Prompt engineering basics: write prompts that get useful answers

"Prompt engineering" sounds technical, but it is really just clear briefing. If you would brief a smart new colleague well, you can prompt AI well. Use this five-part formula.

The RTCFC formula

Before and after

Before: "Write a LinkedIn post about our new course."

After:

"You are a friendly education marketer. Write a LinkedIn post announcing our free 4-week 'Python for Business' course. Audience: busy professionals with no coding background. Mention that each lesson is 20 minutes and ends with a practical exercise. Format: a hook line, 3 short paragraphs, 3 hashtags. Keep it under 120 words and avoid jargon."

The second prompt takes 30 seconds longer to write and saves ten minutes of rewriting.

Iterate like a conversation

• "Make it shorter and more direct."
• "Give me 3 alternative opening lines."
• "What information would make this better? Ask me questions first."

That last one is powerful: asking the AI to interview you before answering often produces the best results.

Common mistakes

1. Asking for everything in one giant prompt. Break big jobs into steps.
2. Not giving examples. If you have a good example of the output you want, include it.
3. Forgetting the audience. "For a CFO" and "for a new intern" produce very different answers.