“What prompts should I use?”
I get this question constantly. And I get it—when you’re staring at a blank AI chat, it feels like there’s some secret formula you’re missing. Some magic combination of words that’ll unlock incredible outputs.
Here’s the truth: prompts matter. But not in the way most people think.
The entrepreneurs getting the best results from AI aren’t collecting massive prompt libraries or memorizing templates. They understand the PRINCIPLES behind good prompts—and more importantly, they’ve moved beyond prompts entirely into trained AI assistants that already know their business.
But let’s start with the fundamentals. Because whether you’re brand new to AI or you’ve been dabbling for a while, understanding how to communicate with AI will change everything.
And spoiler alert: it’s way simpler than the “prompt engineering gurus” want you to believe.
Why Do Most AI Prompts Give Mediocre Results?
Vague prompts get vague outputs. AI isn’t a mind reader—it needs specific context, clear instructions, and defined constraints to produce anything useful.
Here’s what most people do:
They type something like “Write me a social media caption” and hit enter.
Then they’re shocked when they get generic, bland, could-have-been-written-by-anyone content.
The AI isn’t broken. You just didn’t give it anything to work with.
Think of AI like a new team member on their first day. If you said “write me a social media caption” to a human with zero context about your business, your audience, your voice, or what you’re promoting—you’d get garbage too.
AI is the same. The quality of what goes IN directly determines the quality of what comes OUT.
This isn’t about memorizing fancy prompt templates. It’s about learning to communicate clearly.
And honestly? If you can explain what you want to a human, you can prompt AI effectively.
What Makes a Good AI Prompt?
Great prompts include five key elements: role, context, specific task, constraints, and desired format.
I’m going to give you a simple framework. But before I do, I need you to hear this:
Don’t get paralyzed by this.
The biggest mistake I see isn’t bad prompts—it’s people getting so stressed about “prompting correctly” that they freeze up entirely. They read seventeen articles about prompt engineering, save a hundred templates, and then never actually USE AI because they’re afraid of doing it wrong.
Just start talking to AI. Seriously. The more you do it, the better you’ll get at it.
Use voice dictation if typing feels intimidating. Explain what you need like you’re talking to a smart assistant. Ramble a little. It’s fine.
Perfect prompts aren’t the goal. Getting useful outputs is the goal.
Okay, with that said—here’s the framework:
1. Role
Tell AI who it should be or what expertise it should draw from.
Weak: “Write me an email.”
Better: “You’re an expert email copywriter who specializes in online course launches.”
When you assign a role, AI pulls from relevant knowledge and adjusts its tone and approach accordingly. It’s like the difference between asking “someone” for advice versus asking a specialist.
2. Context
Give AI the background information it needs to understand your situation.
Weak: “Write a sales page.”
Better: “I’m launching a $497 course called ‘Email Fix’ that teaches online entrepreneurs how to write better newsletters. My audience is mostly women aged 25-45 who run coaching or course businesses. They know they should be emailing their list but find it overwhelming and time-consuming.”
The more relevant context you provide, the more tailored the output. AI can’t read your mind—tell it what it needs to know.
3. Specific Task
Be crystal clear about what you want AI to actually DO.
Weak: “Help me with my launch.”
Better: “Write 5 subject line options for my launch announcement email. The email introduces my new course and creates excitement without being salesy.”
Vague requests get vague results. Specific requests get specific results. This is the single biggest lever you have.
4. Constraints
Tell AI what to include, what to avoid, and any boundaries to work within.
Weak: (No constraints)
Better: “Keep it under 100 words. Don’t use the words ‘excited’ or ‘thrilled.’ Match my casual, slightly sassy brand voice. Avoid corporate jargon.”
Constraints focus the output and prevent AI from going off in directions you don’t want. Think of them as guardrails.
5. Desired Format
Specify how you want the output structured.
Weak: (No format specified)
Better: “Give me the 5 subject lines in a numbered list, with a brief note under each explaining why it works.”
AI can give you bullet points, paragraphs, tables, step-by-step guides, scripts—but you have to ask for what you want.
Want to create your own brand book? Grab the 2-Hour Brand Book Here →
What Does This Look Like in Practice?
A complete prompt combines all five elements into a clear, specific request that gives AI everything it needs to produce useful output.
Let me show you the difference.
The Weak Prompt:
“Write me a newsletter about AI.”
What you’ll get: Generic, surface-level content that could be about anything. Probably starts with “In today’s rapidly evolving digital landscape…” 🤮
The Strong Prompt:
“You’re an expert email copywriter who writes in a casual, direct, slightly sassy tone.
I’m writing a newsletter to my audience of online entrepreneurs—mostly women who run coaching or course businesses. They’re curious about AI but overwhelmed by all the options and afraid of sounding like robots.
Write a 400-word newsletter that shares my perspective: AI is a tool for execution, not idea generation. The best AI results come from starting with YOUR ideas and using AI to structure and polish them.
Include a personal anecdote about my experience using AI for email (I went from 3+ hours per newsletter to 30 minutes). End with a CTA to check out my Email Fix program.
Don’t use corporate jargon, the word ‘excited,’ or phrases like ‘dive in’ or ‘landscape.’ Keep paragraphs short—max 3 lines each.”
