Prompting: The Skill Behind the Skill
[Open ChatGPT and type:] "Write me an email."
[Read the response aloud. It will be generic. Maybe it says something like "Subject: Hello, Dear [Name], I am writing to reach out to you about..."]
Now watch what happens when we change the instruction. [Type:] "You are a professional virtual assistant writing on behalf of a busy Nigerian entrepreneur. Write a warm but professional email to a potential client who expressed interest in social media management services on Instagram, but has not responded to the first message sent 3 days ago. The email should be friendly, not pushy, and end with a specific call to action. Keep it under 150 words."
[Read that response aloud. It will be significantly better.]
The same AI. The same tool. Completely different results. The only thing that changed was the quality of the instruction. That instruction is called a prompt. And learning to write good prompts is the single most valuable AI skill you can develop.
Why Most People Get Mediocre AI Results
People type vague, short, context-free questions and expect detailed, specific, useful answers. That is like calling a contractor and saying "build me something" and then being frustrated that they did not build what you wanted.
AI is powerful but it is not a mind reader. It works with what you give it. Give it vague: you get vague. Give it specific, context-rich, well-structured instructions: you get specific, useful output. This lesson teaches you exactly how to do that.
Anatomy of a Perfect Prompt: The RCTF Framework
See the framework in a real prompt
Weak prompt vs. strong prompt: side by side
5 Advanced Prompting Techniques
The Prompt Upgrade Challenge
The weak prompt is: "Write me something for social media."
You have 10 minutes. Use ChatGPT or Claude. Write your upgraded prompt in your Workbook first, then run it. When you have both results, we will hear from 3 people about what they changed and why.
Lesson 10 Assignment
- 1
Open your AI Prompt Library Google Doc. Write a new section: "Lesson 10: The Art of Prompting." You will document your 5 prompts here.
- 2
Write 5 original RCTF prompts for real tasks in your VA work or job search. Each prompt must include all 4 RCTF elements. Examples: a cold email to a potential client, a LinkedIn profile summary, a caption for a client's Instagram post, a client proposal outline, a research brief on a topic you need to understand.
- 3
Run each prompt in ChatGPT or Claude. Paste the best output you received into a Google Doc called "AI-Generated Drafts: Lesson 10." Save it in your Drive.
- 4
For each of the 5 outputs: write one sentence evaluating it. Was it good enough to use? What would you change? What did you learn about prompting from this attempt? Add these notes in your Workbook.
- 5
Add this lesson's 5 prompts to your AI Prompt Library document (the ones in the Prompt Library tab of this lesson). Title the section "Lesson 10: Advanced Prompting Techniques."
You Are Now a Prompt Engineer
You now know the foundational framework and the advanced techniques that underpin that work. You are not going to apply for a prompt engineering role tomorrow. But you are going to use this skill every single day as a VA, as a digital professional, and eventually as someone who helps others use AI effectively.
In Lesson 11, we take these prompting skills and apply them to a specific, high-value task: using AI for research and content creation. That is where prompting goes from an exercise to a professional weapon.
- 5 original RCTF prompts written, run, and documented
- Best AI outputs saved in "AI-Generated Drafts: Lesson 10" Google Doc
- One-sentence evaluation written for each prompt output
- Lesson 10 prompts added to AI Prompt Library in Drive
- RCTF framework memorised (Role, Context, Task, Format)
Lesson 10 Student Workbook
Write, Run, and Evaluate 5 RCTF Prompts
Part 1: Set Up Your Documentation
- 1
Open your AI Prompt Library Google Doc. Add a new heading: "Lesson 10: The Art of Prompting." This is where you will document your 5 original prompts and their outcomes.
- 2
Create a second Google Doc called "AI-Generated Drafts: Lesson 10." Save it in your Drive in the same folder as your Prompt Library. This is where you paste the actual AI output for each of your prompts.
- 3
In the AI-Generated Drafts document, add 5 sections labelled "Prompt 1 Output," "Prompt 2 Output," and so on. You will paste each AI response into its corresponding section.
Part 2: Write and Run Your 5 Prompts
- 1
Choose 5 tasks from the following list (or use your own real tasks). For each, write a full RCTF prompt before running it: (a) A cold outreach message to a potential client on LinkedIn, (b) A bio for your VA services page or profile, (c) A social media caption for an imaginary client's business, (d) A summary of what VA services are (to explain to someone asking), (e) A research summary on a topic relevant to your work.
- 2
For each prompt: write the full RCTF prompt in your Prompt Library doc first. Then open ChatGPT or Claude and run it. Paste the output into your AI-Generated Drafts doc.
- 3
After running all 5 prompts, go back to each one and add a short evaluation note: what worked well, what you would change, and whether the output is usable as-is or needs editing. Write this in your Workbook (Part C) and in your Prompt Library doc.
Part 3: Iterate on Your Best Prompt
- 1
Pick the prompt output you found most useful. Now practice the Iteration technique: send a follow-up message asking AI to improve it. For example: "That is good, but make it 20% shorter and change the tone to be more confident and direct."
- 2
Run the iteration once. Save the improved version in your Drafts doc alongside the original. Note: which version is better and why?
- 3
Now try the Options technique on a different prompt: send a follow-up asking for 3 alternative versions. Note which version you prefer and why. This two-step practice builds the habit of treating AI output as a starting point, not a final product.