Skip to Content
Business Trends
Author's profile photo Kristin Kufeldt

Predictive Thursdays: Recommending Next-Best Product Offer in Real Time

Today’s guest blogger is William Tangalos, Decision Scientist, CenturyLink

predcitve_analytics_video_screen-_capatureAt CenturyLink, we recently worked on a problem that many organizations have—helping inbound sales consultants present the best offer for the customer ‘in front’ of them. In too many cases, inbound sales consultants rely on their experience (which in some cases may only be a couple months or weeks) to cull through hundreds or thousands of potential product-bundle offers and then decide which specific product bundle to offer the specific caller.

Our approach is to develop and implement duo empirically-based analytics components in real time in order to enable automated continuous learning in a profitable manner. What our team is striving for is to use:

  1. individual consumer -level, predictive-product propensity models in conjunction with
  2. a “segment”-level based, real-time decision engine that optimizes specific product bundle offerings on a customer segment level

How Does the Solution Work?

This duo analytically-based solution culls through hundreds and thousands of potential product offer bundles in real time for each caller and provides a specific product offer bundle in one second to the sales consultants. The sales consultants then view on their sales screen the distinct product offer bundle that they can offer to the caller.

Optimally, the “results” of the offer are immediately conveyed back into the decision engine portion of the analytics, which adjusts the sales effectiveness of each specific offer to the specific customer segment which the caller is in.

The other component of the duo analytics capability is individual level product propensity scores, which rank orders the product purchase likelihood of hundreds of product bundle components for each individual customer. These scores were built quickly and deployed easily within the SAP HANA environment using the SAP Predictive Analytics Automated Analytics module.

The development and deployment team also found that the SAP Real Time Offer Management decisioning engine worked well with the SAP HANA environment. So that optimal sales rates can be achieved, we’re able to “adjust” the weight given to each of the duo analytics two capabilities (the individual product-propensity component and the real-time decisioning engine/segment component ).

Benefits of the Duo Analytics Solution

There are over 5 million individual customers who can call in, and the call centers have over 2,000 I/B calls each day. The organization is now able to realize efficiencies in this new I/B next-best-offer sales process.

Though more testing and greater organizational adoption is in process, preliminary results have shown that the sales per 100 inbound calls is materially superior using the duo analytics solution when compared to the prior business-as-usual sales process.

Future Uses

Though this initial deployment has been implemented in an inbound call center, the very same capabilities can be utilized in real time in store and on company web sites. But that capability, process, and related measurement is a blog subject for another day.

Learn More

  • Got questions about CenturyLink’s approach? Share you thoughts here.
  • Want to know more about SAP Predictive Analytics Automated Analytics  or SAP Real Time Offer Management ? Visit the SAP help portal.
  • Interested in predictive analytics in general? Read the rest of the blogs in the Predictive Thursdays series.

Assigned Tags

      Be the first to leave a comment
      You must be Logged on to comment or reply to a post.