10 Tips for Writing the Perfect Prompt
The quality of an AI response depends directly on the quality of your instructions. Here are 10 essential tips for guiding AI like a pro.
Here are 10 essential tips for guiding AI like a pro.
Artificial intelligence can write articles, analyze information, generate ideas, solve problems, create images, and assist with complex tasks. But even the most capable AI model can produce disappointing results when it receives unclear or incomplete instructions.
The difference between a generic response and an exceptionally useful result often comes down to one thing: the prompt.
A prompt is the instruction, question, or information you give an AI system to guide its response. It tells the AI what you want, why you want it, and how you expect the result to look.
Writing a good prompt doesn't require programming skills or technical knowledge. It requires clear thinking, precise instructions, and an understanding of how to communicate your expectations.
Whether you're using ChatGPT, Claude, Gemini, or another AI assistant, these 10 practical tips will help you write better prompts, get more relevant answers, and make AI work more effectively.
1. Be Clear and Specific About What You Want
The first rule of prompt writing is simple: don't make the AI guess what you mean.
Broad or vague instructions leave too much room for interpretation. When you ask an AI to do something without explaining exactly what you expect, it must make assumptions about your objective, audience, format, and desired outcome.
The more precise your instructions, the easier it is for the AI to produce a relevant response.
Weak prompt:
"Write something about artificial intelligence."
Better prompt:
"Write a 1,000-word beginner-friendly article explaining how artificial intelligence works. Cover machine learning, neural networks, and real-world applications. Use practical examples, clear headings, and a professional but accessible tone."
The second prompt gives the AI a clear subject, length, audience, structure, and writing style.
Pro tip: Before submitting a prompt, ask yourself whether someone unfamiliar with your intentions could understand exactly what you want. If not, add the missing details.
2. Give the AI Context and Background Information
A good instruction tells the AI what to do. A great instruction also explains the context in which it should do it.
Context helps AI understand the purpose behind your request and make decisions that are more relevant to your situation.
For example, asking for marketing advice without explaining your business will usually produce generic recommendations. Providing information about your industry, customers, product, and objectives gives the AI a much stronger foundation.
Weak prompt:
"Create a marketing strategy for my website."
Better prompt:
"I run an online store selling premium coffee equipment to home coffee enthusiasts aged 25–45. My main products are espresso machines and coffee grinders. I want to increase organic traffic and convert more visitors into customers. Create a six-month content marketing strategy focused on SEO, buying guides, product comparisons, and educational articles."
Notice how the additional information changes the task. The AI now has a business, audience, product category, and measurable objective to work with.
Pro tip: Include the background information that could change the answer. Relevant context improves results; unnecessary details can make prompts harder to follow.
3. Assign the AI a Relevant Role or Expertise
One effective way to guide an AI response is to define the professional perspective it should adopt.
Instead of asking a generic question, tell the AI what type of expertise or approach you expect.
For example, you can ask it to approach a task as a:
- Senior software engineer.
- Experienced SEO strategist.
- Professional copywriter.
- Financial analyst.
- UX/UI designer.
- Academic researcher.
- Business consultant.
Weak prompt:
"Review my website."
Better prompt:
"Act as a senior UX/UI designer and SEO strategist with extensive experience in e-commerce websites. Review my homepage from the perspective of usability, visual hierarchy, navigation, conversion optimization, and search visibility. Identify specific problems and provide practical recommendations for improvement."
A role helps establish the perspective, vocabulary, and priorities relevant to the task.
However, assigning a role does not automatically give an AI professional experience or guarantee expertise. The quality of its analysis still depends on the information, instructions, and evidence available.
Pro tip: Choose a role that matches the actual task. Avoid stacking unrelated professional titles into one prompt.
4. Define the Desired Output Format
Even when an AI understands your request, the response may not be presented in the format you need.
Do you want a detailed article, a comparison table, a checklist, a business plan, a step-by-step tutorial, or a short summary?
Tell the AI explicitly.
Weak prompt:
"Compare ChatGPT, Claude, and Gemini."
Better prompt:
"Compare ChatGPT, Claude, and Gemini in a table. Include their main strengths, weaknesses, context windows, multimodal capabilities, pricing considerations, and best use cases. Follow the table with a short recommendation for students, developers, and business professionals."
You can also specify:
- Word count or response length.
- Number of sections.
- Heading structure.
- Bullet points or numbered lists.
- Tables and comparison criteria.
- Markdown formatting.
- Whether examples are required.
- Whether sources and references should be included.
Pro tip: Describe what the finished result should look like, not just what information it should contain.
5. Provide Examples of What You Expect
Sometimes explaining what you want is harder than showing it.
Examples give AI a concrete reference for the style, structure, quality, or level of detail you expect.
This technique is particularly useful for writing, programming, branding, content creation, and repetitive tasks.
Weak prompt:
"Write product descriptions in a professional style."
Better prompt:
"Write product descriptions following this example:
'Designed for everyday performance, this lightweight wireless keyboard combines comfortable typing, reliable connectivity, and a compact design for a more productive workspace.'
