May 28, 2014
by Anand Rao
Where should the Chief Data Science Officer (CDSO) reside inside the enterprise? The short answer: it depends.
Companies that are deadly serious about using analytics to transform their industry draw a direct line from the CDSO to the CEO. This is the ideal arrangement, but it’s rare. We predict the CDSO will rise up the ranks as businesses gain a greater appreciation for the CDSO’s power. For now, most companies follow one of two reporting paradigms, which have their benefits and drawbacks:
1) Center of Excellence or Shared Service Model
The CEO and senior management establish a ‘center of excellence’ for data science that will assist other functional areas or business units to realize the full potential of data science. In this model, the CDSO reports either to the CIO or the CFO, the two executives who often bring together business units or functional units. By riding on the wings of the CIO or the CFO, the CDSO can gain a bird’s eye view of the entire organization and disseminate business intelligence across the enterprise.
With the CIO/CDSO reporting structure, there are two distinct disadvantages. First, a more centralized function might be too shallow to enable the CDSO to dredge deep enough into the problems, opportunities and data to surface solutions that make a meaningful difference. Moreover, the focus of the Data Science Office might end up being too technology focused and not business driven, which could discredit the initiative.
The second choice is for the CDSO to report to the CFO. Concerned with the performance of different business and functional units, CFOs can be the natural owners of the data science functions. Like the CIO-CDSO reporting model, the CDSO will have access to the entire organization. However, the CFO will view data through the lens of metrics to drive down costs and increase the efficiency of their operations. The drawback of this reporting relationship is that it’s more likely to lead to incremental versus transformational initiatives.
2) Champion Sponsor Model
In this model, the most analytically oriented department is chosen to house the CDSO. Yet, when residing in one department, the CDSO can be co-opted and colored by that department to the detriment of other departments.
For example, in the insurance industry, the marketing and actuary departments are both data-driven. However, marketing is more externally focused, creative and shaped by market and buyer demographics and psychographics. Actuaries are more inwardly focused, keen on historical data, and often require a high degree of certainty and quality of data before incorporating insights into their models. If marketing is chosen as the champion, marketing will likely hire a CDSO who will suit their needs and influence the CDSO to operate as they would. The CDSO will then need to ensure that the needs of the actuaries are sufficiently addressed. The same holds true if the CDSO reports to the Chief Actuary: the CDSO will need to additionally support the needs of the marketing department. The champion should hire a CDSO who can meet the needs of all departments.
Where you place the CDSO depends on the culture of your organization, the industry in which it resides, and the people who are heading the various business departments, but most importantly, it depends on what you want to accomplish with data and analytics. Do you need a center of excellence model that embeds data science into all aspects of the business or are you more suited for a champion model? Do you want data science to drive major advancements and disruptive change or will you be satisfied with incremental improvements? Consider these questions as you find a home for your CDSO.