Using voice transcription in the Messenger
Voice transcription gives your customers a convenient way to communicate without typing. They can simply record a voice message in the Messenger, and it will be automatically converted to text for them to review, edit, and send.
This makes it easier for customers to provide detailed context, making conversations faster and more natural.
How it works for your customers
The process is simple and intuitive for customers using the Messenger on the web.
First, the customer clicks the microphone icon in the message composer. If it’s their first time, they’ll be prompted by their browser to grant microphone permissions.
While recording, they will see a waveform and have options to cancel or stop the recording.
After stopping the recording, the audio is processed for transcription.
Finally, the transcribed text appears in the composer. The customer can then edit the text if needed and send it just like a regular message.
What happens if something goes wrong?
There are a couple of scenarios where a transcription might not work as expected.
Microphone permissions
If a customer denies microphone access, the microphone icon will display a tooltip explaining that they need to grant permission in their browser settings to use the feature.
Transcription failures
In the rare case that a voice message cannot be transcribed, an inline error message will appear in the composer to let the customer know that the transcription failed.
How to enable or disable voice transcription
This setting allows you to control whether or not your customers can send voice transcriptions in the Messenger.
To manage this feature for your workspace:
From the main navigation menu, go to Workspace.
Select the Security page.
Under the security options, click Attachments > Messenger input types.
Find and check or uncheck the box labeled "Voice notes".
Important limitations
Note: There are a few limitations to be aware of with voice transcriptions:
Transcription quality: The accuracy of the transcription can be affected by factors like strong accents, background noise, or speaking clarity. Customers may need to edit the transcribed text before sending.
Network connection: The feature requires a stable internet connection to work in real time. A poor connection may cause a delay in the transcription process.
Length and size limits: Voice recordings cannot be longer than 2 minutes or larger than 10 MB.
iFrame implementation: For the feature to work in a sandboxed iFrame, you must include
allow: “microphone”in your iFrame code.
Need more help? Get support from our Community Forum
Find answers and get help from Intercom Support and Community Experts