Strive for Impact, Not Just Insight

Bruce Kolodziej's picture
 By | June 22, 2015
in Rstat, Bruce Kolodziej, Predictive Analytics, predictive analytics, R, Rstat, RStat
June 22, 2015

Discovery of interesting information is always valuable and can lead to significant insight about an issue. Insight often occurs when organizations uncover facts about their customers, such as:

  • Customers who cancel accounts and leave for the competition
  • Customers under the age of 30 who have low historical spending habits
  • Customers in a specific income range who have called the call center three or more times in the last six months
  • Customers in certain zip codes, with accounts less than a year old, and negative sentiment scores

No doubt these insights are interesting and allow for deep understanding of customer behavior, but to gain maximum value we need to extend the insights into business impact. This means using the insights to predict a customer’s future behavior and also change processes or decisions.

At our recent Summit User Conference, dozens of our customers shared success stories on how they are impacting their business by monetizing data and operationalizing analytic results for better decisions. These stories ranged across virtually all industries, but two I want to specifically mention are student recruitment and retention in higher education, and enhancing data quality in healthcare to impact patient care.

Both use cases are excellent examples of moving from just uncovering insights in data to impacting the business by use of predictive analytics. The ability to generate revenue, reduce costs, and improve care in these cases illustrate how insights are often times not enough: real business impact is the goal.

Let’s revisit the use cases above and review how to turn insights about customer churn into business impact.

The insights we discovered about how a customer’s age, spending history, income, calls into the call center, geography, sentiment, and gender affect churn can be used to make predictions about future churn. We can assign a churn likelihood to each customer and make proactive outreaches to those with a high likelihood. This will allow us to reduce churn, save money, improve customer satisfaction, and maintain market share.

We gain insights when we analyze data holistically, by combining call center and social media data with traditional enterprise data of customer accounts, spend, and demographic details. But business impacts result when we link the repeated calls to the call center and negative sentiment to a business issue, such as churn.

Now we are able to see how these elements drive up costs, how these elements interact with other information sources, and also enable root cause analysis and process improvement. Now the organization can put corrective actions in place to reduce complaint calls and improve customer sentiment on social media outlets.

Want to learn more about the value of Predictive Analytics? Join us for a webcast on June 24 where I’ll be discussing the business value of predictive analytics, use cases across various verticals, deploying predictive results into operational applications and our advanced analytic platform, WebFOCUS RStat.