Who's in the Driver's Seat for Business Analytics Value?
I recently wrote an article concerning the need to extract business value from Business Analytics. A reader then sent me the following question and comment: "Who should be charged with making sure this happens - is it the responsibility of the CIO, line-of-business heads, data scientist (where there is one)? I assume it should be somewhat collaborative but I would think it benefits to have a key leader to drive this." This is an insightful question that business and technology leaders should examine.
Yes, the drive toward accountability is definitely a collaborative effort. It's not unusual to find a difference of opinion as to who in the organization should own responsibility for deriving business value from the BA investments and activities. But most will agree that someone high enough in the organization must be charged with this responsibility in order to bring exposure, profile and accountability to the effort. As a former CIO, I believe this should not be the CIO. Rather, the CIO should partner with the head of actuarial, the CMO, the head of new business, or other function as appropriate heads to lead the charge.
To some extent, just having someone drive the quest to extract value from the business analytics activities will help drive business value. As Dr. W. Edwards Deming taught, you will never be able to improve that which you do not measure. You have to analyze what you're analyzing. Having detailed analytics on a company's analytics investment and activity may seem strange at first, but measurement and process analysis are key to refining business analytics to improve use and drive business value.
Companies are beginning to look at their business analytics activity on a process maturity continuum. When considering process maturity for business analytics, the processes for the three key analytics practices should be examined:
- Measure: the processes to establish, report, and utilize metrics
- Extract business value: the processes to transition business questions to analytics questions, prioritize analytics activity, execute analytics requests and confirm business value
- Support: the processes associated with how a company or department supports the infrastructure and the care and feeding of the analytics organization
It is critical to assess each practice area to determine if the organization has the processes and structure in place to consistently deliver measurable business value. Using a process maturity assessment model, take each process and decompose the associated sub-processes to define how the requisite business value is delivered. For example, using a decomposition process analysis to extract business value, we would begin by translating business questions into analytics questions, prioritizing analytics activity, executing analytics requests and confirming business value. Using "confirming business value" to illustrate the sub-processes, the steps would include:
- Confirming the understanding of the analytics objective (do you have a purpose in the analytics request?)
- Communicating the business activity from the analytics (did you take action on the request?)
- Capture business results (did you quantify the value of the activity?)
In another scenario, when businesses look for improvement utilizing BI initiatives the emphasis is often on metrics reporting. Yet little time is spent creating good processes that are thought out to encompass establishing, reporting and utilizing metrics. If you were to decompose "establishing metrics" into sub-processes, the analysis steps would include: review and updating measures, establishing metric goals and linking metrics to incentives. Success in metrics is greatly increased as organizations mature these sub-processes and as their metrics processes mature, the benefit of BI initiatives is generally higher.
This analysis puts an organization on the path to capitalizing on analytics investments and moves it closer to becoming an analytics-based organization. The maturity scale ranges from an organization at Level 0 (incomplete processes) up to Level 5 (optimized processes). As with other examples of process maturity, an organization at Level 0 may have some success, however, it is hit or miss and often limited to an occasional situation where "things just worked out well." An organization at Level 5 has consistency in the process, ownership of the process, documentation and institutional knowledge of how and why things are done. These mature processes give an organization the means to achieve success.
Whose responsibility is it to make sure business analytics activities deliver business value? Ultimately, it is a shared responsibility. The CIO must be involved just as the business leaders must be. These leaders must set the expectation that the organization will be measured and held accountable for delivering business value.