Optum, Voice Design and Testing for Disease Management
Since a vast majority of Optum’s programs strongly utilize voice experiences, such as voicemail messages, over the phone scripted conversations, IVR interactions, etc. we had the opportunity to examine the end-to-end voice experiences for some of our most impactful and scaled disease management programs. These programs assist members over months and years with their health related to chronic conditions such as Diabetes, COPD, Heart Failure, Asthma, and Coronary Artery Disease.
Our challenge was tri-fold: increase the consumer enrollment, engagement, and overall experience, while improving the business value provided by these programs, and reducing operational costs. This initiative required deep understanding of the consumer, operations, technology, and business metrics used throughout. Furthermore, it was an exciting challenge to design purely for Voice Interactions. Voice Design is an undervalued type of experience in terms of creativity, design, natural language experience, and impact for the consumer.
- Increased client savings by $2 million
- Doubled consumer enrollment and interest through scripting re-design
- Instituted A/B testing methods across voicemail, agent scripting, and Outbound IVR
- Established design principles for voice engagement
- Developed roadmap for continuous improvement of Outbound IVR technology
- Scaled best practices across sister programs
First, we needed to understand all of the consumer touch points that impacted the consumer experience of our disease management programs. This required a thorough analysis of transparent and non-transparent (such as voicemail messages) consumer touch points. Once we had these mapped, we were able to dig into the analytics and flow of consumers to understand most significant areas of impact and opportunity through redesign.
Second, we held creative working sessions with agents, product, consumers and other key stakeholders, as well as conducted industry research, to develop new ideas to be tested across voice interaction touch points. We developed a backlog of ideas and began to prioritize changes to be implemented at strategic intervals. This enabled our teams to establish data and reporting mechanisms on selected variables for a scientific testing approach to making continuous improvements.
Third, we continued A/B testing of prioritized voice touch points and experiences to evolve the program design over time in a productive manner. Continuously prototyping of new scripts were essential to maintain both a creative and effective evolution of the service experience.
In the end, we have established not only improved programatic metrics, but also principles for voice engagement and testing capabilities to be employed across the rest of the business. With hundreds of other programs that utilize voice experience design, these practices will be valuable in scaling improvements to millions of consumers that interact with our health services. It is rewarding to know that our design efforts are improving the health of our consumers by instituting experiences that they find compelling, enticing, and engaging over time.