Voice input removes the friction of typing on mobile. For mobile users — now the majority of web traffic — speaking a question is far more natural than tapping it out on a small keyboard.
The Three-Step Pipeline
1. Speech Capture: When the visitor taps the microphone icon, the browser's Web Speech API activates the device microphone. No plugin required — it is built into modern browsers. Audio is converted to text in real time using automatic speech recognition trained on billions of hours of speech.
2. Transcription Review: The spoken words appear in the chat input field. The visitor can review and edit before sending — catching any misheard words before they reach the chatbot.
3. Standard Processing: Once sent, the text enters the normal pipeline: embedding, vector search, retrieval, generation. Voice input only changes how the text arrives — processing is identical to typed input.
Language and Accuracy
The Web Speech API supports 70+ languages and inherits the user's browser language setting. Modern speech recognition achieves 95%+ accuracy for clear speech on typical customer queries.
Accessibility Benefits
Voice input helps visitors with motor impairments, elderly users who find small keyboards difficult, visually impaired users with assistive technology, and anyone multitasking or hands-free. Enabling it is a one-click toggle in Chatsloop settings.
Businesses whose customers skew mobile typically see a 15-25% increase in mobile engagement after enabling voice input.
