Most people who feel like AI is not useful have a prompting problem, not an AI problem.
The same tool that produces a vague, unhelpful response to one person produces a detailed, exactly-right answer for another. The difference is almost always in how the question was asked.
This guide shows you how to write better prompts — with real examples you can use today across ChatGPT, Claude, Gemini, and any other AI tool.
Why Your Prompts Matter More Than You Think
AI tools are not mind readers. They respond to exactly what you write.
When you write a vague prompt, the AI has to guess your goal, your audience, the tone you want, the format you need, and what success looks like. When it guesses wrong — and it often will — you get a generic response that does not quite hit the mark.
Better prompts reduce that guesswork. They do not require technical knowledge or special formulas. They require the same thing a good brief to a colleague requires — enough context to do the job well.
The Four Things Every Good Prompt Needs
A strong prompt covers four things. You do not always need all four for simple tasks, but understanding each one helps you know what is missing when a response misses the mark.
1. What you want — the request
Be specific about exactly what you are asking for. The more precisely you describe the output, the more targeted the response.
Vague: “Write something about social media.” Better: “Write a 300-word LinkedIn post about why small businesses should post consistently on social media, aimed at business owners who are new to the platform.”
2. Context — who, what, and why
Tell the AI who this is for, what situation it is responding to, and why it matters. Context dramatically changes the quality of the output.
Without context: “Write an email following up on a meeting.” With context: “Write a short follow-up email to a potential client after a first meeting about our marketing services. The meeting went well and we agreed to send a proposal by Friday. Keep it warm and professional.”
3. Role — who should the AI be
Telling the AI to respond from a specific perspective shifts the tone, depth, and relevance of the answer.
Without role: “Explain compound interest.” With role: “You are a financial advisor explaining compound interest to a 22-year-old who has just started their first job and knows nothing about investing. Keep it simple and practical.”
4. Format — how you want the output
Tell the AI how to structure the response. Without this, it often over-produces or formats things in a way that does not fit your needs.
Without format: “Give me ideas for blog posts.” With format: “Give me 10 blog post ideas for a beginner-friendly AI tools website. Format as a numbered list with a one-sentence description for each idea.”
Common Prompting Mistakes and How to Fix Them
Mistake 1 — Being too vague
The most common mistake. If you would not give a human colleague such a vague brief and expect great work, do not expect it from AI either.
Fix: Add specifics — who is the audience, what is the purpose, what does a good result look like.
Mistake 2 — Asking for too many things at once
Long, complicated prompts that ask the AI to research, analyse, write, and format all in one go usually produce mediocre results across all parts.
Fix: Break complex tasks into steps. Do the research first, then the writing, then the formatting — each as a separate prompt.
Mistake 3 — Not specifying length
Without a length constraint, AI often over-produces. A prompt asking for a social media caption can return five paragraphs.
Fix: Be explicit. “In 150 words”, “as a three-sentence summary”, “in one paragraph” — these small additions make a significant difference.
Mistake 4 — Accepting the first response
Most people treat the first AI response as final. It rarely is. The first response is a starting point.
Fix: Follow up. “Make it shorter”, “make it more conversational”, “change the opening line”, “add a specific example” — iterating on the first response is how you get to genuinely useful output.
Mistake 5 — Not giving examples
Telling AI what you want is good. Showing it an example of what good looks like is better.
Fix: Paste in an example of the tone, style, or format you want and say “match this style”. Even a single example dramatically improves consistency.
Practical Prompt Templates You Can Use Today
For writing a blog post: “Write a beginner-friendly blog post about [topic] for an audience of [who]. Use a friendly, clear tone. Include an introduction, three main sections with practical examples, and a conclusion with a clear takeaway. Aim for around [word count] words.”
For summarising a long document: “Summarise the following document in plain English. Identify the three most important points and any action items. Keep the summary under 200 words.”
For writing an email: “Write a [tone — professional/friendly/direct] email to [who] asking [what]. The context is [brief background]. Keep it under 150 words and end with a clear next step.”
For brainstorming ideas: “Give me 10 ideas for [topic]. The audience is [who]. Format as a numbered list with a one-sentence explanation for each idea.”
For rewriting something you have already written: “Rewrite the following text to be more [concise/engaging/professional/friendly]. Keep the key information but improve the flow and make it easier to read. Here is the original: [paste text]”
For explaining a complex topic: “Explain [topic] as if you are talking to someone with no background knowledge. Use simple language, avoid jargon, and include one practical example to make it concrete.”
How to Build a Personal Prompt Library
Once you find a prompt that works well for a task you do regularly, save it.
Keep a simple document — a Notion page, a Google Doc, or even a notes app — with your ten most-used prompts. Label each one by task type. Over time this becomes a personal toolkit that saves you significant time every week.
Good prompts to start saving: your email follow-up template, your blog post brief, your summary request, and your social media caption format.
A Simple Rule to Remember
Before you write a prompt, ask yourself: if I gave this instruction to a smart colleague who knew nothing about my situation, would they be able to do a good job?
If the answer is no, add more context. That single question catches most prompting mistakes before you make them.
Summary — The Prompting Checklist
Before you send a prompt, run through these five questions:
Have I said exactly what I want? Have I given enough context — who it is for, why it matters? Have I assigned a role if the task needs a specific perspective? Have I specified the format and length? Have I included an example if style or tone matters?
Three or more of these covered and your prompt will consistently produce better results.
Related reading: How to Use ChatGPT for Work How to Use AI to Write Emails Faster How to Pick the Right AI Tool for Your Needs