What you’ll get: Something actually useful that sounds like a human wrote it for a specific audience.
See the difference?
The second prompt isn’t complicated. It’s just… complete. It gives AI everything it needs.
What Are the Biggest Prompting Mistakes?
The three most common mistakes are being too vague, asking for too much at once, and getting paralyzed by overthinking the “right” way to prompt.
Mistake #1: Too Vague
I see this constantly. “Write me content.” “Help with my business.” “Create something for social media.”
These prompts give AI nothing to work with. You’ll get generic outputs that require heavy editing—which defeats the whole purpose.
The fix: Add context. What kind of content? For whom? About what? In what tone? For what purpose?
You don’t need to write a novel. But you need to give AI more than five words.
Mistake #2: Asking for Too Much at Once
“Write me a complete launch sequence with 15 emails, a sales page, 30 social media posts, and a webinar script.”
AI isn’t magic. When you ask for too much at once, the quality of everything suffers. You end up with shallow, generic content across the board.
The fix: Break it down. Work on one piece at a time. Get the first email right, then move to the second. Build the sales page section by section.
Better outputs on smaller tasks beats mediocre outputs on massive requests.
Mistake #3: Analysis Paralysis
This one’s sneaky. People get so worried about “prompting correctly” that they never actually use AI. They collect templates, watch tutorials, read articles—but freeze when it’s time to actually type something.
The fix: Just start. Seriously.
Talk to AI like you’re explaining something to a smart friend. Use voice dictation if typing feels like pressure. Ramble a little—AI can handle it.
Your first prompt doesn’t need to be perfect. Send it, see what you get, then refine. The conversation IS the process.
The more you use AI, the better you’ll get at communicating with it. There’s no shortcut—you have to actually practice.
Does It Matter Which AI Platform I Use?
The prompting principles work across all major AI platforms—Claude, ChatGPT, Gemini—though each has slightly different strengths.
Everything I’ve shared works whether you’re using ChatGPT, Claude, Gemini, or any other AI assistant.
Personally, I use Claude as my AI platform of choice. I’ve migrated almost entirely to it for my daily work—content creation, strategic thinking, brainstorming, you name it. But the prompting fundamentals are the same across platforms.
The bigger factor isn’t which platform you choose—it’s how consistently you use it and how well you communicate what you need.
Pick one platform, get comfortable with it, and go deep. You can always explore others later.
Are Prompts Even the Real Answer?
Good prompts are the starting point, but the real unlock is trained AI assistants with your Brand Book and knowledge files built in—so you don’t have to prompt from scratch every time.
Here’s what most “prompt engineering” content won’t tell you:
Prompts are just the beginning.
If you’re writing detailed prompts from scratch every single time you use AI, you’re doing it the hard way. Yes, you’ll get better results than vague prompts—but you’re still spending way more time than necessary.
The real unlock? Trained AI assistants.
When you build a custom assistant with your Brand Book, your voice guidelines, your audience insights, and your methodology baked in—you don’t need elaborate prompts anymore.
Instead of this:
“You’re an expert email copywriter who writes in a casual, direct, slightly sassy tone. My audience is online entrepreneurs, mostly women who run coaching or course businesses. They value freedom and flexibility. Don’t use corporate jargon or the word ‘excited.’ Keep paragraphs short…”
You just say:
“Write a newsletter about why AI shouldn’t generate your ideas.”
And the assistant ALREADY KNOWS everything else. Because you trained it once, and now it works for you every time.
This is the difference between using AI as a tool and having AI as a team member.
I teach exactly how to build these trained assistants inside AI All Stars. It’s the system that lets me run my business in 25 hours a week—not by prompting better, but by building AI that already knows my business inside and out.
How Do You Get Started?
Start by using AI daily with the framework, then level up to trained assistants once you’re comfortable with the basics.
Here’s the progression I recommend:
Level 1: Learn to Prompt Well
Use the framework: Role + Context + Task + Constraints + Format.
Practice daily. Don’t stress about perfection—just get reps in. Talk to AI like a human. Refine based on what you get back.
This alone will dramatically improve your outputs.
Level 2: Create Your Brand Book
Document your voice, audience, offers, and personality in one comprehensive document. (I covered this in depth in my article on training AI to write in your brand voice.)
Upload this document at the start of AI conversations. Instant improvement in how “you” the outputs sound.
Level 3: Build Trained Assistants
Create custom AI assistants with your Brand Book and relevant knowledge files built in. Different assistants for different tasks—email writing, social content, podcast notes, client support.
Now you’re not prompting from scratch. You’re directing a team that already knows you.
Level 4: Build Systems
Connect your AI assistants into workflows and automations. Content that creates itself. Support that handles itself. A business that runs smoother than ever.
This is where the real time savings happen.
Want to see exactly how this works? Join my free AI workshop →
The Bottom Line
Prompts matter. Learning to communicate clearly with AI is a skill worth developing.
But don’t get stuck at the prompt level.
The entrepreneurs winning with AI aren’t the ones with the fanciest prompts. They’re the ones who’ve built trained assistants that already know their business—so every interaction is fast, efficient, and on-brand.
Start with good prompts. Graduate to trained assistants. Build systems that scale.
That’s the real path.
Now stop reading and go talk to some AI. 🔥
Related Reading:
