Who Writes on This Account
Your AI agent writes every reply in the voice set on the agent: the persona, the tone and the rules in its business context. That is right when the agent runs on your own LinkedIn account. It is wrong the moment the same agent replies from a colleague’s account, because their leads would be reading your name, your way of writing and your claims about the company.
Who writes on this account fixes that per LinkedIn account. It lives on the account row in Settings, on the LinkedIn accounts tab, under the timezone.
What you set
- Name: the name the replies are written as. Use the first name people see on that LinkedIn profile.
- Refers to themselves as: she, he, or no preference. Some languages change word endings depending on who is speaking, and this keeps them right when the agent replies in Spanish, French, Dutch and other languages that do that.
- Who they are: one line on their place in the team, for example “part of the team, talks to agency founders about their outreach”. The agent uses this instead of the founder’s story, so a colleague never says they built the product.
- How they write: casing, punctuation, emoji, formality and length. This replaces the writing-style rules on the agent for this account only. Everything else the agent knows and does stays the same: the knowledge, the offer, what it never promises, the rules about calls.
Choose Save. From the next message onward, replies and follow-ups in that account’s conversations are written as this person.
What does not change
- Leave it empty and the account writes in the voice set on the agent, exactly as before. Your own account needs nothing here.
- The agent, its knowledge and its rules are shared. A persona changes who is speaking and how, not what they know.
- Conversations a person is handling by hand are not touched.
Remove persona takes it off again; the account goes back to the agent’s own voice.