Key Considerations When Creating a Successful Centre of Excellence

2017 has been dubbed the year in which data and analytics finally reaches mainstream status. This fact has been reflected in a report by IDC, which found that revenue growth from information-based products will double the rest of the product/service portfolio for one third of Fortune 500 companies. Organisations are now in agreement that a well developed data analytics strategy is no longer an add-on, but rather a necessary component of modern day business which can effectively boost growth, deepen customer relationships, improve products and services, and optimise organisational performance.

However, a well defined data analytics strategy needs to be supported with an agile architecture, built to sustain the developments in technological innovation and scale solutions across the complex enterprise. This is by no means an easy feat and in response, some forward thinking companies have looked to implement a Centre of Excellence (CoE). A CoE may be created for a number of reasons, however, most CoEs are fundamentally incubators, which develop the standards, best practices and ideas for data analytics usage which are then applied to the rest of the enterprise.

Although CoEs do provide numerous benefits, their mere existence will not be a magical solution to all of your organisation’s most pressing concerns. Creating a CoE is an intensive process, one which requires a great deal of planning, time and resources. So, as you can imagine, these teams can be under great pressure and scrutiny to deliver on their investment. But the critical issue here isn’t whether a CoE can work, but ensuring their effort isn’t wasted.

Having spoken to a number of Chief Data and Analytics Officers, I have compiled some key factors to consider when creating and sustaining a successful CoE:

Establish the value proposition

Define the vision for your CoE early. This will also play a crucial role in measuring the value of its implementation later on in your journey. Understand that in most cases there exists a disparity in data and analytics understanding amongst the experts and business users. In order to close this gap, bring the business users into the conversation pre-conception and understand their needs. You are likely to have a greater degree of success by aligning the objectives of your CoE to the objectives of the business as a whole.

Bridge the gap with a robust engagement model

Once the vision for the CoE has been defined, the engagement model and scope must be established. This is perhaps one of the most critical stages in developing a CoE, as its ultimate success does lie in the ability for the organisation to understand and consume its output. In order to overcome the information asymmetry which may exist, common definitions, languages and metrics can help communicate the results of the CoE. A process of implementation also needs to be considered, which brings about the notion of actionable insights. Will the CoE operate as a service? Will the CoE be project-based, application-based or product-based entity? How will the communication chain look and how will the findings be reported?

Diversity and skills inclusion is a balancing act

“Unicorns are almost impossible to find, so there is no reason why these skills could not be distributed across the CoE. Ingraining diversity into the very fabric of your CoE and aligning the business users, story tellers, analysts and Data Scientists under one umbrella could create some devastatingly positive results.”

A CoE is only as it good as its people. When assembling a team, the combinations of skills needed should be carefully considered in order to optimise the performance of the CoE. The concept of the unicorn comes to mind! When I talk of a unicorn in the context of Data Science, I am referring to an individual who possess deep technical expertise in maths, computer science, statistics or engineering but also understands business issues and can communicate effectively. Unicorns are almost impossible to find, so there is no reason why these skills could not be distributed across the CoE. Ingraining diversity into the very fabric of your CoE and aligning the business users, story tellers, analysts and Data Scientists under one umbrella could create some devastatingly positive results. However, it is unwise to also extract all of the best talent into the CoE and great consideration must be given to the organisational make-up after its creation.

If you can’t measure the value, it does not exist

As I mentioned earlier in this article, the CoE is a great investment by any means. The continuation of the CoE is highly dependant on the ability to tie in the outcomes of projects to the value created for the business. Empowering a core team with decision rights and accountabilities for measuring the efficacy of data analytics projects can greatly improve the communication of wins to the business.

Support your initiatives with the right technology and infrastructure

The infrastructure of your business is paramount to the success of your CoE. Drawing in the expertise of the IT function to aid in your decision making when selecting vendors or building in-house data analytics solutions can help them develop the architectures needed. The right technological solutions can vastly improve the output of work by minimizing duplication and creating a layer of consistent data analytics across the enterprise.

Please feel free to share your experience in building a CoE in the comments section. If you would like to learn more about this topic and hear case studies from data analytics leaders, join Chief Data & Analytics Officer, Winter taking place in Miami from February 6-7th, 2018. Visit the website here.