Review of Tom Davenport’s "Competing on Analytics"

Competing on Analytics: The New Science of Winning, Davenport, Thomas H. and Jeanne G. Harris


Tom Davenport has turned his attention of late to the prospects for business intelligence and information analytics. Competing on Analytics offers a managerial introduction to the topic. It emphasizes why organizations ought to be interested in the topic, what kinds of payoffs they might expect, and how organizations will need to adapt to take advantage of robust analytics. Davenport and co-author Jeanne Harris of Accenture split the book into two major sections. The first deals with describing how analytics can be used as a competitive tool; the second with the organizational challenges of building analytical capabilities. Overall, it’s a relatively short book and is well-suited to its target audience. On the other hand, if you’re on the receiving end of a mandate to build an analytical capability after someone higher in the food chain has gotten excited about the topic, don’t expect quite as much in the way of detailed implementation advice.

Davenport and Harris set out a stage-model of analytical capabilities starting with "analytically impaired" and ending with "analytical competitors." Partly, this is to support an argument they make that there’s an advantage to managing analytics at an enterprise level. My cynical side suspects that this advantage lies primarily in providing a clear target for the likes of Accenture or SAS to sell to.

Given that every new capability benefits from senior executive attention and that everyone wants to get on the CEO’s calendar, are there, in fact, compelling reasons that analytics deserves to be on this short list? Two come to mind. One is that the expertise called for in effective analytics is scarce. Better to have that expertise directed at the targets of greatest opportunity by those best positioned to judge. Two, the competitive business opportunities that might yield to analytics are more likely to be found from the perspective of those with an integrative view of the enterprise.

The authors walk through major functions of the enterprise identifying opportunities and examples of how analytics have been successfully applied. There are clearly an abundance of opportunities to apply analytical tools and techniques to improving internal processes, optimizing supply chains, and leveraging marketing.

One problem with the focus on describing the business opportunities for analytics is that the variety of potentially applicable tools gets short shrift. All books have to make decisions about what to put in and what to leave out. Given the intended audience, I can understand the decision to focus on the business side of the equation rather than on the tools side. On the other hand, glossing over the complexities of the statistical tools and algorithms has its own risks. Organizations risk creating a new class of wizards whose dark arts must be taken on faith or they risk putting dangerous tools in the hands of amateurs who will be blind to both the limits and the dangers of the tools.

This brings us to the second part of the book and the challenges of building an analytical capability. I

Attitude, hypothesis, experiment, and evidence

Doing science is fundamentally a state of mind more than any particular set of tools or any particular domain of knowledge.

How do you know when you’re doing science wrong?



Read the comments on this post…


More in the same vein from xkcd.

Fostering these attitudes is increasingly relevant in organizational settings. We’re awash in data and in advocates of data mining, information analytics, super crunching, and other forms of extracting insight from the data. Too often, however, the emphasis elevates a new set of experts with a new set of mysterious tools saying “trust me.” Trusting them is no better than trusting your gut or someone else’s gut.

Fundamentally, the scientific method is no more than a method for how to be productively skeptical in the face of pressures and dispositions to believe and the multiple ways to be mistaken.