Last week I reported on the first of two analytics-intriguing articles from the Fall 2011 Issue of the MIT Sloan Management Review. “Creating Business Value with Analytics” presents survey research findings that articulate factors distinguishing companies transforming the way they do business with analytics.

This week's article, “The Secrets to Managing Business Analytics Projects,” is a bit less strategic but perhaps more practical for companies and managers embarking on the analytics journey.

SMBAP starts where CBVA leaves off – with Aspirational, Experienced and Transformed analytics companies looking to enhance their capabilities. For analytics projects in these organizations, the research behind SMBAP seeks to identify “Which best practices do they employ, and how would they advise their less experienced peers.” A major goal of the inquiry is to highlight best practices in the structure and management of analytics projects.

The study methodology involves 32 in-depth, 1-to-2 hour, face-to-face interviews with business analytics project managers using an open-ended questionnaire. The sample includes experienced PMs only – an average of over 16 years. Interviewees are both internal and external managers drawn from a wealth of industries.

The main findings from the research involve common themes that surround both people and  planning/execution considerations of analytics projects. On the people side, analytics projects are enjoined to gain business commitment – to engage stakeholders early and often while managing involvement through all phases of the project. In addition, the research suggests all parties must recognize that analytics initiatives are similar to other internal business endeavors where true partnerships between major stakeholders yield the best results. Expectation management is also critical: “Setting the right expectations at the beginning and managing them as the project progresses increases both acceptance and the chances that the project will be successful.”

Study results indicate it's especially important to promote the smart use of information technology in analytics initiatives. Leading companies leverage IT to “revolutionize the way in which they innovate by playing on four dimensions simultaneously: measurement, experimentation, sharing and replication.”  Paradoxically, these dimensions promote decentralized IT to develop solutions, and centralized IT to effectively implement them across the enterprise.

If the authors are big in support of aligning the stakeholders of business analytics, the same cannot be said for deep, top-down strategy/project planning processes. Indeed, heavy planning seems now to be non grata in much of business thought, replaced by lighter efforts that “start with the premise that the initial plan will have to change as the project progresses. This is what we mean by a bias toward execution”  

The research contrasts production-oriented and specifications-based planning that details early requirements specification with minimal ongoing change and exploration, with experimentation-based approaches that emphasize evolutionary design with significant ongoing learning and change.

According to respondents, the latter, more adaptive approach is better suited to analytics projects. “Adaptive methods assume there's a need to gather information as you go along. These methods typically emphasize rapid delivery of prototypes … Many project managers have learned through experience that they can't expect to be right the first time ... it is better to attempt to execute good ideas quickly than to attempt to impose the 'perfect' plan.”

Part and parcel of the more agile management approach for analytics projects is the commitment to intelligent experimentation to fuel learning. Study respondents are advocates of “good experimentation” consistent with the scientific method that articulates “measurable hypotheses about the expected outcomes and controlled testing of these hypotheses.” They take a more pragmatic approach to experimentation, recognizing that laboratory style science is generally impractical.

After I finished reading SMBAP, I reviewed a blog I wrote a few years ago, “Planners, Searchers and BI.” SMBAP's emphasis on adaptivity driven by learn-as-you-go experimentation in contrast to heavy-handed, top-down planning reminds me an awful lot of the differences between Searchers and Planners. “Planners are top-down experts; Searchers are bottom-up experimenters. For Planners, the plan with fixed objectives is too often the end game; Searchers, in contrast, are not unduly attached to their ideas, but instead are flexible and nimble, willing to vary objectives with learning.”