Predictive Analytics Helps Vodafone Ring Up Sales

Understanding the rapid and radical pace of change in today’s business environment is no simple task. Customer sentiment changes rapidly, cross-selling and up-selling services are essential and, for mobile phone service providers, churn rates wreak havoc with bottom-line results. As a result, there’s a fundamental need to understand the business and act on data in the most efficient way possible.

This concept rings true for Vodafone Netherlands (VN), the second largest mobile carrier in the Netherlands and part of the Vodafone Group Plc., which claims about 4.8 million subscribers. “We have a reasonably large number of customers, a limited marketing budget, and the need to understand how to apply the money effectively and get the best results,” says Viliah Overwater, senior information architect for BI Strategy & Architecture.

Vodafone recognized the need for an analytics-based approach that could help the company accurately forecast and predict customer behavior across consumer postpay, prepay and enterprise markets. The company wanted to avoid complicated software and the need to have teams of data scientists pore over all of the firm’s data and make sense of it. Instead, it needed an approach that allowed business users to draw results.

As a result, Vodafone turned to SAP InfiniteInsight, a predictive analytics solution. The company has used the analytics tool to build upward of 700 predictive models for churn and cross-selling.

“In the past, we had conducted some churn modeling, but we had used a manual tool that required a lot of oversight and intervention,” Overwater explains. However, the velocity of business in the digital age presented mounting challenges for Vodafone.

“We were seeing bottlenecks,” she adds, “and we faced difficulties executing on analytics and making sure that campaigns were taking place on schedule. We realized that if we wanted to be smart about campaigns, it [would be] necessary to embed models inside the campaign in an automated fashion.”

The current solution allows Vodafone to dissect data and gain insights that would have escaped the company in the past. For example, the firm recently wanted to better understand winter roaming patterns for subscribers and how better to communicate with them. This required a deeper understanding of customers’ activities.

“We wanted to understand who goes skiing, so we could target a campaign and provide a roaming service that fit their needs more closely,” Overwater says. “Once we understood who roams and why, we were able to more closely address customer needs.”

As big data takes hold, and organizations find themselves swimming in more data and more complex data sets, analytics tools are becoming increasingly essential, Overwater points out. “We want to make sure we are reaching the right people, but we also don’t want to over-contact them,” she says. “We want to find ways to help our customers save money.”

Predictive analytics has also helped Vodafone identify factors that may signal churn. “It’s not one thing; it’s a group of things and how they relate to each other,” Overwater notes. “We can drop hundreds or thousands of variables into the model and find out which are the most relevant for the final model. It fundamentally changes the way we run the business.”