The point of departure for Analytics at Work is a domain space matrix generated by the cross-product of dimensions time frame and innovation. The values of time frame are past, present and future, while the levels of innovation are information (traditional BI) and insight (analytics). Analytics is thus concerned with the insight row of the 2x3 matrix, where insight past addresses how and why things happened; insight present asks what's the next best action; and insight future has to do with what's the best/worst that can happen
Analytics at Work coins a clever acronym, DELTA, to describe the capabilities and assets needed as a foundation for a successful analytics program. D stands for accessible, high-quality data; E represents an enterprise focus; L is enlightened analytic leadership; T represents strategic targets; and A depicts the analysts who power the initiative. Much of Part One of Analytics at Work takes a deeper dive into each of the DELTA factors in turn.
Of the five DELTAs, the most intriguing to me is T, strategic targets. The authors note that strategic targeting is all about finding analytics opportunities and setting the analytics ambition. I've found in 30 years as a BI consultant that many enterprises don't adequately address these preconditions critical to the success of their BI and analytics programs. Indeed, I'm often asked to help companies considering intelligence initiatives to determine the ROI of BI as part of the proposal process. And while I certainly can and do provide value in such cases, it's pretty scary if an organization is embarking on BI/analytics without thoughtful consideration of benefits and costs – – how the information and insight gleaned from BI will help the business increase revenue, reduce costs, make better decisions, etc. My big fear in such instances is that BI will become an IT initiative, only tangentially engaging the business units it's to serve. That's not a formula for long-term success.
Analytics at Work borrows from a seminal Harvard Business Review article, Putting the Service-Profit Chain to Work, as an important guide for discerning analytics opportunities. The article details the “theory” of how service firms operate, positing a series of strategic hypotheses of the form “the more of X, the less (more) of Y”, linked in a causal chain that articulates how the business works. In the case of the service company, the chain links from internal service quality, to employee satisfaction, to employee productivity, to external service value, to customer satisfaction, to customer loyalty to, ultimately, revenue and profitability.
Component hypotheses in this “theory of the service firm” might look something like the following: a greater internal service focus promotes employee satisfaction, leading to higher productivity. More productive employees, in turn, generate higher service value for customers. That value leads to enhanced satisfaction among customers, who then become more loyal, thus increasing company revenue and profitability. Outcome metrics early in the chain are called leading indicators, while revenue and profitability are lagging measures.
Strategy as theory can be a major enabler of the business search for analytics targets. Strategic assertions of how the business works are cast as causal linkage hypotheses, operationalizied with metrics, and tested with research methods (including experiments) and analytics. BI and analytics in this instance are front and center to the management of performance -- leading, hopefully, to better business outcomes through feedback loops. Finally, there's the collateral benefit of the collaboration of strategy and BI – aligning business and technology – to develop the “theory of the firm”.
Steve also blogs at miller.openbi.com.