Should Organization Build a SINGLE Center of Excellence for Business Analytics

Analytics is not for division, but for unification. 

More and more companies are realizing that analytics is a core skill and all the learning of analytics is applicable across the organization. So the evolution of Data Analytics combined with greater use of predictive, big data is leading companies to think further about the 'Analytics Center of Excellence'. Should there be a central CoE Analytics that oversees and governs all business analytics, or multiple CoE's based on pillar, business unit or domain?

• There are high level corporate issues that need to be defined and passed to branches so there's some consistency in KPIs and decision processes. Analytics being performed to support decision making would be more effective if they are aligned closely to the senior managers making the decisions. The level of involvement from the decision makers would be quite high in business critical and time sensitive requests compared to an instance where the center is producing reports based on mature and established guidelines and processes. 

• Having a single center of excellence is ideal, but is not feasible for every company. A single source of Truth makes sense, it brings with it more control, governance, flexibility.  However, the success of a center of excellence will work best if the company is uniform in structure and the data that is required. It does not work so well when there is not any uniformity. Utilizing regional centers of excellence are more feasible when there is no uniformity, even in such a situation it is important to have this regional/divisional centers work closely together ideally under one management. 

• Larger corporations require a more centralized structure, to ensure basics like data governance is complied to and ensure minimal disruption to operational deliverables. At the same time every line of business should have its own analytical champion to lead the analytics for that specific LoB but in very close cooperation with the center of excellence or the BI-CC. But how feasible is it in any company of reasonable size where data is all over the place, and with different data owners/gatekeepers? It depends on the organizational maturity of information management. 

• The analytics team structure may also depend on the BI maturity of the organization, dynamics of the BI adoption in the enterprise. As organization matures towards “Sage” stage in analytics, the goal is eventually creating a symbiotic federated system. You need to empower each organization with the ability to define what analytics are needed for their individual business situations for local (whether that's region, department, etc) strategy and tactics 

• An Effective CoE Analytics may help unify the view and target the customer with tailored, evolutionary analytics solution. Because you can discover a lot more "truth" about your business when you have a unified view across all silos of departments/business. Operational leaders need KPI-oriented analytics that are grounded in measurements that are owned by and used consistently in operations (business analytics). Strategic decision makers will need data scientists who build new models that identify trends and project them into the future (predictive analytics). Both need descriptive analytics that effectively standardize internal transactions across the enterprise and relate this to external sources for benchmarking with industry standards 

• CoE Analytics or the data governance body such as ‘Data Analytics Steering Committee’ share the  sets of objectives such as 
  
  
  1). Ensure information is consistently defined and well understood.
  2). Create trusted data as an enterprise asset
  3). Improve the consistency of data use across an enterprise
  4). Identify external data to compliment internal data
  5). Ensure the availability of data at the right time and in the right format to enable data-driven decisions.
  6). Create vision for how analytics will be used in the organization and identify the specific capabilities necessary.

7). Formalize the process of targeting as a collaborative process between business/IT/ analytic leaders. Focus on high value, high impact targets

8) Set the expectation that decisions will be based on data and analysis. .
9). Promote and provide delivery enablement through a consistent set of analytics skills, standards, and best practices
10). Enable repeatable successful analytics deployment through the development and focus of people, technology and process—in ways that makes sense to an entire organization or division, rather than just a ‘single project’ 

Posted in: Big Data,Business Intelligence