Q2 2014 Intelligence Quarterly: Big data and the power of prediction

Intelligence Quarterly Q2 2014Business leaders have always made predictions about the future of their organizations. The difference today is that our predictions no longer have to be based on gut feel and inexact analyses of the past. With big data and predictive analytics, we have the ability to leverage collective knowledge and larger volumes of data. As a result, our predictions can be fact-based, not based on the experiences of one person.

Predictive analytics can be used in two powerful ways: for prevention or for creation. One is about stopping the undesirable from happening, and the other is about fulfilling desires.

First, let’s look at prevention. When banks can predict what leads to fraud, they can take steps to stop fraud before it happens. When public safety officials can predict what leads to crime, they can lower crime rates by curtailing the elements that lead to crime. When telcos predict the factors that lead to losing customers, they can step in to prevent churn before those factors align.

The clear advantage with prediction is that you are not merely reacting to fraud, crime or churn after the fact. You are taking action earlier to help reduce the factors that lead to fraud, crime and churn. You are preventing it from happening in the first place. I like to call this “predict to prevent.”

On the creation side, prediction can help you anticipate customer needs and fulfill those needs before demand strikes. Retailers can deliver products that customers want before they can even articulate the desire. Utility companies can anticipate spikes in energy use and produce the right amount of energy before demand increases.

More importantly, as economies shift from a product to a services focus, “predict to create” can give organizations an even bigger advantage.

Thinking back to the Industrial Revolution, consumers were suddenly able to purchase things they didn’t have before: cars, shoes, televisions and refrigerators are just a few examples. As consumer goods became produced on a mass scale, there were enough products for nearly everyone with the means to purchase them.

Now, in the digital revolution, the focus has moved from the product to the experience. Goods are still plentiful, but there’s a stronger demand for customer service and personalization. As a result, the feelings surrounding a brand can become even more important than the products. To compete in this new environment, companies are bundling products with services to create experiences, both online and off. Analyzing consumer and behavioral data has become one of the best ways to satisfy consumers, by determining not just what they want, but when they want it and how they want it – creating the complete package.

In this issue of Intelligence Quarterly, we’ve included multiple stories that illustrate how to use prediction for prevention and creation, including:

  • A hospital in Norway predicts what factors lead to patient injuries and prevent accidents and adverse reactions from occurring, resulting in huge improvements in patient safety (Page 3).
  • Public safety programs in the UK are analyzing public sources of data to predict and prevent terrorism, cybercrime and gun violence (Page 16).
  • A mobile marketing company predicts consumer preferences by analyzing location data and mobile activity, and creates relevant offers for registered users based on their preferences and whereabouts (Page 19).

With advanced analytics and the predictive capabilities of SAS, you can accomplish similar goals. Open the covers of this journal to learn how to use your data to prevent fraud, crime and churn – and to create product and service bundles just in time for demand to strike.

tags: big data, intelligence quarterly, predict to prevent, predictive analytics

Post a Comment

Your email is never published nor shared. Required fields are marked *

*
*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <p> <pre lang="" line="" escaped=""> <q cite=""> <strike> <strong>