Follow us...

Measuring the Success of a Business Analysis Center of Excellence

By Joy Beatty, Seilevel, Vice President of Blue Ocean Services

In implementing a Business Analysis Center of Excellence (CoE), it is important to measure the success of the endeavor, because these measurements keep your funding in place to continue the effort, and they motivate improvement within the BA organization. Further, if you aren’t trying to improve something in your organization, then you wouldn’t implement a BA CoE!

Management buy-in will literally make or break a BA CoE effort—simply put, if management does not buy-in to the importance of the CoE, when budgets get tight, they will cut this effort long before they cut an actual software project. Management must not only understand how the BA CoE is important to the organization’s success; there must be at least one management champion for the CoE. The easiest way to gain management support, and to find your champion, is to have metrics that demonstrate the need for a BA CoE and periodically show improvement on those metrics. Further, BAs know they need to do what they are measured on; if there are success metrics for a CoE in place, the BAs will behave accordingly to improve those metrics.

Types of Metrics for a BA CoE
The success of the BA organization should align with delivering overall business value on a project. The very best metric to measure the success of a BA organization is whether the business objectives for each project were met. That said, whether or not business objectives are met is difficult to measure; you may need to use other metrics to get good indicators of success.

Metrics by Project  Explanation
% Business Objectives met Track whether overall value is delivered as an indicator that BAs understand the Business Objectives and align scope accordingly
% User Adoption  Track that the percent of users who adopt the new solution is moving towards 100% as an indicator that BAs are working with the business to build something acceptable to the actual end users
Average missed requirements as a % of total requirements by phase Track when missed requirements are identified, looking for a reduction in later phases as an indicator of overall BA skills
Number of out of scope requirements added as % of total Track “scope creep” beyond the original scope as an indicator that BAs are tracking to business objectives
BA turnover Track that the number of BAs leaving the organization because they aren’t growing is improving as an indicator that the COE clarifies and supports BA career paths

When setting metrics to measure success of a CoE, understand the current baseline and set a reasonable target improvement. A goal that is too lofty will serve to demotivate the team rather than inspire improvements. It is also helpful to involve the BA team to set target metrics so they are more likely to help achieve them.

Don’t Forget to Actually Measure Your Metrics
Once the metrics are set, they also must be measured periodically (monthly or quarterly at minimum). If there isn’t any improvement in the metrics, then change the CoE approach accordingly. This is very much like running an iterative project: Implement, measure, iterate and implement more, measure, iterate, etc.

If you are not taking regular, periodic measurements, there is no way to know if the effort is on track for success. If progress and likelihood for success is unknown, and management has to cut a project for budget reasons, this effort is likely to be cut. Simply put, if your BA CoE doesn’t have the support needed in the organization, it will fail; these success measurements are an easy way to build that support.

Joy Beatty is Vice President of Blue Ocean Services at Seilevel, a professional services company focused exclusively on helping clients change the way they create requirements in order to enhance business outcomes. She writes about Seilevel methodologies and studies in whitepapers that can be found at http://www.seilevel.com/resources and blog posts that can be found at http://requirements.seilevel.com/blog.