How to Build a Business Intelligence Center of Excellence: Advice from 3 BI Thought Leaders
One of the major BI trends for 2017 that we’re seeing is an increasing number of organisations that are establishing a BI and analytic center of excellence (BI CoE). This trend, to a large extent, is being fuelled by the huge explosion of data and data tools and the apparent benefits of in-house analytics services. However, merely deploying a BI CoE doesn’t guarantee effectiveness. Rather, the organization should set up its BI center of excellence properly to facilitate quick delivery of insights on the firm’s products, services, customers, market initiative, and consequently adjust its strategies for optimal business benefits. In this article, Three BI CoE experts share tips on how to set up a business intelligence center of excellence properly, but first things first:
What Exactly, Is a Business Intelligence Center of Excellence?
A BI center of excellence is a shared platform (tools and people) within an organization for supporting effective implementation of performance management, information management, BI and analytics across the organization. A properly structured BI center of excellence drives firm-wide data integration capabilities, including data mining, data warehousing, migration, consolidation, synchronization, cross enterprise data sharing, as well as set up of data hubs.
A BI CoE team has the collective know-how, time, and other resources required to optimize BI investment by developing BI strategies and producing business insights about the firm’s operations, services and products.
The primary goal of setting up a BI CoE is to capitalize on the economies of scale that arise from pooling and sharing expertise, process, tools, and people.
BI CoE Return on Investment
Each day, organizations are inundated by data from more sources than they can cope with. A BI CoE helps an organization make sense of this data. For example, if we consider a firm that wants to know how their clients will react to a marketing campaign, besides analyzing current trends in the market, the BI CoE could shed more light by comparing the specific campaign against past campaigns, when conditions were different, and mapping customer behaviours accurately. A BI CoE produces the needed insights to guide business decisions, and as such, it’s a vital tool for optimizing ROI on marketing programs and other initiatives.
Organizations that set up BI systematically prove more adept in producing powerful and focused business intelligence than organizations that implement BI CoE in a haphazard way, leaving the workload to groups or individuals without the skills, time, or mindset to drive deep business intelligence.
But How Should You Build a BI Centre of Excellence? Advice from Three BI Thought Leaders
When it comes to implementing a BI CoE, it’s important to keep an eye on what others in your niche are doing, but there is no one size fits all solution. Dan Johnson says, “In my experience the size, shape and organizational position of the CoE will change significantly over its first five years, so make the start that feels right given the organization culture and politics.” He advises, “Perhaps implement BI CoE on a test basis initially – incubate the CoE in the function with the most potential and enthusiasm and expand from there.” The following aspects are used to evaluate the worth of a BI CoE”:
Operation efficiency – Will the BI CoE facilitate more effective use of information? Will it lead to more relevant, consistent, accurate and timely insights?
Strategic Transformation – Will the CoE facilitate discovery of ways through which the organization can use data to transform its operations, markets, products, etc.
Total Cost of Ownership – Will the BI CoE facilitate efficient use of technology by eradicating operational redundancies and technical obstacles?
Johnson’s suggestion is supported by Tom Mucha of Charles Taylor InsureTech Limited. He posits, “The enterprise should first determine what it wants, otherwise, the project will fail.” Mucha suggests a step by step approach. He says, “The big thing is step by step. A lot of IT projects and BI CoE falls under this, fail because everyone is looking at the big bang approach – they want to deliver everything at once rather than take baby steps.”
Kunal Gupta, says that for successful implementation of BI CoE, the organization should start by evaluating its capability and the value of the possible business insights. He says, “This comes down to capability. You must have a very good understanding of where capability and appetite for insights sits within your business. Once you understand what you are good at and what you are not good at, and which areas need more support, and you have assessed the cultural change that needs to take place, then you can start to put a plan in place.” Make sure the BI investment plan is tightly tied to the organization strategy; better reporting, cost avoidance, cost saving, and risk reduction and avoidance.
Choose the BI CoE Sponsor
Johnson says, “Securing effective executive sponsorship of the BI CoE is vital. Hopefully this is the ExCo member that owns the vision for how data and analytics will create value for the organization – the greater the perception that BI is critical to the success of the organization, the more help the CoE will receive.
Choose the BI CoE Implementation Model
Gupta says, “Once you are done determining the cultural change that needs to happen for people to adopt the new way of working, you can start thinking about the right BI CoE implementation model.” He adds, “On one end you have a fully centralized CoE and on the other end you have a fully decentralized model with a good level of analytical capability distributed throughout the business.” He adds, “The end goal for me, always, is to have the capability embedded within the business to facilitate decentralization as much as possible. “However, Gupta agrees with Johnson that each organization is unique. He says, “Every organization needs to take a really hard look and figure out what the right model is.” In case the organization does not have data scientists, Mucha suggests that the BI CoE should be made as user friendly as possible, but cautions that “doing so limits the number of options the user has to manipulate the data.”
Governance, Governance, Governance
It’s highly unlikely (although it would be nice!) that the CoE will have unlimited resources and funds, so the priorities of the BI CoE have to be carefully set, costs have to be carefully controlled, and the benefits case should be proven. This is where good governance is essential.
Governance takes on many forms and operates at various different levels, but it all starts with the strategic. The strategic governance board is comprised of senior business leaders from the organisation’s major business units, divisions and functions, and chaired by the executive sponsor. It acts as the custodian of the BI vision and strategy and collectively decides how and where to spend the companies money and focus its precious BI resource.
Other governance cascades from there: operational steering commitees are focused on the success of specific projects; programme-level steering committee takes a birds-eye view of the various projects to avoid dangerous resource crunches; and then there’s the ‘design authority’ type of governance that ensures that the BI solution as a whole hangs together properly with the best user-experience.
Moreover, governance is a really good way of ensuring that the whole organization has a stake in the BI CoE.