Here's a quick test. Think about the last time you used AI. Did you open it, type a question, skim the answer, tweak it a bit, and move on?
If yes, you're using AI the way most people do. And that's exactly why most people aren't getting the results they're hoping for.
I see this constantly when I'm speaking at conferences or running workshops. Business owners are using AI every day, but they're still frustrated. The outputs feel generic. The back-and-forth takes longer than just doing it themselves. And eventually, they decide AI just isn't that useful for their situation. The problem isn't the tool. It's the approach.
Average business owners use AI like a search engine. The best ones treat it like staff. Same tool. Completely different outcomes.
Why does treating AI like a search engine hold you back?
When you treat AI like Google, you're asking it a question and expecting a good answer based on almost no context. That's fine for looking up a restaurant. It's not fine when you're asking for strategic business advice, marketing copy, or client communications.
Think about how you'd feel if a new team member turned up on their first day, you gave them zero briefing, and then expected them to produce work that matched your standards. You wouldn't do that. Yet that's exactly what most business owners do with AI every single day. The good news is that once you shift your approach, the results change dramatically. Here are the four things I've seen top business owners do differently.
What do the best AI users actually do differently?
1. They start with personalisation
This is a setting inside ChatGPT that most people never touch. Great AI users begin with what I'd call an orientation. Before they ask the AI to do anything at all, they tell it who they are, what they do, and what good looks like.
Something like: "You're helping me run a professional services business. My clients value clarity, trust, and judgment. I don't want hype, jargon, or clever tricks. Just practical, calm recommendations. If something is risky, flag it. If something is speculative, say so." My personal favourite to add: "Be a ruthless mentor and challenge my thinking." This isn't just prompting. This is onboarding.
2. They use projects for continuity
Most people start from scratch every time they open ChatGPT. Top AI users don't. They organise their work into projects, which are essentially folders that carry context forward. I have separate projects for content creation, workshops, speaking engagements, and client work. I also use them in my personal life, health and nutrition, finances, parenting, even surfing.
When I open a project, the AI already knows who my clients are, the standards I work to, and everything relevant about that particular context. It stops feeling like a chatbot and starts feeling like a genuine workspace.
3. They set system instructions
Inside each project, you can write system instructions. This is where you stop hoping for good answers and start setting standards. You're telling the AI how to think and how to behave, not just what to do. For example: "Act as a senior business advisor. Default to structured thinking. Challenge weak assumptions. Optimise for leverage, not activity. Use plain English. Never overpromise. Ask clarifying questions."
Once that's in place, I don't need to explain myself every single time. Think of it as the brain of your AI setup. The part that makes it think the way you think.
4. They upload knowledge files
If system instructions are the brain, knowledge files are the memory. This is the piece most business owners never get to, and it's the one that makes the biggest difference. Knowledge files are documents you upload into your ChatGPT project, things like your writing style guide, business stories, client frameworks, methodologies, and examples of your best work.
Without them, the AI is improvising. With knowledge files, it knows exactly what you want, how you sound, and how you approach problems. The value compounds. This is the idea I've built into a full framework, the AI Business Brain.
How do you start using AI in business the way the top 1% do?
You don't need to overhaul everything at once. Start with the personalisation step, write a short orientation paragraph that tells the AI who you are and what good looks like. Do that today, and you'll already notice a difference. From there, create one project for the area of your business where you use AI most. Add a few lines of system instructions. Then gradually build out your knowledge files.
The model here is simple: treat AI like a new team member. Provide a brief. Give it training. Offer context. When you do, you stop getting generic outputs and start getting genuine leverage.
Frequently asked questions
What does it mean to treat AI like staff?
It means briefing it properly before asking it to do anything. Give it context about who you are, what you do, and what a good result looks like, the same way you'd onboard a new team member. When you do this, the quality of every response improves immediately.
What is the personalisation setting in ChatGPT?
It's a built-in feature that lets you tell ChatGPT about yourself and your business so it can adjust its responses. You access it through your profile settings. Most business owners never set it up, which is why their results feel generic.
What are ChatGPT projects, and how do I use them for business?
Projects are folders in ChatGPT that carry context between conversations. Instead of starting from scratch every time, you organise your work by topic, marketing, client work, speaking, operations, and the AI retains relevant context each time you open that project.
What are system instructions in ChatGPT?
System instructions are guidelines you write inside a project that tell the AI how to think and behave, not just what to do. They let you set standards upfront so you're not repeating yourself in every conversation.
What should I include in my AI knowledge files?
Start with four things: your writing style guide, your business background and stories, any frameworks or methodologies you use with clients, and examples of your best past work. These give the AI the context it needs to sound like you.
Do I need to be technical to set this up?
Not at all. Personalisation, projects, system instructions and knowledge files are all set up through plain text and file uploads. There's no coding involved. If you can write an email, you can do all of this.
