Module 06 — Practical Prompting Skills

📖

Lesson Content

Read & Understand

A prompt is everything you give to an AI before it responds. The quality of the prompt determines the quality of the output — and most people write terrible prompts, then blame the AI when results are poor.

The most important principle: be specific. "Write me an email" gives the AI almost no information. "Write a friendly follow-up email to a client who hasn't responded to my proposal in two weeks — keep it under 100 words and end with a clear call to action" gives it everything it needs.

A strong prompt has four dimensions: Role (tell the AI who to be), Context (give background), Task (state exactly what you want), and Format (specify how the output should look). You don't need all four every time, but knowing them helps you identify what's missing when outputs disappoint.

Prompting is also iterative. Your first prompt rarely produces the best result. Think of it as a conversation — review the output, identify what's off, and refine. Treat each prompt as a hypothesis and each response as evidence about how to prompt better.

Key Takeaways

Specificity is the #1 lever for better AI outputs

Good prompts have: Role, Context, Task, and Format

Prompting is iterative — treat it like a conversation

Bad outputs usually mean bad prompts, not bad AI

More context is almost always better than less

🔬

Anatomy of a Great Prompt

The four ingredients

🎭

Role

Tell the AI who to be or what expertise to draw on

"You are an experienced UX designer reviewing a mobile app prototype..."

📋

Context

Provide background the AI needs to understand the situation

"...The app is for seniors aged 65+. Our main issues are low retention and confusing navigation..."

🎯

Task

State clearly and specifically what you want the AI to do

"...List 5 specific UX improvements with a one-sentence rationale for each..."

📐

Format

Specify how the output should be structured or presented

"...Use a numbered list. Keep each improvement to 2 sentences max."

🔍

Before and After Prompts

Specificity in action

Before

Write me a bio.

After

Write a third-person professional bio for me for LinkedIn. I'm a freelance graphic designer with 8 years of experience specializing in brand identity for tech startups. Tone: confident but approachable. Length: 3 sentences. End with a line about my passion for minimalist design.

Before

Explain machine learning.

After

Explain machine learning to a complete beginner with no technical background. Use a real-world analogy (not a computer science example). Keep it under 150 words. Avoid all jargon — if you must use a technical term, define it immediately.

📊

The Specificity Spectrum

From vague to precise

VagueBetterPrecise

"Help me with my presentation."

"Improve the structure of this 10-slide sales presentation."

"Review this 10-slide sales deck for a B2B SaaS product pitched to CFOs. Suggest 3 structural improvements and rewrite the opening slide hook to be more compelling. Keep the tone formal."

✅

Self-Check Quiz

Click an answer to check your understanding

Q1 of 4

Which of the four prompt dimensions tells the AI who to be?

Context

Role

Task

Format

✓ Role tells the AI what expertise or persona to draw from — 'You are a senior copywriter...' or 'Act as a skeptical investor...'

✗ That's 'Role' — it tells the AI who to be or what perspective to take.

Q2 of 4

If an AI gives you a bad output, what should you do first?

Switch to a different AI

Accept that AI isn't capable of the task

Review your prompt and add more specificity or context

Rephrase the exact same prompt

✓ Bad outputs usually signal missing context or specificity in the prompt — iterate by adding more information.

✗ Bad outputs usually mean the prompt needs work. Review what's missing — role, context, task clarity, or format.

Q3 of 4

Which prompt would produce the most useful output?

"Write an email."

"Write a professional email."

"Write a follow-up to a client who hasn't responded in 2 weeks. Friendly but firm. Under 80 words. End with one clear next step."

"Write the best email you possibly can."

✓ Option C provides situation, tone, length, and format — all four dimensions working together.

✗ Option C is the strongest — it specifies the situation, tone, length, and structure. The others leave too much ambiguous.

Q4 of 4

What does the "Format" dimension of a prompt control?

The file type the AI saves its response in

How the output is structured — length, layout, style

The language the AI responds in

Whether the AI uses markdown or plain text

✓ Format tells the AI how to structure the output — length limits, bullet points vs. prose, headers, etc.

✗ Format specifies structure and presentation: 'Use a numbered list', 'Keep it under 150 words', 'Write in paragraphs, not bullets'.

🧩

Exercises & Worksheets

Apply what you learned

Anatomy Labeling

Take this prompt and label which parts cover Role, Context, Task, and Format — and identify what's missing: 'You're a nutritionist. I'm trying to eat fewer processed foods. Give me a 5-day meal plan with breakfast, lunch, and dinner. I don't like fish.'

📝 Analysis

Rewrite a Bad Prompt

Take one of these vague prompts and rewrite it using all four dimensions: (a) 'Help me write something for social media.' (b) 'Explain what a contract is.' (c) 'Give me advice about my job.'

✍️ Writing

The Iteration Test

Send a vague prompt to an AI. Read the output. Write one sentence describing what's missing. Add exactly that to the prompt and run again. Repeat 3 times. Document how the output improved each time.

🔄 Experiment

Build Your Prompt Template

Create a reusable prompt template for a task you do regularly. Include Role, Context, Task, and Format as labeled placeholders. Test it on 3 different examples and refine until it works reliably.

🛠 Practical