How to Write Better AI Prompts and Get More Useful Results

Most people use AI tools the same way they use a search engine, a few words, minimal context, hoping for the best. Then they’re surprised when the result is generic, off-target, or missing the point entirely.

The AI isn’t the problem. The prompt is.

AI assistants like Claude, ChatGPT, and Gemini are capable of genuinely impressive work. But they’re not mind readers. The quality of what you get out is almost entirely determined by the quality of what you put in. Here’s how to put in better.

Be Specific About What You Actually Want

Vague prompts produce vague results. This is the most common mistake and the easiest to fix.

Compare these two prompts:

“Write me an email.”

“Write a professional email to a client who missed our 2pm meeting without notice. Keep it brief — two short paragraphs. The tone should be polite but make clear we need to reschedule this week. Sign off as Sarah from the Techwey team.”

The first prompt gives the AI almost nothing to work with. The second gives it a recipient, a situation, a length, a tone, a deadline, and a sign-off. The output from the second prompt will be dramatically more useful — and will require far less editing.

Before you write a prompt, ask yourself: what do I actually need? Who is this for? What tone, length, or format do I want? What shouldn’t it include? The more of those questions you answer in the prompt, the less work you’ll need to do after.

Give the AI a Role

AI tools perform better when you give them a context to operate within. Telling an AI to act as a specific type of expert — a senior editor, a financial analyst, a beginner-friendly teacher — shapes how it approaches the task and the language it uses.

“Explain machine learning to me” produces a decent explanation. “Explain machine learning to me as if I’m a non-technical business executive who needs to understand enough to evaluate vendor claims, not to build anything myself” produces something far more targeted and useful.

The role also sets expectations for tone and depth. “Act as a harsh editor reviewing this for clarity” will give you more direct, useful feedback than simply asking “can you improve this?”

Provide Context — Even When It Feels Obvious

AI tools don’t know your situation unless you tell them. They don’t know your industry, your audience, your constraints, or your previous work unless you provide that information in the prompt.

If you’re asking for help with a presentation, tell the AI who the audience is, what the goal is, and what you already know they know. If you’re asking for code, specify the language, the framework, and any constraints the solution needs to meet. If you’re asking for a marketing email, include the product name, the target customer, and the main benefit you want to highlight.

Context isn’t padding — it’s instruction. Every relevant detail you add reduces the gap between what you want and what the AI produces.

Ask for a Format

If you want bullet points, say so. If you want a numbered list, say so. If you want a table, a script, a step-by-step guide, or a single paragraph with no headers, say that too. AI tools will default to a format if you don’t specify one, and the default is often not what you actually need.

Format instructions also help with length. “Summarise this in three bullet points” produces something very different from “summarise this.” “Write a 150-word product description” is clearer than “write a short product description.” The more precise you are about format, the less you’ll need to edit the output.

Iterate Rather Than Restart

One of the most underused features of AI assistants is the ability to continue a conversation. Most people send a prompt, get a result, decide it’s not quite right, and start again from scratch. That throws away all the context the AI has already built up.

Instead, follow up. “Make it shorter.” “Change the tone to be more direct.” “Add a section on X.” “The third paragraph is too vague — can you be more specific?” Each of these builds on what already exists rather than starting over. This is faster, produces better results, and teaches you more about how the AI interprets your instructions.

Use Examples

If you have a specific style, tone, or format in mind, show the AI an example. “Write a product description in this style: [paste example]” is more effective than trying to describe the style in words. AI models are trained on vast amounts of text and can match styles, formats, and tones precisely when given a reference.

This is especially useful for writing tasks where “professional but approachable” or “technical but readable” might mean different things to different people. An example removes the ambiguity.

Tell It What to Avoid

Negative instructions are just as useful as positive ones. “Don’t use jargon.” “Avoid bullet points.” “Don’t mention price.” “Skip the introduction and start with the main point.” These guard rails prevent the AI from doing things that would make the output less useful — and they often produce cleaner results than just hoping the AI will make the right choices on its own.

The Bottom Line

Better prompts produce better results — consistently, across every AI tool. Specificity, context, role, format, iteration, examples, and constraints are the seven levers you can pull. You don’t need all of them every time. But the more deliberately you use them, the more you get out of every AI interaction.

The people who get the most value from AI tools aren’t the ones who use them most. They’re the ones who communicate with them most clearly. For more on how to choose the right AI tool for different tasks, see our guide to choosing the right AI tool for what you actually need.

Read more tech related articles here.