Using predictive analytics to improve customer engagement
- November 8, 2015
- 332 views
Note: The following is sponsored content from Richardson, Texas-based Vertex.
In retail, as well as other industries where there is significant competition, predicting customer behavior through predictive analytics is no longer an option. It's become a critical must-have. Companies that have failed to adopt this new approach have gone out of business, while those that have embraced it have thrived.
Predictive analytics, when paired with smart customer engagement, can significantly improve performance. But how can utilities take advantage?
Dennis Rominger, marketing manager with Puget Sound Energy, and Micah DeHenau, vice president of Analytics and Consulting Services at Vertex, discussed that question and others during a recent Energy Central webinar titled, “Better Customer Engagement: Increasing Marketing Success with Predictive Analytics.”
The following is an edited excerpt of the webinar.
Q: How can predictive analytics be used to improve customer engagement at a utility?
DeHenau: One example is to better target offers. If my customer base is like a deck of cards, [predictive analytics] allows me to stack that deck in my favor. So I am achieving a higher return within a smaller portion of the population. By stacking the higher propensity customers toward the top of the population through predictive analytics, and scoring them on their likelihood to exhibit a certain behavior, I can then market to the customers most likely to enroll.
This is important because consumers are going to need to take a more active role in meeting energy demand in the near future. Leveraging predictive analytics to pair specific customers with the right energy efficiency programs is ensuring just that. Better targeting means a utility can spend less on outreach, yet achieve a much better response. For instance, analytics can help predict the likelihood customers will enroll in energy efficiency, e-billing or auto-pay programs – programs that both the customer and the utility find incredibly valuable. Predictive analytics allows you to identify your prime candidates for those programs and craft messaging that appeals to their specific values to drive increased enrollment.
Q: What are other ways predictive analytics can be used, in addition to improving marketing programs?
DeHenau: Utilities can also apply analytics to predict the likelihood of a customer to default on their debt. This can allow a utility to increase their debt-collection effort effectiveness by prioritizing collections resources and driving a higher return on collection-dollar investment.
In addition, in competitive utility markets where utilities are attempting to acquire or keep high-value customers, predictive analytics can identify those customers that are likely to leave as well as those likely to be acquired, which allows the utility to focus their retention and acquisition efforts on those specific high-value customers.
In addition to these customer-based applications, there are also a number of asset and equipment-based applications – such as predicting the likelihood of an asset to fail – and better prioritizing maintenance schedules.
Q: Can you point to specific examples?
DeHenau: We began working with Puget Sound Energy around four years ago to help them find a better way to deliver direct marketing for their energy-efficiency programs. The goal was to begin to pair customers up with PSE's energy-efficiency programs in a smarter way. PSE had nearly 150 different energy-efficiency products and programs then. The thought was that if we could identify which customers were likely to enroll in each one of these programs, we could significantly improve marketing effectiveness. Using predictive analytics, PSE marketing campaigns are now directed only at the high propensity customers, and they are designed to appeal to the values of those high propensity customers.
In the future, this will also allow PSE to conduct contact center upsells. So when an agent is discussing really anything with a customer, the agent is popped a message that says, “This customer is a high candidate for refrigerator recycling,” or any one of those 150 programs. So the agent can then actually add value to the conversation and the interaction, driving increased enrollments, and make customers happy as well.
Another example – the Metropolitan Sewer District of St. Louis has been able to increase its total arrears dollars collected by 30 percent since it began predicting which of its customers are most likely to pay back debt after going late on their bill. They have also been able to decrease the amount of accounts assigned to collection agencies each month by nearly 80 percent.
Q: What’s the best way for a utility to get started with predictive analytics?
DeHenau: It’s critical to begin the analytical journey by creating a strong reporting and business intelligence foundation. From this foundation, you can begin to reveal insight from data and drive change into the business. Progressive companies can then begin to anticipate events and predict customer behavior to really change the game.
To listen to the webinar, click here. This webcast was the third in a three-part series. Click here for the entire series.