Onboarding concept explorations
January 2025 · Senior Product Designer
Overview
Reveri's onboarding was designed to educate users before their first hypnosis session, but it could take 15–20 minutes before they experienced the product itself. We wanted to see if that time could feel more like the beginning of a session with Dr. David Spiegel, and less like setup.
At the time, we were exploring how AI could become part of the Reveri experience. We already had a voice model trained on Dr. Spiegel's voice, which gave us room to experiment with more responsive, conversational interactions. I explored how that could translate into onboarding: bringing more of his presence and guidance into the experience while helping users reach value sooner.
We ultimately decided not to take the onboarding exploration further. Instead, it helped us narrow the problem and redirect our effort toward a smaller, more useful interaction: Reveri's first voice-responsive hypnosis session.
Concept 01
Video call
The first direction imagined onboarding as a one-on-one video call with Dr. Spiegel. His trained voice guided the experience while questions and simple tap interactions appeared throughout, making the setup feel closer to a real conversation than a traditional onboarding flow.
Concept 02
Voice-first beacon
For the second direction, I removed the video and introduced a soft visual beacon as a representation of Dr. Spiegel's voice. The experience became simpler and more voice-led, while still giving users the option to tap when speaking wasn't convenient.
Flexibility became friction
Engineering built an early version of the experience so we could test the interaction beyond static concepts. Once we tried it, the problems became much clearer: the conversation felt rigid, switching between voice and tap added complexity, and the overall experience required more effort than the onboarding it was meant to improve.
Supporting both interaction modes also introduced a lot of complexity behind the scenes without giving users a clear enough benefit. What looked flexible on paper didn't feel particularly simple or natural in use.
We decided not to take the concept further. For me, the biggest takeaway was that giving people more ways to interact doesn't automatically make a product easier to use. Sometimes the better design decision is to narrow the interaction, reduce what the system needs to handle, and focus on the moment where the technology actually adds something useful.
