a piece of the bigger initiative to show incremental value.
- Assess competencies to determine distribution of responsibilities. To know which competencies will be housed in the hub versus decentralized in business units or functional areas, data and analytics leaders must get a good understanding of who can do what across the company. Some spokes might be more advanced than others with teams of business analysts and data scientists in-house. Think finance with skills in forecasting or risk detection or customer insights with expertise in demand trending and personalized offers. Other groups, such as legal or HR, might not have such skills. To ensure broad access to insights, the hub might have a few roving data scientists who go on assignment to help build HR retention models or understand the impact of new workers compensation laws and optimize legal spend.
- Clearly communicate service offerings and value delivered. At a minimum, most CoEs coordinate data governance policies and the allocation of resources across the organization. They are not necessarily shared service units. However, some CoEs act as service bureaus providing data and analytics expertise to both business units and function teams, and sometimes even to partners and customers. Over time these CoEs develop a portfolio of repeatable services and, most importantly, promote its successes.
How to Ensure Ongoing CoE Success
A data and analytics CoE is not a project management office with a fixed start and end date. This is a permanent organizational framework designed to advance business goals through data and analytics. Ongoing success requires both top-down and bottom-up support. Successful CoEs demonstrate business value and promote their contribution to the strategic goals of the organization. Always remember what Thomas Edison said: “The value of an idea lies in the using of it.” The same goes for data. The ultimate goal of the CoE is to ensure that data is accessible and is used across the organization to drive value. Delivering value is the ultimate measure of success.