Exafluence | Why are Analytics COEs so important for managing data initiatives these days

Why are Analytics COEs so important for managing data initiatives these days

Why are Analytics COEs so important for managing data initiatives these days


It is an accomplished fact today. Data analysis and preparation is as important – if not more – as building sophisticated models. It is also understood that the customer is spoilt for choice today – far more than ever. While Customer Centricity seems to be the byword for every major Digital Transformation initiative, there is little realization that organizations today have a very fragmented technology landscape – in most scenarios, a spanking new, state-of-the-art Big Data implementation could be sitting with myriad other data hubs and solutions it was expected to replace. Many healthcare companies see their provider data sitting in MDM tables across various data warehouses, besides dedicated MDM instances. Such issues are often exacerbated by inorganic growth – acquisitions and ineffective integration of IT departments. Small wonder that major BI or data migration or integration initiatives get started in siloes – siloes that cut across Governance, Knowledge and Project layers. The problem with fragmentation does not blight the world of databases alone. A typical healthcare or financial services company today sits on multiple sets of reporting tools – from enterprise tools bursting scheduled reports to newer data exploration and visualization solutions and good old spreadsheets. With such fragmentation of technology and data, is customer centricity truly being pursued by organizations? Do one-time data quality and governance initiatives solve chronic, on-going problems?

Today, more than ever, we need our experienced IT staff to focus on orchestration than troubleshooting. Ironically, managing the ever-increasing growth in data within organizations requires documenting and managing the How-Tos, thereby creating some knowledge on the growing pool of knowledge. You cannot afford to keep creating data assets without some discipline – standards, best practices, etc. Enter the Analytics Center of Excellence, or ACOE. Also called BI Competency Centers (BICC) or BICOE in other organizations, today, various companies are taking a hard look at running shared services within their IT departments focused on the data value chain (data generation to visualization).

So, what truly is an ACOE? Quite like how Operational Data Stores in some organizations could conveniently be labeled Enterprise Data Warehouses, a COE could be anything – a group of Subject Matter Experts chugging frameworks and models as a shared cost to the enterprise. Or they could be active revenue models providing analytics services on demand. Ideally speaking, an ACOE, in its mature “future state”, is an insights-driven organization that helps its stakeholders make quick, actionable decisions. The ACOE, in such a case, becomes an Information Management entity that integrates the entire data value chain, while working as a cohesive face of IT to the business. As a service oriented model, it ends up providing a low cost, high performance, scalable, industrialized analytics platform.

In some healthcare organizations we have worked for in the past, First Time Quality is an important byword. FTQ forms an important metric that makes the ACOE so compelling. Eventually, FTQ is not the only expectation for organizations. As we know already, organizational levers still remain the same as they were ages back – profitability, productivity, efficiency, etc., still drive all initiatives across industries today. Therefore, FTQ blended with tools and process reengineering results in FTQ with agility. In turn resulting in customer centricity.

At this stage, an important question circles around the Hows. How does one implement a successful Analytics Center within the IT enterprise? Of course, it all starts with a management vision and adequate commitment from the leadership. With the management’s blessings, a gap analysis and roadmapping exercise helps baseline your organization from the standpoint of its maturity as an insights-driven organization. A comprehensive review of processes, people and tools helps articulate the organization’s data goals. First Time Quality? Certainly an important goal. Can it be quantified and made SMART? Sure. A target could be to reduce QA overheads or even the number of production tickets by means of investing in upstream activities. Do you need to have data modelers for each silo working in a decentralized manner, adding entities to data models with little reuse of existing ones? Basically, the gap analysis reveals aspirational goals and sub-objectives and helps define key SLAs or operational metrics that you would track your performance with.

Once the foundation of the ACOE is established with some fundamental tools, standards, processes and people in place, all you need is to test its efficacy by means of demonstration. Handle a new data project, if you can, and measure your performance. Or, depending on what mandate is given to you, start as a support entity and test your skills in managing Continuous Improvement initiatives, and L3 requests.

An SLA driven ACOE then forms the template for the rest of the IT organization in the company. We have seen that it works – and it delivers across the board. Organizations need to reduce their time to market significantly, and the ACOE is the organization that can deliver them insights with First Time Quality and agility.