A voice is how Extrovert writes as you. It holds your background, how you want to sound, and the material the AI is allowed to bring up. Every comment and DM is drafted with one.

Each campaign uses exactly one voice, and a campaign can't exist without one. That's what lets you write differently for different audiences: longer takes in a topics campaign, short friendly comments for your creator peers, a low-pressure tone for reactivation.

Open any voice and you get three tabs:

What to actually write in each field, with character limits and examples, is covered in How to improve comment and DM draft quality. This article covers the voices themselves.

Go to Settings → Voice library. Each row is one voice and shows:

Click a row to open the voice. The menu at the end of the row holds Duplicate voice and Delete voice.

If you're an admin or a manager, the picker at the top of the page lets you switch to a teammate's or a managed account's voices and edit those too. Everyone else sees only their own.

The Comments and DMs columns tell you whether each channel is ready to write with. A column reading Not set up means the tone instructions for that channel are empty, and drafts there fall back to a generic voice.

One number is worth knowing: three. With fewer than three approved examples, or fewer than three DM templates, the AI has too little to pattern-match against, and the column keeps asking for more. Background isn't part of that check, but fill it anyway - it feeds both channels.

Click Add voice in the Voice library and pick one of three routes:

The same three routes appear inside each tab, in the Set up with AI card, so you can fill one section at a time without redoing the rest. Either way, review what the AI produced before you save it: cut anything generic, repeated, or unlike you.

Give the voice a name you'll recognise in a dropdown. "Casual outreach", "Enterprise CFOs", "Reactivation" beat "Voice 2".

The count is per person, so each team member has their own allowance. Inside a single voice, every plan gets the same room: 20 approved examples, 40 talking points, 40 selling points, and 40 DM templates. See Plan features and differences.

You pick the voice when you create the campaign, on the last step. The dropdown shows the voices belonging to the campaign owner, not to you, and it's prefilled with the one they edited most recently.

To change it later, open Campaign → Voice and pick a different one. The page also shows what that voice currently contains, so you can see at a glance whether anything is missing.

Two consequences worth knowing:

When a second audience needs most of the same setup, don't start over. Open the row menu in the Voice library and choose Duplicate voice. You get a copy named "Original Copy" - rename it, then change only what differs, usually the tone instructions and a few talking points.

This is also the safe way to try a different tone: duplicate, point one campaign at the copy, and compare the drafts before you change anything on the original.

Open the row menu and choose Delete voice. You can't delete a voice while a campaign is still using it - point that campaign at a different voice first, or delete the campaign.

Deleting removes the voice's examples, talking points, selling points, and templates with it. Comments and DMs already posted are unaffected.

One voice, filled properly, beats three half-empty ones. Fill Background first, then the Comments tab, and add the DMs tab when you actually start sending DMs. Add a second voice only when you catch yourself wanting to write differently for a specific audience.

For what goes in each field, read How to improve comment and DM draft quality.