← Resources

How to Make ChatGPT Sound Like Your CEO

The limit usually isn't the prompt. It's what the model knows about the person.

You can spend an hour improving your prompt. Sometimes that helps. But if ChatGPT doesn't know what your CEO actually believes, better prompting can only take you so far.

The problem usually isn't:

"How do I tell AI to sound more human?"

It's:

"What information does AI need in order to sound like this particular human?"

Why "write this in my CEO's voice" doesn't work very well

AI models are extremely good at recognizing patterns. Give one a few LinkedIn posts and it can often approximate:

That is style imitation. Executive voice is deeper.

The model also needs to understand:

Without that information, the model fills in the blanks. And the blanks are where generic thought leadership comes from.

Give ChatGPT a source of truth

Instead of rebuilding the executive every time you open a new conversation, create a reusable document containing the context the model needs. That can include:

Background

Who is this person? What experience informs their perspective?

Beliefs

What ideas do they hold strongly?

Point of view

Where do they have something distinctive to say?

Stories

What experiences can support those ideas?

Voice

How do they naturally speak and write?

Guardrails

What language, opinions, or framing would feel wrong?

Audience

Who are they trying to influence?

Then prompt from the source

A strong workflow looks like:

Executive → source material → voice brief → ChatGPT → draft

rather than:

Executive → vague prompt → ChatGPT → endless revisions

The model hasn't changed. The inputs have.

A useful test

Generate a piece without giving the model the executive context. Then generate the same assignment after providing a strong Executive Voice Brief. Ask:

If the second version isn't materially better, the Brief needs more work.

Studio doesn't replace ChatGPT

We think you should use it. And Claude. And whatever comes next. Studio exists one layer above the writing model. It captures the information those tools need before you ask them to produce something.

The model isn't the source of your voice. You are.