Prompt engineering is writing clear instructions so an AI tool gives accurate, useful output. A strong prompt states the role, the task, the context, any constraints, an example if possible, and the format you want back. Then you review the answer, ask follow-up questions and refine. Clarity beats clever tricks.
Why the prompt matters so much
Ask an AI "explain photosynthesis" and you get a generic paragraph. Ask "Explain photosynthesis to an O-Level student who keeps confusing it with respiration. Use a simple analogy, then a 5-question quiz with answers at the end" and you get something genuinely useful. Same tool, same model. The difference is the prompt.
AI models do not know who you are, what you already understand, or what you will do with the answer. Prompt engineering is simply the habit of telling them.
The six parts of a strong prompt
You will not always need all six, but checking them is the fastest way to improve.
- Role: who should the AI act as? "You are an experienced chemistry teacher."
- Task: what exactly do you want? Use a clear verb: explain, summarise, compare, draft, critique, list.
- Context: the background it needs. Your level, your goal, your audience, the material to use.
- Constraints: length, tone, what to avoid, what must be included.
- Example: a sample of the output style you want. This is often the single most powerful addition.
- Format: a table, bullet points, a step-by-step plan, a short paragraph.
Before and after
Weak: "Write about climate change."
Strong: "You are a geography tutor. Write a 200-word explanation of the enhanced greenhouse effect for a 14-year-old preparing for IGCSE Geography. Use one real-world example, avoid jargon except key exam terms in bold, and finish with two exam-style questions."
Techniques that work
Ask for steps before answers
For maths, logic or planning, ask the model to work through the problem step by step, or to outline its approach first. This reduces careless errors and lets you spot where reasoning goes wrong.
Give examples (few-shot prompting)
If you want a particular style, show it. Paste two examples of flashcards in the format you like, then ask for twenty more on a new topic.
Iterate, do not restart
Treat the first answer as a draft. Follow up: "Make it shorter", "Explain step 3 differently", "Which part of this are you least sure about?"
Ask it to critique itself
"Review your answer for factual errors and anything a teacher might mark wrong." Models often catch their own mistakes when asked directly.
Ground it in your material
Paste your notes or upload the chapter and say "Use only the text below. If the answer is not there, say so." This cuts hallucinations dramatically.
Ready-to-use prompts for students
| Goal | Prompt |
|---|---|
| Understand a topic | "Explain [topic] as if I am a [level] student. Start with a simple analogy, then the proper explanation, then a common misconception." |
| Test yourself | "Quiz me on [topic] one question at a time. Wait for my answer, tell me if I am right and why, then ask the next question." |
| Improve your writing | "Here is my paragraph. Do not rewrite it. List the three biggest weaknesses and suggest how I could fix each one." |
| Plan revision | "I have [x] weeks until my exam on [subjects]. I can study [y] hours a day. Make a realistic plan using spaced practice." |
Notice the writing prompt asks for feedback, not a rewrite. That keeps the work yours, which matters for honesty and for actually learning.
Common mistakes
- Being vague: "make it better" gives random changes. Say what better means.
- Overloading one prompt: split big jobs into steps.
- Trusting the first answer: always check facts that matter. See how to fact-check AI answers.
- Sharing private information: never paste passwords, ID numbers or confidential documents into public tools.
Where to go next
Once the basics feel natural, learn the prompt patterns professionals use in our guide to prompt patterns that work, or join our live prompt engineering course, which includes premium AI tools and a personal prompt library built around your goals.
Frequently asked questions
Is prompt engineering a real skill or a fad?
The tricks change, but the core skill, giving clear goals, context and constraints and judging output, is lasting and valuable in almost every job.
Should prompts be long?
As long as needed and no longer. Context and examples help; padding and repetition do not.
Do prompts work the same in every AI tool?
The principles do. Each model has quirks, so test important prompts in the tool you will actually use.
