Traditional voicemail was a destination: dial in, press through prompts, listen in order, forget the number halfway. Modern voicemail is a delivery: the message arrives in email seconds after recording, as an audio attachment with an automatic transcript, sender, time, and duration in the subject line. The storage moved; more importantly, the interaction model inverted, from you visiting messages to messages visiting you.
What actually happens to a message
The caller records; the system stores the audio, runs speech recognition over it, and sends the result wherever configured, typically email to the extension owner, often also into the mobile app. One message can fan out to several inboxes (a shared "info" box emailing three people), and boxes need not belong to humans: an after-hours menu option can be a mailbox whose transcripts land with the whole team by morning.
Honest expectations for transcription accuracy
Modern transcription is genuinely good on clear speech and struggles predictably: heavy accents, names, addresses read fast, bad speakerphones, wind. Treat the transcript as triage, who called, roughly why, is it urgent, and the audio as the record for details worth acting on. A phone number in a transcript deserves one listen before dialing. That triage function alone changes the workflow math: fifteen voicemails after a morning of meetings become a scannable list, urgent ones surfacing in seconds, without eleven minutes of sequential listening.
Workflow effects people notice
- Voicemail becomes searchable: "the guy about the roof, three weeks ago" is an inbox search, not an archaeology dig.
- Messages become forwardable: a transcript+audio email routes to whoever should own it, with context, in one forward.
- Coverage becomes visible: shared boxes emailing a team mean a missed message is everyone's to see, and no single person's to bury.
- The callback queue develops memory: unread email is a better to-do list than a blinking light ever was.
The retention question nobody decides on purpose
Emailed voicemail is stored voicemail: transcripts and audio persist in mailboxes under your email retention rules, which may be longer, or shorter, than you would choose for phone messages. For most businesses this is a quiet benefit (a paper trail of commitments); for some content it deserves a policy decision, aligned with how you treat recordings and transcripts generally. The point is to decide, rather than discover the answer during a dispute.
Configuration worth stealing: give every menu dead-end a mailbox that emails a real, monitored inbox with transcripts. "Leave a message" is only respectable if the message demonstrably reaches someone by morning.
Still have a question?
Real people answer our phones. Ask anything about business phone systems, no pitch attached.
Ask a human