A visitor’s preferred way to ask a question changes with their situation. Someone exploring a service on a phone may want to explain a complex need aloud. The same person in a shared office may prefer a short written message. Offering both voice and text gives the conversation room to fit that moment.
Parev AI brings the two channels into one website widget. The useful starting point is the visitor’s choice: how can they ask comfortably, understand the response and decide what to do next?
Where voice can help
Speaking can be practical when a question takes several sentences to explain or when typing on a small screen is awkward. A visitor might describe the problem they are trying to solve without searching for the exact name of a service. A useful assistant can then ask a relevant clarification and explain what matches that need.
Voice depends on the visitor’s environment and device. It needs a compatible browser and microphone permission. Background noise and network conditions can affect the experience. A clear invitation helps people understand that voice is available; it should not be confused with starting a microphone automatically.
Where text makes more sense
Text works well when someone wants to read quietly, ask a precise question or avoid speaking in public. It also lets visitors participate without giving microphone permission. A written explanation can include links to relevant pages, so someone can move between the conversation and a fuller service description.
Parev AI displays text answers progressively with an indication that the question is being processed. That visible progress helps the visitor understand what is happening. The actual wait still depends on factors including the connection, the question and the information needed for the answer.
Keep the conversation relevant across channels
The channel is one part of the experience. A visitor asking “what does it include?” also needs the assistant to understand the page they are viewing. Parev AI uses page context and relevant approved knowledge, alongside the conversation history, to make follow-up questions more meaningful.
Conversation context can continue within the same browser and website, including across voice and text sessions. This can reduce the need to describe the same situation again. Treat it as continuity within that context: it does not automatically identify someone using a different device.
Make the state understandable
Visitors need to know when the assistant is connecting, ready to listen, processing or speaking. Clear status text and visual cues make the interaction easier to follow. During setup, check whether people can tell when to speak and whether text remains easy to reach.
Review the widget on desktop and mobile. Its placement should work with the page’s other fixed elements and keep important actions accessible. Parev AI offers separate adjustments for the closed widget on these devices, including size, position, spacing and background colour.
Test the same need in both forms
Try a typical visitor need aloud and in writing. Review the substance of the answer, the clarification and the next step, rather than judging the experience only by its appearance. Check microphone permission, mobile placement and behaviour when a channel’s configured limit has been reached.
Then use actual conversation records to find improvements. Voice records are transcripts, so review them as text when evaluating explanations. Explore voice and text features and see the setup steps before deciding how both channels should work on your site.