How to Write Clear, Effective Prompts

The single biggest factor in getting useful output from an AI is the quality of your input. A vague, ambiguous prompt will produce a vague, generic response. A clear, specific prompt — one that tells the model exactly what you want, in what format, for what purpose — will produce dramatically better results.

This is not a matter of finding secret keywords or magic phrases. It is about communicating clearly. If you can explain what you want to a reasonably intelligent colleague, you can write a good prompt.

Why Vague Prompts Get Vague Answers

Remember from Lesson 1 that LLMs predict the most likely continuation of your text. If your prompt is vague, there are thousands of possible valid continuations. The model will pick one — but it might not be the one you wanted.

Example: A Weak Prompt

"Tell me about climate change."

This could lead to a 10,000-word essay, a three-sentence summary, a list of causes, a policy analysis, a children's explanation, or a debate between perspectives. The model has no way to know which one you want, so it picks the "most likely" type of response — which is usually a generic, medium-length overview that is not particularly useful to anyone.

Example: A Strong Prompt

"Summarise the three main causes of climate change in 150 words, suitable for a Year 10 geography student. Use simple language and include one real-world example for each cause."

Now the model knows:

What to produce (a summary of three causes)
How long (150 words)
For whom (Year 10 student)
What style (simple language)
What to include (real-world examples)

The difference in output quality is enormous — and all it took was being specific about what you wanted.

The Four Pillars of a Good Prompt

While there is no single formula that works for every situation, most effective prompts include some combination of these four elements:

1. Be Specific About the Task

Tell the model exactly what you want it to do. Use action verbs: summarise, compare, list, explain, translate, rewrite, generate, analyse.

Weak	Strong
"Help me with this essay"	"Proofread this essay for grammar errors and suggest improvements to the introduction paragraph"
"Write about dogs"	"Write a 300-word informational paragraph about why golden retrievers are popular family pets"
"Explain photosynthesis"	"Explain photosynthesis in three steps, suitable for a 12-year-old, using an analogy to cooking"

2. Provide Context

The model has no idea who you are, what you are working on, or why you need this information — unless you tell it. Context dramatically narrows the range of possible outputs.

Useful context includes:

Your audience: "This is for university-level biology students"
Your goal: "I am preparing a presentation for a client meeting"
Background: "I have already written the introduction; I need help with the methodology section"
Your level: "I am a beginner learning Python for the first time"

3. Specify the Format

If you want bullet points, ask for bullet points. If you want a table, ask for a table. If you want prose, say so. If you want the response in a specific structure — introduction, body, conclusion — describe that structure.

Common format instructions:

"Reply in bullet points"
"Format your response as a table with columns for [X] and [Y]"
"Give me a numbered list"
"Write this as a formal email"
"Keep your response under 200 words"
"Use markdown headers to organise your response"

4. State What to Avoid

Sometimes it is just as important to say what you do NOT want. This is especially useful when you have noticed the model defaulting to a pattern you dislike.

Examples:

"Do not include a motivational closing paragraph"
"Avoid jargon — explain any technical terms"
"Do not use bullet points — write in flowing paragraphs"
"Do not repeat information I have already provided"