Get started Bring yourself up to speed with our introductory content.

Give your predictive analytics projects a business reality check

Predictive modeling is great for optimizing operations to the current business environment, but a Harvard professor says its value diminishes in the face of disruptive innovation.

Predicting the future is always a dicey proposition, yet businesses increasingly are placing their chips on predictive analytics. What should they realistically hope to get out of their predictive analytics projects?

Speaking at software vendor SAS Institute Inc.'s 2014 Premier Business Leadership Series conference in Las Vegas, Harvard Business School Professor Clayton Christensen said it depends on what a company is trying to predict and the state of the industry it's in. If the business future looks similar to the present and past, an organization can use predictive models to score some big business wins, he said. But Christensen added that when major disruption occurs in a particular industry, predictive analytics techniques can become useless. For example, he said, computer vendors that specialized in minicomputers in the 1970s and 1980s couldn't have accurately predicted the sales impact of PCs and commodity servers because those systems were new innovations and there was no data about them available to analyze.

In this edition of the Talking Data podcast, Tech Target editors Ed Burns and Jack Vaughan discuss Christensen's points on the business relevance of predictive analytics projects and how companies should incorporate predictive modeling into their decision-making processes -- or whether they shouldn't bother if the required investments are unlikely to pay off.

Burns and Vaughan also explore the thoughts of SAS conference attendees on how the need for data analytics is changing the face of executive boards at many businesses by adding the chief analytics officer to the table. This move reflects a realization on the part of some businesses that they have things like their data warehouse and data governance policies figured out and now they need to focus more on how to use these things.

Ed Burns is site editor of SearchBusinessAnalytics. Email him at and follow him on Twitter: @EdBurnsTT.

Next Steps

Learn how predictive analytics projects can work in healthcare

See what Gartner has to say about the latest predictive analytics tools

Test your knowledge on creating effective predictive models

Dig Deeper on Predictive analytics

Join the conversation

1 comment

Send me notifications when other members comment.

Please create a username to comment.

It's a good point - if your models are based on the current situation, and the current situation changes drastically, then your models aren't much good anymore. Still, that's why they call it disruption - you don't know it's coming. You're always taking a risk by investing in planning and prediction, but there's also a risk in not doing that.