How to build an Analytic Center of Excellence

0

'Develop an Analytic Center of Excellence (COE)' could be translated to – develop a culture of analytics in your organization such that every department and division sees value in analytics and pursues measurement for all strategic decisions.

Chuck Kincaid said that there are two reasons that he proposed his paper for NESUG 2012. “One of them is my own life…,” he said. “I used to think I was a statistician, but now, with all of these fancy terms that are coming around, I’ve found that I’m actually a Data Scientist or an ‘Analytician.’”

His second reason was because analytics is important. He says that is removed from software, technology, hardware and SAS; it’s more of the people side. According to Kincaid, it’s about convincing the people in your organization to rely on data.

Kincaid said that there are three things that lead to analytics excellence: people, processes and support.

People

“The most important contributor to analytics excellence is having the right people,” he said. “When you look for people who are Analyticians, there are two kinds of things you are looking for: hard skills and soft skills.

Hard skills

  • Statistics – A statistical background and education is nice, but don’t assume that you have to look for people with graduate degrees or Ph.Ds in statistics. Look at people with physics, economics or operations research degrees. And, Analyticians can be brought up from within the organization and trained.
  • Database programming – This isn’t limited to knowing PROC SQL because people are now accessing data more directly with fewer layers between the data users and the data.
  • Programming – Today’s analyticians need to have a broader skill set than SAS programming alone. Look for multiple skills such as programming in Perl, PHP, Python and Java.
  • Business knowledge - You have to understand the industry and organization’s issues and translate those into solutions.

 

Soft skills

  • Problem solving
  • Collaboration
  • Presentation skills

Finding an Analytician

Remember that people with a quantitative mindset can come from many different sources. He says that it’s easy to determine if the applicant has the hard skills you are looking phone before you have them come in for an interview, but the soft skills require a different, more involved interview.

At Experis, interviewees are given a problem-solving scenario to that is ambiguous – this mimics problems clients have. “The customer always lies. They don’t know what they want, but they say they do,” said Kincaid, quoting a former colleague. The problem-solving challenge is more about listening to the problem, probing for requirements and presenting options.

“We also have every candidate give a presentation – about 20 minutes – that says ‘Here who I am and what I can bring to this company,’” said Kincaid. The presentation shows whether or not the interviewee can sell themselves – “If they can’t sell themselves, then they can’t sell other things.”

Operations

“This is a picture of a woman standing on a slab of bacon with a big spatula in her hand,” Kincaid explained, while showing a picture on the screen. The point is, if you don’t have operations, you may be doing this.

An important part of your operations is internal marketing. According to Kincaid, you always have to be selling the COE to people and departments in the organization. Tell them what you can do for them and what they can expect. Use pamphlets, videos, webinars and training to train people in your organization about the importance of analytics and how it can get their department and the organization closer to its goals.

Processes

“Project management and software engineering are two sets of processes that we can to that make what we do more robust, more valuable to our customers – and a better experience for them,” said Kincaid.

Requirements gathering, change management and version control make your processes reliable and repeatable and help sell your services. Also give a good estimate of when the project will be done. “Your customers have deadlines that you have to work within so that they will come back to you,” he explained.

Support

To provide the best support, you have to be selective in the projects that you take on. It’s important to be able to work on projects that contribute to the bottom line rather than spending time on EVERYTHING. To be a valuable COE - valuable to the organization - it’s important to measure what you do, so that you can know what types of projects contribute to the bottom line.

Kincaid said that there may be times that you take on less strategic work. For instance, you might want to introduce your team to departments that might not readily see the possibilities with analytics. An alternative to taking on less strategic may be to give the project to a less mature analyst to build additional skills or build a custom product for manual or repeated projects so that the end-user can be autonomously.

A connected Center

You may not need all of these roles in the beginning, but some roles that Kincaid recommends for your COE include statisticians, database programmers, programmers, application developers, business analysts, project managers, a PR specialist, an IT liaison, and a trainer.

The COE should be centralized – a place where these people are connected. “You might send somebody out to a group – like [the marketing group]for a year - but they are always tied to the center,” said Kincaid. According to Kincard, a decentralized COE makes it more difficult for the organization to grow in analytics development and use.

If you can’t find people, there are a few things that you can do: automate projects that are long manual or less strategic. Use recruiters to search for talent, offer off-site work, hire part-time people or hire contractors.

Reaching the next level

“You have to have the right talent. You have to have the right skills – talent is the key. Bring up talent from inside and give them a centralized structure,” said Kincaid. "Build repeatable processes – this builds the sense that you are reliable.”

 

Download Kincaid's paper and search for other great papers from NESUG 2012.

 

Share

About Author

Waynette Tubbs

Editor, Marketing Editorial

Waynette Tubbs is a seasoned technology journalist specializing in interviewing and writing about how leaders leverage advanced and emerging analytical technologies to transform their B2B and B2C organizations. In her current role, she works closely with global marketing organizations to generate content about artificial intelligence (AI), generative AI, intelligent automation, cybersecurity, data management, and marketing automation. Waynette has a master’s degree in journalism and mass communications from UNC Chapel Hill.

Leave A Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Back to Top