Prompt Engineering Basics: How to Ask AI the Right Way
Learn the Role–Task–Context–Format method and other simple techniques that make AI answers dramatically better.
- Beginner
- 14 min
- Updated
What you will learn
- Use the Role–Task–Context–Format method
- Give examples to guide AI output
- Break complex work into steps
A prompt is what you type to an AI. Prompt engineering simply means writing prompts that get reliable, useful results. It is the single most valuable AI skill for beginners.
The R-T-C-F method
- Role — who should the AI act as? “Act as a friendly maths teacher.”
- Task — what exactly do you want? “Explain percentages.”
- Context — background that changes the answer. “For a class 7 student who finds maths scary.”
- Format — how the answer should look. “Three short steps and one practice question.”
Weak prompt vs. strong prompt
| Weak | Strong |
|---|---|
| Write a post about my shop. | Act as a social media writer. Write an Instagram caption for my bakery in Surat announcing eggless Diwali sweets. Friendly tone, 3 emojis maximum, under 50 words, end with a question. |
Five techniques that always help
- Give an example. Show one sample of the style you want, and the AI will follow it.
- Set limits. Word count, reading level, tone, language.
- Ask for questions first. “Before answering, ask me 3 questions to understand my needs.”
- Break big tasks into steps. Outline first, then each section.
- Ask it to check itself. “Review your answer for mistakes and missing points.”
Common mistakes
- Asking several unrelated things in one prompt.
- Not saying who the output is for.
- Accepting the first answer without refining it.
Practise with the prompt box below and the free Prompt Starter Pack. Within a week, you will notice a big difference.
The R-T-C-F prompt in action
ROLE: Act as an experienced Excel trainer.
TASK: Explain how to create a pivot table of total sales by region.
CONTEXT: I am a complete beginner using Excel on Windows.
FORMAT: Numbered steps (max 8) and one common mistake to avoid.