Three minute video summary:
Consortium members met recently in Phoenix to explore “Predictive Customer Engagement,” or “Necessary Knowledge:” a topic suggested at the Leadership Committee Meeting in December 2013. How do we identify and deliver relevant information or knowledge to customers without them asking or searching for it? We set out to build the first draft of a framework for customer engagement.
We began by discussing the difference between unsolicited information that is valuable and helpful versus irritating and intrusive. We agreed that being able to deliver personalized, relevant, timely information in the context and best interest of the customer has great potential to reduce customer effort, increase customer loyalty, and be a competitive differentiator.
We heard from Louis Tetu at Coveo about what’s possible when you use unified search instead of federated search, which inspired a great conversation about the democratization of content.
Rick Hansen explained BMC’s AMIGO program, a great example of a low-tech proactive strategy that reduces both customer and agent effort when customers are installing upgrades.
Steve McMillan, our host from the University of Phoenix, described their goal of frictionless service, and their work on a system of automated triggers to identify and provide appropriate resources to students who need help, in some cases before that student even knows.
Jim Moran at Red Hat proposed that perhaps it’s more about TIMELY customer engagement than PREDICTIVE customer engagement, and talked through some ways that Red Hat is reaching out to customers based on activity, time frame, usage, and role.
Tony Nachman shared a pilot program that Sage is currently running. After observing higher NPS scores from customers who had opened one or more support cases, Sage is seeing if they can increase NPS scores and renewal rates by having support reach out to customers they haven’t heard from. They expect to have preliminary results soon.
During Open Space, the group discussed content implications of a predictive program, looked for non-survey methods to assess customer experience, and build the first draft of a model for customer engagement.
This first draft model reflects the idea that every interaction is an opportunity to improve the next interaction. It starts with what we know about people and known issues. The analysis of this information should yield patterns or events that are predictive. For example, if you downloaded this article, you might also be interested in these two articles. The recognition of a pattern triggers an interaction that will result in one of three outcomes: value realized, neutral, or value erosion. The outcome will also strengthen or erode the brand. The outcomes and what we learn from the interaction become additional data for analysis and should result in improvements in the business rules.