Having a Voice on Android: How hard can it be?
summary
Mobile voice features succeed or fail on concrete details: use reliable speech‑to‑text and pleasing TTS, show what users can say and offer multiple input modes, handle fallbacks and errors (Smart Life asks to “please add one” instead of explaining a missing device), leverage APIs and intents (Wolfram, Algolia, implicit intents) and installed apps, prefer graceful natural‑language parsing (Lyra handles variations), measure 71% voice preference, test with recorded audio or hardware playback rigs (169 Labs approach) and build predictable, recoverable flows.
AI-written summary · about
chapters (13)
- 0:00Introduction and speaker background
- 1:02Voice UX on mobile today
- 3:52Intentional design for voice interfaces
- 6:38First disappointments in voice apps
- 10:01User persistence and success moments
- 14:39Error handling and user expectations
- 17:28Babel and conversation flow
- 19:14Fallbacks and ecosystem leverage
- 22:23Common UX principles for voice apps
- 25:36Making users feel in control
- 27:21Delighting users with voice
- 29:52Summary and future of voice UIs
- 31:17Q&A on voice testing
related talks
47mBuilding Voice-First Android Apps
42mThe State of Voice Interactions on Android- 47mBuilding Voice-First Android Apps
43mThe State of Voice Interactions on Android
16mHow to voicify your Android app
40mConversational power for Android apps
36mVoice Enabling Mobile Apps and Custom Hardware Devices
29mAndroid Gets a Voice: Building Real-Time Agents