Learn how to write a prompt the right way and every AI tool suddenly feels smarter. The same question can give a lazy paragraph or a full-blown expert answer, depending on five small habits. Here is what beginners get wrong and how to fix it in two minutes.
The Gist
- A prompt is not a question. It is a briefing you give to an assistant
- The same five rules work across ChatGPT, Claude, Gemini and every AI tool
- Three beginner mistakes explain most of the disappointing answers you get
Have ChatGPT Recap This Article
ChatGPTA Prompt Is Not a Search Query
The first mental switch to make is simple: you are not googling anymore. When you type into Google, you throw two or three keywords and it returns a list. When you type into ChatGPT, Claude or Gemini, you are talking to an assistant that will actually do the work for you.
That means the game is different. Instead of guessing the right keywords, you tell the assistant who it is, what you want, and how you want the answer. The more clearly you brief it, the better the output. This is why the same tool can feel useless one day and brilliant the next. The tool did not change. The prompt did.
Think about how you would brief a new intern on their first task. You would not just say “write me something about marketing”. You would say what the project is, who the audience is, how long the piece should be, and what format you expect. Prompting is the same thing, done in one shot.
Once that mental model clicks, most of your daily frustrations with AI tools disappear. You stop asking short vague questions and start writing short structured briefs. Same time investment, completely different results.
The 5 Rules That Fix 90% of Bad Answers
Every good prompt has the same skeleton. You do not need to hit all five points every time, but the more you hit, the better. Think of these as ingredients you can mix depending on the task.
Rule one, give a role. Start by telling the AI who it should act as. “You are a senior marketing consultant” or “Act as an experienced English teacher explaining to a beginner”. A role narrows the tone, the vocabulary, and the depth of the answer instantly. Skip this and you get a generic middle-of-the-road response. This is exactly the trick that makes tools like Claude Tag inside Slack feel like a real teammate rather than a chatbot, because the tool has been briefed once on what it is.
Rule two, describe the context. Two or three sentences of context change everything. Who is the audience, what is the goal, what has already been done. If you are writing a birthday message for your grandmother, say so. If it is for a corporate LinkedIn post, say so. The exact same subject with two different contexts should produce two very different outputs, and the AI needs you to say which one.
Rule three, state the task in one line. Not “help me with X”, but “write a 200-word email that convinces my landlord to lower the rent”. Verb, output, target, constraint. A clean task line is the difference between an answer and a real deliverable.
Rule four, specify the format. Do you want a bulleted list, a table, a short paragraph, three options to pick from, a step-by-step tutorial? Say it. If you do not, the AI will pick a format for you, and it will not always match what you had in mind. This single line often saves the whole revision cycle.
Rule five, give one example when possible. Not always needed, but when you can paste a sample of what “good” looks like (“Here is a previous email that worked well, do the same style”), the AI will match that reference much more closely than any adjective you could throw at it. One good example beats five adjectives every time.
Keep learning on AI Noobies:
- ChatGPT Hallucinations: How to Spot Made-Up Facts
- Gemini Spark Mac Now Does Your Tasks for You
- 5 Free AI Tools Every Beginner Should Try in 2026
The 3 Beginner Mistakes That Kill Your Results
The biggest gains come less from doing the five rules perfectly and more from stopping three habits that quietly ruin your answers. If your prompts underperform, one of these three is usually the culprit.
Mistake one, asking too many things at once. A single prompt that says “write me an email, translate it into Spanish, then explain the reasoning behind each sentence” is a prompt destined to produce a mediocre version of all three. Give one prompt one job. If you need multiple things, chain them one after the other in the same conversation. The AI keeps the context and each output stays sharp.
Mistake two, giving up after one try. The first draft from any AI tool is a starting point, not a final answer. Real users iterate. “Make the tone warmer.” “Cut the second paragraph in half.” “Rewrite it as if you were talking to a friend.” Every follow-up sharpens the output without you having to rewrite the whole prompt. Users who get magical results are usually the ones who do three or four rounds of polishing, not the ones who write the perfect first prompt.
Mistake three, trusting invisible information. Every AI tool will happily invent a source, a date, a statistic or a person that does not exist. It is called a hallucination and it happens more often than most beginners realize. If a fact matters (a legal reference, a medical claim, a financial number), verify it in a real source before acting on it. Use AI to draft, use humans and official sites to fact-check. The gap between confident-sounding output and reliability is exactly why only 10% of Americans currently use an AI chatbot for news, despite the daily push around these tools.
Put the five rules in your notes app, keep them within reach for a couple of weeks, and the habit forms itself. Very quickly, briefing an AI feels as natural as briefing a human colleague, and the difference in the results is the difference between a tool you use once a week and a tool you cannot imagine working without.
Stay tuned on AI Noobies.
About the Author
