By Jacob Rucker | Apex CAS
A client called me last month, frustrated. She’d been using ChatGPT for three months and called it “a fancy search engine that writes worse than Google.” She’d type a question, get a generic paragraph back, and close the tab. She was ready to cancel her subscription.
I asked her to show me how she was using it. She typed: “How do I improve my onboarding process?” and hit enter.
That’s when I saw the problem. It wasn’t the tool. It was the conversation.
What Changed in 15 Minutes
I walked her through a simple AI prompting framework for business that we teach every client at Apex. Four steps. No technical knowledge required.
Step 1: Tell it who to be. Instead of just asking a question, start with a role. “Act as an operations consultant who specializes in client onboarding for professional services firms.” That one sentence changes the lens the AI uses for everything after it.
Step 2: Tell it what you need. Be specific about the deliverable. Not “help me with onboarding” but “give me a prioritized list of the five biggest bottlenecks in a 30-day client onboarding process, with one fix for each.”
Step 3: Give it the full picture. This is where most people fall short. They give two sentences of context and expect brilliance. That’s like briefing a new hire in 30 seconds and expecting them to run your department. Talk through what’s working, what’s not, what you’ve tried, and what the constraints are. If you’re speaking instead of typing, you can dump 3-4x more context in the same time. Five minutes of talking gives the AI more to work with than most people provide in a week of typed prompts.
Step 4: Ask it to ask you questions first. This is the step nobody does. End with: “Before you answer, let me know if you have any questions about what I just explained.” That gives the AI permission to clarify instead of guess. It surfaces blind spots you didn’t know you had. The response after that round? Dramatically better.
The Difference It Made
We ran through all four steps together. The AI came back with five specific bottlenecks she recognized immediately, including one she hadn’t thought of. She turned to me and said, “That’s more useful than three months of Googling.”
PwC’s 2026 Global CEO Survey found that 56% of CEOs report getting “nothing” from their AI adoption efforts. McKinsey reports only 1% of executives consider their AI rollouts mature. The gap isn’t the technology. It’s the interface between the human and the tool.
Think of AI like the smartest intern you’ve ever had. Knows everything, works 24/7, desperately wants to help. But like any intern, it won’t ask clarifying questions unless you give it permission. It’ll fill gaps with assumptions rather than admitting it needs more information. Step 4 fixes that.
If That Sounds Familiar
Most businesses don’t need more AI tools. They need to learn how to talk to the ones they already have. This framework is what we call Phase 1. Phase 2 is connecting AI to your meeting notes, CRM, and project tools so it already has context before you start talking. Phase 3 is automated workflows where AI handles the repetitive work and your team handles the judgment calls.
You can’t skip to Phase 3 without getting Phase 1 right. And Phase 1 is free. Open your AI tool, try the four steps, and see what happens.
If you want to see what Phases 2 and 3 look like for your business, we can show you. It starts with one conversation.
Frequently Asked Questions
How do I get better results from ChatGPT or Claude?
Use a 4-step framework: define the role (who should AI be), define the outcome (what deliverable you need), give full context (background, constraints, what you’ve tried), and ask AI to clarify before answering. This structured approach produces dramatically better output than typing a short question and hoping for the best.
Why does AI give generic answers?
The most common reason is insufficient context. Most people type 1-2 sentences and expect AI to understand their full situation. AI responds generically because it’s filling gaps with assumptions. Give it more context, either by typing or by using voice-to-text, and explicitly invite it to ask clarifying questions before responding.
What is the best AI prompting framework for business owners?
Start with role, outcome, context, and clarification. Tell AI what expertise to use, what deliverable you need, give it your full situation, then ask “do you have questions before answering?” This works across ChatGPT, Claude, and any other AI tool. No technical knowledge required.