Use a similar tone: clear, benefit-focused, professional, and free from exaggerated marketing claims. Each description should be between 80 and 100 words."
You can provide examples of:
- Good and bad answers.
- Preferred writing styles.
- Existing documents.
- Brand guidelines.
- Formatting conventions.
- Sample code.
- Reference images.
When you provide examples, explain which characteristics the AI should reproduce. Otherwise, it may copy irrelevant details.
Pro tip: One or two carefully selected examples are often more useful than a long collection of unrelated references.
6. Break Complex Tasks into Smaller Steps
A complicated request can overwhelm a prompt when too many independent objectives are combined without a clear structure.
Instead of asking AI to complete everything in one loosely defined instruction, divide the task into logical stages.
For example, building a complete website involves research, planning, design, development, testing, optimization, and deployment.
A structured prompt might ask the AI to:
- Analyze the target audience and business requirements.
- Define the website structure and content hierarchy.
- Establish the visual design system.
- Design the key pages and user journeys.
- Implement the website.
- Test functionality and responsiveness.
- Review SEO and performance.
- Prepare the final deployment.
Breaking down complex work makes it easier to inspect intermediate results, identify mistakes, and correct the direction before moving forward.
For particularly important projects, ask the AI to complete and verify each stage before proceeding.
Pro tip: Use one prompt for a focused task. For larger projects, establish the overall objective first and then work through clearly defined stages.
7. Set Clear Constraints and Boundaries
A good prompt explains not only what the AI should do but also what it should avoid.
Constraints prevent unwanted assumptions, unnecessary additions, and outputs that don't meet your requirements.
For example, you might specify:
- Do not exceed 500 words.
- Use only the information provided.
- Do not invent statistics or references.
- Avoid unnecessary technical jargon.
- Do not change the existing code architecture.
- Do not include promotional language.
- Do not make claims without supporting evidence.
Weak prompt:
"Improve this article."
Better prompt:
"Improve the clarity, structure, and readability of the article below. Preserve its original meaning, factual claims, and important information. Do not introduce unsupported statistics, invent sources, or add promotional language. Keep the final article under 1,500 words."
Constraints are especially important when working with research, technical documentation, financial information, and other tasks where accuracy matters.
Pro tip: Clearly distinguish between mandatory requirements, preferences, and things the AI must never do.
8. Ask the AI to Clarify Missing Information
You don't always need to have every detail figured out before writing a prompt.
If your request is missing important information, you can instruct the AI to ask questions before starting.
This is particularly useful for complex or high-impact tasks where assumptions could lead to the wrong result.
For example:
"Help me create a content strategy for my business. Before developing the strategy, ask me up to five essential questions about my industry, target audience, products, competitors, and business objectives. Wait for my answers before creating the final strategy."
This approach makes the process collaborative rather than forcing the AI to fill gaps with assumptions.
You can also ask the AI to identify unclear requirements, point out contradictions, or explain what additional information would improve its answer.
Pro tip: For simple questions, provide the necessary information and get straight to the answer. Reserve clarification steps for tasks where missing information materially affects the result.
9. Iterate, Test, and Refine Your Prompts
Your first prompt doesn't have to be your final prompt.
Prompt writing is an iterative process. Even a well-structured instruction may produce an answer that needs adjustments.
Instead of starting over every time, use follow-up instructions to improve specific aspects of the result.
For example:
- "Make the explanation more practical and less theoretical."
- "Add three real-world examples."
- "Remove repetitive information."
- "Explain the technical concepts for beginners."
- "Strengthen the introduction without changing the article's meaning."
- "Identify unsupported claims and distinguish facts from assumptions."
You can also ask the AI to evaluate its response against your original requirements.
For example:
"Review your previous answer against my original instructions. Identify missing requirements, unsupported claims, and areas that need improvement. Then provide a corrected final version."
This creates a feedback loop in which each interaction helps move the result closer to your expectations.
Pro tip: Give specific feedback. "Make it better" is less useful than identifying exactly what needs to change.
10. Treat Prompt Writing as a Skill You Can Improve
The best prompt writers don't rely on complicated formulas or magical words. They understand how to communicate objectives, provide relevant information, and evaluate results.
Over time, you'll discover which instructions work best for your tasks and which approaches produce inconsistent results.
A useful habit is to save your most effective prompts and reuse them as starting points.
Build a personal prompt library organized around your common activities:
- Writing and editing.
- Research and analysis.
- Marketing and SEO.
- Programming and development.
- Business planning.
- Education and learning.
- Image generation.
- Productivity and automation.
You can refine these prompts as your requirements evolve and as AI tools introduce new capabilities.
Remember that different AI models may interpret the same instruction differently. A prompt that works exceptionally well with one model may need adjustments when used with another.
Pro tip: Don't just collect prompts. Record what each prompt is designed to accomplish, what information it requires, and what a successful result looks like.
Bonus: A Simple Framework for Writing Better Prompts
The ten tips above can be combined into a practical structure that works for many everyday AI tasks.
