You typed something into ChatGPT and got back a wall of waffle. Been there. The problem usually isn't the AI — it's the prompt. A vague input produces a vague output, every single time. The good news is that writing better prompts isn't some dark art. There are clear, learnable patterns behind the prompts that get results. This guide breaks them down, with practical techniques you can use today — across ChatGPT, Claude, and Copilot.
Why most prompts fall flat
Most people type into an AI the same way they’d type a quick Google search — short, vague, no context. That approach works fine for “weather in Manchester” but it’s a poor fit for an AI that can write, analyse, plan, and summarise.
AI tools like ChatGPT, Claude, and Copilot are essentially very capable assistants waiting to be briefed. The clearer your brief, the better the output.
The most common prompting mistakes
Too short and vague — “Write me an email” gives the AI almost nothing to work with. Who is it to? What’s the purpose? What tone?
No context — AI doesn’t know who you are, what you do, or what you’re trying to achieve unless you tell it.
No format guidance — If you want bullet points, a table, or a two-paragraph summary, say so. The AI will make its own guess otherwise.
Treating it like a search engine — Keywords don’t work the same way. Full sentences and natural instructions tend to produce far better results.
Giving up after one attempt — Your first prompt doesn’t have to be perfect. You can follow up, refine, and ask the AI to adjust.
The fix for all of these is the same: give the AI enough to work with. That starts with understanding what a good prompt actually looks like.
The anatomy of a great prompt
Good prompts tend to share the same ingredients. You don’t need all of them every time, but knowing each one helps you build better requests instinctively.
Think of it as a simple framework — Role, Task, Context, Format (RTCF). Cover these four elements and most AI tools will return something useful on the first try.
Role: tell the AI who it is
Starting with a role sets the tone for everything that follows. “Act as an experienced HR manager” or “You are a plain-English editor” tells the AI how to approach your request before it even reads the rest.
Task: be specific about what you want
This is the actual instruction. “Write a short email” is fine; “Write a two-paragraph email declining a meeting request, keeping the tone warm but firm” is much better. Specificity isn’t about over-engineering — it’s just being clear.
Context: tell it what it needs to know
Context is anything the AI couldn’t reasonably guess. Your audience, the situation, relevant background, any constraints. The more relevant context you include, the less the AI has to improvise.
Format: say how you want the output
Do you want a bulleted list, a table, a numbered plan, or a paragraph? Say so. If you need the response to be under 200 words, include that too. Anthropic’s prompt engineering guidance highlights output format as one of the most consistently useful elements to specify.
A quick example
Weak prompt: “Write a LinkedIn post about our new product.”
Stronger prompt: “You are a B2B copywriter. Write a LinkedIn post announcing a new project management tool for small teams. The tone should be conversational and confident. Keep it under 150 words. End with a question to encourage comments.”
Same task. Completely different output quality.
Ten prompting techniques worth knowing
You don’t need to memorise these. Try a few, see which ones fit your workflow, and build from there. These techniques work across all major AI tools.
Give it a role first. “Act as a financial analyst” or “You are a plain-English editor” sets the register immediately.
Specify the audience. “Explain this to someone with no accounting background” produces very different output than the same prompt without it.
Ask for a specific format. “Give me this as a numbered list” or “Format this as a table with three columns: task, owner, deadline.”
Set a word or length limit. “In under 100 words” stops the AI padding things out unnecessarily.
Give examples. Paste in a sample of writing you like and say “match this tone.” Examples are one of the fastest ways to steer output quality.
Ask it to think step by step. For anything analytical or complex, adding “think through this step by step” before the task tends to produce more accurate, reasoned responses.
Use positive instructions, not just negative ones. “Write in a warm, direct tone” is clearer than “don’t be too formal.”
Ask for alternatives. “Give me three versions of this” or “Suggest two different approaches” opens up options you might not have considered.
Follow up and iterate. If the first response is 80% there, say what needs fixing: “Make it shorter” or “The second paragraph needs to be more assertive.” Prompting is a conversation.
Use a prompt improver. If you’re not sure where to start, the ivee Prompt Improver Chrome extension rewrites your prompts automatically — paste in what you’ve got and it makes it sharper before you even hit send.
The OpenAI prompt engineering guide covers several of these same principles in detail if you want to go further. The fundamentals translate across tools.
Prompting across different AI tools
The core prompting principles above apply everywhere. But each major AI tool has its own character worth knowing.
ChatGPT
ChatGPT handles a wide range of tasks well — writing, summarising, coding, brainstorming. It responds well to step-by-step instructions and tends to produce longer, more structured answers by default. If you want brevity, ask for it.
Good for: drafting, ideation, Q&A, explaining complex concepts
Tip: use the system prompt or custom instructions feature to set a persistent role or style
Claude
Claude (made by Anthropic) is particularly strong at nuanced writing, long documents, and tasks requiring careful reasoning. It tends to be more cautious about making things up, which is helpful when accuracy matters. Claude also handles long documents and large pastes of text well.
Good for: analysis, long-form writing, editing, research summaries
Tip: Claude responds well to detailed context — don’t be afraid to give it background
Microsoft Copilot
Copilot is built into Microsoft 365 — Word, Excel, Outlook, Teams — which makes it uniquely useful in a work context. It can reference your actual documents and emails, which changes what’s possible. We covered the most useful Copilot forward-slash prompts in a recent post if you want a practical shortcut list.
Good for: work tasks inside Microsoft 365, summarising meetings, drafting in Outlook
Tip: use forward-slash prompts (type “/” in any Copilot chat) to access ready-made commands instantly
Whichever tool you use, the fundamentals stay the same. If you want to get up to speed across all three, the ivee AI Masterclass series covers prompting and practical AI use in a format built for people who don’t have a tech background.
FAQs: Writing prompts for AI
What is a prompt in AI?
A prompt is any input you give to an AI tool — a question, instruction, or request. The quality of your prompt directly affects the quality of the response. Better prompts get better answers.
How long should an AI prompt be?
There is no perfect length, but more context usually helps. A one-line prompt works for simple tasks. For complex requests — like writing a report or analysing data — a few sentences giving context, format, and tone will get you much better results.
What is prompt engineering?
Prompt engineering is the practice of crafting inputs to AI models to get the best possible outputs. In practice it just means being clear, specific, and giving the AI enough context to do its job well. No coding required.
Do different AI tools need different prompts?
Broadly the same principles apply across ChatGPT, Claude, and Copilot — be clear, give context, specify format. Each tool has its own strengths, but good prompting habits transfer between all of them.
Is there a tool to help me write better prompts?
Yes — ivee built a free Chrome extension called the Prompt Improver that rewrites your prompts to make them clearer and more effective. You can install it and use it in any AI tool in your browser.
Ready to build real AI skills?
Writing better prompts is one of the fastest ways to get more value from any AI tool — and it’s a skill that compounds. The better you get at briefing AI, the more useful every tool becomes.
If you want to go further, the ivee Core AI Programme covers prompting in depth alongside practical AI skills for everyday work tasks — no coding, no jargon. And if you’re specifically looking to improve your prompts right now, grab the free ivee Prompt Improver Chrome extension — it rewrites your prompts before you hit send.
Book onto one of our AI programmes at ivee and start learning today.
