Using Predictive Analytics to Uncover Root Causes and Solve Problems vs. Treat Symptoms

Introduction

Despite significant investments in technology and an abundance of data, many companies repeatedly struggle to improve results on their marketing efforts. Why? In many cases, they are addressing the symptoms rather than the causes of their customers' issues.

What is a symptom versus a cause? Let's take the example of someone who visits the doctor for a persistent headache. If the doctor just prescribes aspirin, he is treating the symptom. If he does additional research and uncovers a more serious condition, then he would apply a different tactic to treat the cause of the headache. If the root cause is not diagnosed, a more serious condition could go untreated.

Predictive analytics can be utilized to uncover root cause issues for companies so they can treat problems, not just symptoms. For example, one company compared its marketing communications channels in order to measure the effectiveness of its offers. It discovered that one channel was experiencing higher attrition rates than others. Statistical behavioral analysis helped reveal that the sales force was selling the lowest price offering to obtain the highest response rate. Further analysis revealed that the initial price was lower only if the customer's usage was low. Heavy-usage customers incurred additional charges, which resulted in them paying significantly more than they expected.

This was the root cause of the increase in churn rate for this channel. The churn was merely a symptom of customer dissatisfaction over the pricing mismatch. By uncovering the real cause of the high churn rates, the company then was able to create training and incentive programs for the sales force on how to match pricing levels with customer needs. In treating the root cause, the company successfully eliminated all the systemic symptoms and eradicated the problem.