A useful prompt generally includes five elements:
| Element | What to include |
|---|---|
| Role | What expertise or perspective should the AI use? |
| Context | What background information does it need? |
| Task | What exactly should it accomplish? |
| Constraints | What requirements and limitations must it follow? |
| Output | What should the final result look like? |
The perfect prompt template
Use the following template as a starting point and adapt it to your needs.
ROLE: Act as [relevant professional role or expertise].
CONTEXT: Here is the background information you need: [provide relevant details].
TASK: Your objective is to [describe exactly what you want accomplished].
CONSTRAINTS: Follow these requirements: [list limitations, standards, and things to avoid].
OUTPUT: Deliver the result as [specify format, structure, tone, and length].
QUALITY STANDARD: The final result should be [describe the qualities that define success].
Example: Turning the framework into a real prompt
Imagine you want to create an educational article about artificial intelligence.
ROLE: Act as an experienced technology journalist and educational content writer.
CONTEXT: The article is for a website that explains artificial intelligence to beginners and technology enthusiasts.
TASK: Write a comprehensive guide explaining how large language models work.
CONSTRAINTS: Use technically accurate information, avoid unsupported claims, explain complex concepts in accessible language, and include practical examples.
OUTPUT: Create a structured article with an introduction, descriptive H2 and H3 headings, examples, a comparison table, and a frequently asked questions section. Target approximately 2,000 words.
QUALITY STANDARD: The article should be informative, engaging, logically structured, factually responsible, and useful to readers without a technical background.
The result is a much more complete instruction than simply asking an AI to explain large language models.
Final Thoughts: Better Prompts, Better AI Results
Artificial intelligence is only as useful as the way you direct it.
A vague instruction often produces a generic response. A clear, contextual, and well-structured prompt gives AI a much better opportunity to deliver something genuinely useful.
The most important principles are straightforward:
- Be specific about your objective.
- Provide relevant context.
- Define the expertise you need.
- Specify the output format.
- Include examples when useful.
- Break complex tasks into manageable steps.
- Establish clear constraints.
- Clarify missing information.
- Refine the results through feedback.
- Save and improve the prompts that work.
You don't need to become a programmer or learn a complicated technical language to communicate effectively with AI. You need to know what you want, explain it clearly, and evaluate what you receive.
The real skill isn't writing longer prompts. It's writing better instructions.
Start with a clear objective, provide the right context, and give AI the direction it needs. The difference in the results can be substantial.
Remember: Don't just ask AI a question. Give it a clear direction.
Prompts that put this into practice
Marketing Strategy Prompt
Turns your business facts into a focused strategy: a diagnosis, one priority segment, positioning, three channels at most, and a phased plan sized to your budget.
# Role
You are a marketing strategist who works with small teams and limited budgets. You choose a few things to do well instead of listing every available chan…Blog Post Writing Prompt
Turns your own notes and experience into a focused blog post with one clear argument, and marks every spot where it needs a real example or a fact check.
# Role
You are a blog writer and editor. You write for one specific reader, make one point per post, and would rather publish 600 useful words than 1,500 padded…Complete Website Design Prompt
A structured brief that walks an AI model from business goals to sitemap, page layouts, visual direction and responsive rules.
# Role
You are a senior web designer and UX strategist. You make decisions from user needs and business goals first, visual style second.
# Project
- Website t…Editing and Proofreading Prompt
Edits a draft at the level you choose, protects the author's voice, and returns a change log plus queries instead of silent rewrites.
# Role
You are a copy editor working on someone else's writing. Your job is to make the draft cleaner and clearer while it still sounds like the person who wrot…AI App Builder Prompt Template
A reusable template for any app: app type, audience, key features, design and edge cases.
Build a [App type] called “[App Name]” for [Describe your target user in one sentence]. The primary goal is for users to [Core action or workflow].
Key feature…Problem Solving
Help me break it down, identify root causes, and suggest practical solutions ranked by impact and ease.
I’m facing this work problem: [Describe the problem]
Help me break it down, identify root causes, and suggest practical solutions ranked by impact and ease.Keep reading
Guide
How to Write a Good AI Prompt
The five parts of a complete prompt (goal, context, task, constraints, output format), with a worked rewrite and a pre-send checklist.
Guide
What Is Prompt Engineering? Core Techniques Explained
What prompt engineering is, the techniques the major model vendors agree on, the working loop behind them, and what prompting cannot fix.
Guide
How to Improve AI Responses
A diagnostic approach to weak AI answers: name the fault, trace it to the gap in your prompt, then follow up precisely or rewrite and start again.
Technical guide
Few-Shot Prompting: How to Use Examples in a Prompt
How to include worked examples in a prompt so the model follows a format, label set or style, and how to choose examples that generalize.
Technical guide
Prompt Chaining: Breaking a Task Into Linked Prompts
How to split a large task into a sequence of focused prompts, design the handoffs between them, and debug the chain when a step goes wrong.
Guide
Prompt Templates vs Prompts: What Is the Difference?
A prompt is one complete request; a template is a reusable prompt with named variables. When to use each and how to turn one into the other.
Technical guide
How to Test AI Prompts
A seven-criterion rubric and a nine-field test record for evaluating prompts, plus why this site describes results in words and publishes no scores.