What is a prompt? A practical guide to better AI results

Most people who feel let down by ChatGPT or Claude aren't using the wrong model. They're giving the wrong instruction. They type "write a proposal" and get back a bland, generic text, then conclude that AI is overhyped. But the model did exactly what was asked: something generic, because nothing specific was said.

The instruction you give to an AI model is called a prompt. And the difference between a mediocre and a useful result almost always sits in that prompt, not in the model. The good news: prompting is a skill, not a talent. You can learn it in an afternoon.

What is a prompt?

A prompt is the text with which you tell an AI model what you want. That can be a question ("what's the difference between revenue and profit"), an instruction ("summarise this email in three sentences"), or a complete work brief with details like background and desired formatting.

The word "prompt" means to cue or nudge into action. That captures it: a prompt is the starting signal that determines what the model does. An AI model has no idea what's in your head. It only has the words you type. The more relevant information sits in there, the closer the answer lands to what you actually meant.

Many people treat a prompt like a Google search: short, a few keywords, then see what comes out. That's exactly the wrong reflex. Why that is, and how to have a conversation with AI instead, I wrote about in why you need to be patient with AI. This guide is about the structure: what a good prompt actually looks like.

Prompting is a skill, not a talent. You can learn it in an afternoon.

The anatomy of a good prompt

A strong prompt almost always contains the same building blocks. Not every prompt needs all of them, but the more important the task, the more of them you bring along.

Context. The background. Who is this for, what's the situation, what's already happened. "I run an installation company with eight engineers" gives the model a completely different starting point than nothing at all.

Role. Who should the AI be? "Act as an experienced copywriter" or "you're an accountant explaining this to me" steers the tone and level of detail surprisingly strongly.

Task. What exactly needs to happen, in verbs. Not "something about marketing" but "write three subject lines for an email to existing customers".

Format. How should the answer look? For example: a table, a list, or a short text of at most a hundred words. Say nothing and the model picks for you, usually not what you wanted.

Examples. One good example often does more than three paragraphs of explanation. Paste an earlier proposal you liked and say "in this style".

Two frameworks that give you traction

If you don't want to think through those building blocks every single time, use a fixed template. Two that work in practice:

CRAFT. Context, Role, Action, Format, Test. You describe the situation, give the AI a role, say what it should do, in what shape, and against which criteria the result should hold up. Handy for one-off jobs.

CRISP. Context, Role, Instructions, Specifications, Persona. Especially useful for recurring business tasks where you want the output to feel the same every time. You lock in who the AI is and how it should sound, so you don't have to type it out again.

Which template you pick matters less than the fact that you use one. It forces you to give the information the model needs, instead of hoping it will read your mind.

Five examples: before and after

Nothing shows the difference better than the same task, done twice.

Replying to a customer email. Before: "Write a reply to this complaint." After: "A customer is complaining that their order is three days late. Write a reply of at most a hundred words, friendly but without overdone apologies, offer a concrete solution, and close with a question to check whether that works for them. Write in English, informal tone."

A job ad. Before: "Create a job ad for an engineer." After: "Write a job ad for a maintenance engineer at an installation company in Zeeland. Target audience: vocationally trained, 25 to 40 years old, values a stable team and short lines of communication over a high salary. At most three hundred words, no clichés like 'dynamic environment', close with how to apply."

Explaining numbers. Before: "Explain these numbers." After: "Here's my monthly revenue over a year. Summarise in three sentences what stands out, name the strongest and weakest month, and give one possible explanation I could look into myself. Don't present assumptions as fact."

Developing a blog idea. Before: "Write a blog post about AI." After: "Write an outline for an eight-hundred-word blog post for SME owners who don't work with AI yet. Topic: where they should start. Practical, no hype, with one concrete first step. Give me only the headings and one sentence per heading."

Translating while keeping the tone. Before: "Translate this into English." After: "Translate this product description into British English. Keep the informal, enthusiastic tone, but avoid literal translations of Dutch expressions. Target audience: British webshop customers."

The pattern is the same every time: you replace "do something" with "do this, for this person, in this shape, and watch out for that".

A five-word prompt gives you a five-word-quality answer.

The mistakes almost everyone makes

Too short. By far the most common. A five-word prompt gives you a five-word-quality answer.

Wanting everything in one go. A prompt that asks for multiple outputs like a strategy and copy usually delivers everything at half strength. Break it into steps.

Not saying what you don't want. "No jargon", "no longer than a hundred words", "no made-up numbers": the negative instructions are often just as important as the positive ones.

Not following up. The first answer is a draft, not a finished product. "Make it shorter", "more businesslike", "give me three variants": that's part of the job.

Not giving an example when you actually have one. If you have a good example, show it.

How do you save and reuse good prompts?

A prompt that works well once is worth its weight in gold, provided you can find it again. A lot of time savings get lost because people start from scratch every time. So keep a simple file or note with your best prompts for recurring tasks like proposals or summaries. Tools like ChatGPT and Claude also have built-in features for this, like custom instructions or saved projects, so you don't have to retype your context each time. Which tool suits you best depends on your work; I put the options side by side in ChatGPT alternatives for SMEs.

2026: from prompt engineering to context engineering

An important shift worth knowing about. Models have become a lot better over the past few years at understanding what you mean, even when your phrasing is imperfect. At the same time, it's clear that the biggest gains come from the context the model has, not the exact phrasing of the question.

That's called context engineering. Instead of polishing the perfect sentence, you make sure the AI has the right background at hand: your earlier documents, your house style, examples of good work, the agreements inside your business. For SMEs this means, in practice: the more relevant material you bring in, the less it matters whether your prompt is grammatically perfect. The question shifts from "how do I phrase this cleverly" to "what does the AI need to know to help me well".

How do you get better at this?

By doing it. Take a task you had to do this week anyway, write a prompt for it using the building blocks above, and improve the result in two or three rounds instead of settling for the first attempt. After a week or two it starts to feel obvious.

If you want your team to get good at this on your own actual work, with examples from your own practice, there's AI training for that. And for the mindset behind good prompting, the patience to have a conversation instead of firing off one instruction, read why you need to be patient with AI. And if you want to know first where AI can save you the most time, the free AI scan gives a first analysis based on your own situation.

A good prompt is simply a clear description of what you want, for someone who can't read your mind. That's a skill that comes in handy outside AI too.