Most people treat AI like a search engine. They type a few words, get a generic answer, and walk away thinking AI "isn't that useful."

The problem isn't the AI. It's the prompt.

If you've ever asked AI for a LinkedIn post and gotten something that sounded like it was written by a robot for a robot — you know what I mean. The output is technically correct and completely unusable. You asked for a professional summary and got a paragraph that sounds like a job description from 1997.

AI is only as smart as the instructions you give it. Vague in, vague out. Give it real context, a clear role, and specific guardrails — the output changes completely.

That's the premise behind CRISPY™, a prompt framework I built to help sales professionals and business leaders get genuinely useful output from AI.

Here's how it works.

What Is the CRISPY™ Framework?

CRISPY™ is a six-part structure for writing AI prompts. It stands for:

  • C — Context
  • R — Role
  • I — Inspiration
  • S — Scope
  • P — Prohibitions
  • Y — You

Each section solves a specific problem with the way most people prompt. Together, they turn a vague request into a clear brief — the kind a talented human collaborator would actually be able to act on.

Here's how each piece works.

C — Context

What it is: The background the AI needs to do its job well.

Most prompts skip this entirely. The result is generic output because the AI is filling in blanks with guesses.

Context answers: Who is this for? What's the situation? What has already been tried or said? What does the audience already know?

Example of a prompt without context:

"Write a LinkedIn post about prospecting."

Example with context:

"I'm writing for B2B sales professionals who are already active on LinkedIn but struggle to get responses from their outreach. They've tried connection requests and generic messages and gotten nothing back. They're skeptical of AI content because it sounds fake."

Same topic. Completely different starting point for the AI.

R — Role

What it is: The persona, expertise, and perspective you're asking AI to inhabit.

This one matters more than most people realize. AI doesn't have a default point of view — it has millions of them. If you don't specify a role, it averages across all of them, which is why you get that vague, cover-all-your-bases tone that doesn't sound like anyone.

When you define a role, you anchor the AI's voice, vocabulary, and frame of reference.

Weak role: "You are a helpful assistant."

Strong role: "You are a B2B sales strategist with 20+ years of experience helping sellers build trust-based pipelines on LinkedIn. You write with a direct, conversational tone. You've seen every outreach mistake in the book and you're not here to sugarcoat them."

The second version tells the AI how to think, not just what to do.

I — Inspiration

What it is: Examples, tone references, or style guides the AI can learn from.

This is where most prompt writers leave significant quality on the table. Telling AI to "write in my voice" means nothing if you haven't shown it what your voice sounds like.

Inspiration can be:

  • A sample of your own previous writing
  • A few bullet points describing your tone ("punchy, direct, no buzzwords, use real examples")
  • A link or excerpt from content that hits the right register

If you're trying to create on-brand content consistently, this section is what makes the difference between "sounds like me" and "sounds like AI trying to sound like me."

S — Scope

What it is: The format, length, structure, and specific deliverables you want.

AI will fill whatever container you give it. If you don't give it one, it will make one up — and it will usually be longer, more structured, and more formal than you actually wanted.

Scope answers: How long should this be? What format — bullets, prose, a numbered list? What sections do I want? How should it start? What's the one thing I want the reader to walk away with?

Vague scope:

"Write me a blog post."

Clear scope:

"Write a 600-800 word blog post in three sections: an opening that names a common mistake, a middle section that explains the better approach, and a closing that ends with a question to the reader. No subheader for the opening paragraph."

The clearer your scope, the less editing you'll do on the back end.

P — Prohibitions

What it is: The things you explicitly don't want — content, tone, format, phrases.

This is the section people skip because it feels obvious. It isn't.

AI has defaults — patterns it falls back on when it's not told otherwise. If you don't name what you don't want, you'll get it. Bullet points when you wanted prose. Corporate-speak when you wanted conversational. A generic conclusion when you wanted a specific call to action.

Prohibitions give you control over those defaults.

Examples of useful prohibitions:

  • "Do not use the words 'leverage,' 'utilize,' or 'synergy.'"
  • "Do not start sentences with 'In today's world.'"
  • "Do not use bullet points — write in paragraph form."
  • "Do not summarize at the end. End on a forward-looking question."
  • "Do not make claims I can't back up with data."

The more you use AI, the better your prohibition list gets. Keep a running list of the things that keep showing up in your edits — those are your prohibitions.

Y — You

What it is: Who you are, what you're trying to accomplish, and any intake questions you want the AI to ask before it starts.

This is the last section — and that's intentional. By the time the AI reaches this part of the prompt, it has fully absorbed the context, role, inspiration, scope, and prohibitions. Now it has the full brief. Now it can ask smart questions.

The Y section is where you can turn a one-shot prompt into a template. Instead of always pre-filling every detail, you can instruct the AI to ask you the critical questions before it starts writing.

Example:

"You are creating this content for Brynne Tillman, CEO of Social Sales Link. Before you begin, ask me: (1) What's the core insight for this post? (2) Who specifically is the audience? (3) Is there a client story or example I want to include?"

Now the AI interviews you before it writes. That's not just efficient — it's how you build a repeatable system instead of starting from scratch every time.

The Difference Is Hard to Miss

Without a framework, prompting feels like gambling. You type something, hit enter, hope for the best. Sometimes it works. Usually you're editing more than if you'd just written it yourself.

With CRISPY™, prompting becomes something you can repeat. You build a template once, refine it as you go, and get output that actually sounds like you — not like everyone else using the same chatbot with the same vague instructions.

That distinction matters more every month. AI content is everywhere and readers are developing a nose for it fast. The way to stand out isn't to avoid AI — it's to use it in a way that amplifies your voice instead of replacing it with an average of everyone else's.

Start With One Prompt

Pick something you've tried to get AI to write and were unhappy with the result. Run it through CRISPY™. Give the AI the background, the role, a tone reference, a format, your no-go list, and your intake questions.

See what comes back.

My bet is it won't sound like AI content anymore. It'll sound like you.

What's the one type of content where AI keeps missing the mark for you? Drop it in the comments — I'd love to dig into the prompt with you.

Next Steps

Try the CRISPY™ prompt framework with any of the FUEL AI instructors. brynne.ai

Check out Prompt Writing Made Easy on Amazon - it is the CRISPY™ playbook and will help you prompt perfectly.