I saw my first application of advanced mathematics to a strategic business problem in 1970. Since then, I've seen hundreds of such applications. In over 95 percent of the cases, those charged with making decisions didn't want to rely on the math, didn't understand the math, and stopped using the math within a few years. Ten years later, no one even knows that the math was ever used.
There's a second problem: A lot of the advanced math looked better than it was. Nice graphs suggested certainty where the numbers and assumptions shouldn't have permitted such impressions to be formed.
Beyond that, a lot of the data being used had no predictive value . . . a particular problem with correlation-based conclusions and time series.
Finally, the mathematicians often solved the wrong problem.
Have there been a few places where advanced math has made a lot of difference? Sure, especially where real time decision making would overload an organization. Load management in airlines, logistical optimization in supply chains, and in providing alerts that service is needed.
The most valuable applications that I've seen came in places where proprietary data added new perspectives that no one else could imagine. These advantages came from new ways of gathering data . . . not just compiling all transactions into large data bases. In fact, the best math solutions I've seen for strategy wouldn't strain any body's calculator to solve. Typically, these are done on personal computers anyway because the graphical choices are better for presenting what's been learned.
Can more advanced math be employed for strategy and operations? Sure. But the failure rate will be high, the cost will be enormous, and many managements won't engage.
People like Gary Loveman are unusual: Most executives don't appreciate and pay attention to analytics while running a large company. They prefer accounting reports instead. That's not going to change very fast except among start-ups by mathematically literate leaders.
What's really going to happen is that the off-the-shelf business intelligence software companies are going to make progress in selling their offerings to those who want and can use better data and analysis. But I suspect it will take another generation before you'll see much company-wide use of analytics.
You'll notice that I didn't discuss this book very much so far. Why? It doesn't reveal much of anything other than what you read in business periodicals and press releases by various vendors who want to sell offerings related to analytics. I recommend you skip the book. It won't tell you what you need to know. You would do better to spend a few hours with someone who understands analytics discussing what might be done to improve your performance.
I've read and appreciated a number of excellent books by Thomas H. Davenport in the past, so I'm surprised this book turned out to be so over optimistic based on so little evidence . . . and stated awareness of the problems. I can only conclude that this book is intended to sell services related to analytics rather than to give people an objective sense of what they are up against.
Ultimately, there's another problem with this book: If you use analytics to fine tune the current business model, you'll steal time, money, and effort from the more important task of creating an improved business model. The authors fail to make a distinction between business-model-optimizing analytics and analytics for business-model improvement. The former runs the risk of making companies less flexible and less able to compete.
The Balanced Scorecard approach, by comparison, is a healthier way to go by encouraging quantification of what needs to be done and tracking of how you are doing. From that discipline, you define the areas where innovation is needed . . . including analytics. Hiving off analytics as a separate subject simply creates the potential for misuse of a potentially valuable discipline.