|
Page 1 of 2
By Ellen Pearlman
Strategic Thinkers:
Thomas Davenport and Jeanne Harris
Credentials:
Davenport is the President's Distinguished Professor of Information Technology at Babson College; Harris is executive research fellow and director of research for the Accenture Institute for High Performance Business. They are the authors of Competing on Analytics: The New Science of Winning
Big Idea:
Prediction and recommendation technologies are proliferating, but they are not a substitute for decision-making.
Article:
"What People Want (and How to Predict It)" published by MIT Sloan Management Review, Winter 2009
Why is it that some people seem to have an uncanny ability to make smart decisions? We all know people like that who seem to lead a charmed life: they make the right choice at the right time and reap the rewards. But what about the rest of us - aren't there some useful tools to help make difficult decisions easier? There are, but how good are they?
In the current issue of MIT Sloan Management Review, Babson College Professor and author Thomas Davenport and Accenture Institute for High Performance researcher Jeanne Harris take a look at prediction and recommendation technologies to determine whether they are helpful in forecasting what customers want. They focus much of their article on consumer taste and whether it's a science or an art to figure out what the next great movie hit, gold record or best-selling toy will be.
To come to their conclusions, the authors reviewed the academic research on recommendation engines and conducted interviews with 18 organizations involved in predicting or recommending cultural products. They also interviewed movie studio executives, distributors and theatre companies about their use of predictive models. And while they focused on cultural products they note that any consumer product company will increasingly rely on these technologies to help guide them in making the right decisions. "No company will launch any expensive-to-create product or content offering without subjecting it to some form of systematic prediction or test," they say.
Traditionally, creators and distributors of cultural products have not used analytics to predict success. Instead they have relied on the instincts of tastemakers to determine what people will spend their money on. Today "the balance between art and science is shifting," the authors say, with the access to data and sophisticated technology providing options that didn't exist even a few years ago.
Recommendation engines are becoming popular with consumers because the number of choices they have to make is overwhelming. Producers too need to make smart decisions at a time when there are a plethora of products to buy and consumers are reluctant to open their wallets. In addition, the cost of producing many products has skyrocketed and developers can't afford to make expensive mistakes.
Next: Fine Tuning Recommendation Engines
|
