Client Use Cases: 1-800 CONTACTS Personalizes Actions to Curb Churn
Updated: Sep 30, 2019
The following is the final post in a five-part series on how 1-800 CONTACTS uses Compellon20|20 to drive change within their organization. Thank you to Neil Wieloch, PhD, Director, Marketing Strategy and Insights, for writing these posts and for sharing 1-800 CONTACTS’ successes with us. In case you missed them, the first post in this series was about identifying the drivers of churn, the second one was about creating predictive variables, the third one was about unifying Marketing silos, and the fourth was about flagging customers with look-alike data.
Who Do You Love?
Do all of your new customers need the same level and type of attention to become loyal? Probably not. But how do you know who needs what?
These were questions that the Insights Team at 1-800 CONTACTS pondered. Their hypothesis was that “who” someone was (demographically and psychographically) probably determined the level and type of attention required. To test this hypothesis, they worked with their data warehouse team to append customer transactional behavior data with both demographic and psychographic data. They then ran this data set through Compellon.
The Team was correct: customer re-order was predicted by individual customer demographic and psychographic variables. To make this insight actionable, the Insights Team at 1-800 CONTACTS (based on recommendations from their Compellon consultants) used the “confusion matrix” from the Compellon predictive model to identify customers as members of one of four segments based on their likelihood to re-order:
True positive = likely to reorder
True negative = likely to not reorder
False positive = look like they’ll reorder, but don’t
False negative = look like they won’t reorder, but do
As a result, the Marketing Team at 1-800 CONTACTS is now using these four customer designations to serve up variations on four different types of actions that look something like this:
True positive = continue to provide same great service (no change)
True negative = test possible messages and offers
False positive = provide reminders close to their estimated re-order time
False negative = message frequently with high appreciation and value
Working with the team at Compellon and using Compellon20|20’s breakthrough prescriptive analytics capabilities has changed how 1-800 CONTACTS’ teams work together, uniting the doers and the thinkers and putting them all on the same page. Silos are united. Data are better utilized. And most importantly, data-driven decisions have resulted in successful business outcomes.