Competing on Analytics

Competing on Analytics by Thomas Davenport and Jeanne Harris is an important book for business leaders.  The title explains it well: some companies are setting up analysis of information as a major competitive strategy.  If you are competing with them with gut-feel decisions, you're going to be in a world of hurt.

Simple example: you are insuring teenage drivers.  You price insurance for teenage boys higher than for teenage girls, based on either the gut feel of the managers, or based on an old study done years ago.  Your competitor regularly updates its loss estimates, and matches pricing to these up-do-date estimates.  One gender or another finds a better deal at your competitor.  That's the gender which you were overcharging.  Now you have fewer of these.  The other gender, however, finds your pricing attractive.  These are the kids you have been undercharging.  So you now have more of the drivers with losses in excess of premiums, and fewer of the drivers with losses lower than premiums.

Now let's expand the idea of identifying customers a bit.  Suppose that your most loyal customers come in without discount coupons or promotions.  They provide high profit margins.  Can you identify them?  Can you keep them happy?  Can you keep them coming in?

Or you've built a business based on fast service.  But there are times when you get more business than you serve quickly.  Can you identify those times, so that you either staff up or you alert your customers that they won't get the usual speed this time?

These are all cases in which analysis of data can help you beat the competition.  Here's the BIG WARNING:  even if you don't choose this strategy, you're in trouble if your competitors do.

Remember quality?  Total Quality Management and its variations were the fad sweeping business a decade and a half ago.  We call it a fad, but the fact is that quality of products and services has improved tremendously since then.  Analytics is the same type of thing.  (In fact, the book reads like an old quality book in which the word processor's search-and-replace function was used to make it an analytics book.  However, even if the subject sounds a bit hackneyed on the first reading, it's still vitally important.)

Davenport and Harris talk about the roles of executives, of analysts, and of "analytical amateurs."  They emphasize that the top management do not need to be analysts, but they must understand analysis and encourage fact-based decision-making throughout the organization.  The company must also recruit non-analysts who are comfortable working in an analytical organization.  The analytical amateurs are not rocket scientists, but can use the results from the propeller-heads to make better decisions.

Competing on Analytics is not a book about analytic techniques.  They describe some briefly.  Regarding the ones I'm familiar with, I nodded my head.  Regarding the ones that were new to me, I found myself wanting to see more details, maybe some equations.  Perhaps that's just me.  After all, one doesn't go to graduate school in economics just to pick up chicks.  The book is vital to any corporate executive's shelf.