How to Create a Voice Recipe (4 Steps to Sound Like You)

You use AI to write. So does everyone else. That is why everything sounds the same.

The problem is not that AI writes badly. It writes generically. It averages the internet and produces content that could belong to anyone. A voice recipe fixes that. It gives the model a set of rules so the output sounds like you, not like everyone else.

What is a voice recipe

A voice recipe is a structured set of constraints that shape how AI generates text. Think of it like a brief for a ghostwriter: here is how I talk, here is what I never say, here is the energy I bring to a room. Without that brief, the ghostwriter guesses. With it, they sound like you.

A voice recipe typically includes:

  • Your personality archetype (the person behind the brand)
  • Your audience energy (how your reader should feel)
  • Your anti-patterns (words, tones, and structures you never use)
  • Your voice anchors (real examples of how you actually communicate)

Step 1: Name your personality

Every brand has a personality, whether you have defined it or not. The question is whether you are controlling it or leaving it to chance.

Think about how you talk to your best customer in a relaxed conversation. Not your pitch voice. Not your LinkedIn voice. Your real voice. Are you direct and blunt? Warm and encouraging? Dry and understated? That is your starting point.

Pick 2 to 3 adjectives that describe how you naturally communicate. Not how you wish you sounded. How you actually sound when you are not trying.

Step 2: Define your audience energy

Voice is not just about how you talk. It is about how your reader feels after reading. Do they feel challenged? Reassured? Energized? Curious?

This matters because AI defaults to neutral. Without direction, it writes content that lands flat. When you define the emotional response you want, the model adjusts tone, pacing, and word choice to create that feeling.

Step 3: Identify your anti-patterns

This is the most underrated step. Knowing what you never say is more useful than knowing what you do say. Anti-patterns act as guardrails.

Common anti-patterns include:

  • Corporate jargon ("leverage," "synergy," "innovative solutions")
  • Exclamation marks and forced enthusiasm
  • Passive voice and hedging language
  • Buzzwords that mean nothing ("cutting-edge," "game-changing")
When you tell AI what to avoid, the output immediately sounds more human and more distinctive.

Step 4: Find your voice anchors

Voice anchors are real examples of your communication at its best. A LinkedIn post that got great engagement. An email that a customer replied to saying "this is exactly how I feel." A paragraph from your website that sounds unmistakably like you.

These examples give the AI a concrete reference point. Instead of generating from scratch, it calibrates against your actual voice.

Putting it together

A voice recipe is not a creative exercise. It is a strategic constraint. The more specific your recipe, the less generic the output. That is the trade-off most people miss: creativity comes from constraints, not from freedom.

servo's voice tool walks you through these four steps in about 5 minutes and produces a recipe you can paste into ChatGPT, Claude, Jasper, or any AI tool. The recipe works because it encodes decisions, not preferences.

Start your session →