Tips for writing effective prompts for research
Creating effective prompts is both an art and a science. Here are the top 7 prompt-writing guidelines, developed through this hands-on, iterative approach, with easy-to-follow examples and expanded explanations.
1. Be Clear and Specific
A good prompt is not generic. It clearly states what specific output is expected, including the desired length, tone, and focus. Think about the exact information you need from the AI.
✅ Good Example: 👉 “Summarize this 500-word article for a 10-year-old in 3 bullet points, focusing on the main characters and their motivations.”
🚫 Bad Example: 👉 “Summarize this article.”
The AI models perform significantly better and provides more relevant results when it understands your exact needs and the specific context of your request. Being precise minimizes ambiguity and reduces the need for follow-up prompts.
2. Include Context When Necessary
Include any background information, relevant details, or surrounding context that could influence the AI’s answer. This helps the AI understand the specific situation and tailor its response accordingly. Consider the audience, purpose, and desired outcome.
✅ Good Example: 👉 “Translate this informal email written in English into professional French for a key business client who values formal communication. Ensure the tone is respectful and uses appropriate business terminology.”
🚫 Bad Example: 👉 “Translate this into French.”
AI can tailor its responses much more effectively when it understands the specific context and nuances of your request. Providing context ensures the output is relevant, accurate, and aligned with your goals.
3. Set the Format or Structure You Want
“If you want bullet points, say so. Don’t expect the AI to guess. What about tables, numbered lists, code blocks, or specific paragraph structures?”
Always explicitly define the desired format or structure of the AI’s output. Specify whether you need a list, table, paragraph, code snippet, JSON format, etc.
✅ Good Example: 👉 “List 5 pros and cons of electric cars in a table format with two columns: ‘Advantage’ and ‘Disadvantage’.”
🚫 Bad Example: 👉 “What are pros and cons of electric cars?”
Clearly defining the format saves you time by receiving the information in a usable structure from the outset. It avoids the need for manual reformatting and ensures the AI presents the information in a way that meets your needs.
To learn more about the fundamental principles for responsible AI development and deployment, see our article on 7 Essential Principles for Ethical AI: A Guide for Everyone.
Additionally, for a beginner-friendly introduction, the “AI Prompts: Essential Guide with Types & Best Practices“ by GetGuru offers a solid foundation. It covers the basics of AI prompts, different types, and optimization strategies to achieve the best possible results.
4. Use Role Play or Persona-Based Prompts
Assign a role or point of view to guide tone and expertise.
✅ Good Example:
👉 “Act as a nutritionist and explain the benefits of a low-carb diet to a beginner.”
🚫 Bad Example:
👉 “Explain the benefits of a low-carb diet.”
Roles give the AI a frame of reference for the response.
5. Give Examples
“If you’re looking for a certain style, tone, or even logical reasoning, showing the AI an example is incredibly helpful.”
Including a clear example of the desired output style, tone, or reasoning pattern can significantly improve the AI’s ability to match your expectations.
✅ Good Example: 👉 “Write a short product description like this: ‘Soft, cozy, and perfect for winter nights. Crafted with premium wool for ultimate comfort.’ Now describe a lightweight summer T-shirt targeting young adults.”
🚫 Bad Example: 👉 “Write a product description.”
Examples act as a powerful visual or textual guide for the AI, reducing ambiguity and increasing the likelihood of receiving output that aligns with your desired style and quality.
6. Limit the Scope to Avoid Overload
“Don’t ask the AI to do too many distinct things in a single prompt. It can lead to less focused and potentially inaccurate results.”
Break down complex tasks into smaller, more manageable prompts, each focusing on a specific aspect of the overall goal.
✅ Good Example: 👉 “Step 1: List the three main arguments presented in this research paper. Step 2: Based on those arguments, write a one-paragraph summary suitable for a social media post.”
🚫 Bad Example: 👉 “Write a social media post summarizing this research paper, highlighting its key arguments and potential impact.”
7. Test, Review, and Refine
Always test your prompt. If the result isn’t quite right, tweak it.
✅ Good Strategy:
👉 Run your prompt, assess the output, then clarify, simplify, or reformat if needed.
The best prompts evolve through trial and error.
Focused prompts allow the AI to concentrate on individual tasks, leading to more accurate, detailed, and ultimately more usable results. Breaking down complex requests improves the quality and coherence of the AI’s output.
Mastering the Art of AI Prompting
Prompt engineering is becoming an increasingly vital skill for a wide range of professionals, from marketers and writers to developers and business analysts. By applying these seven expert-backed guidelines, you can significantly enhance your ability to communicate effectively with AI models. Remember to start simple, be clear and specific, provide necessary context, and iteratively refine your prompts for optimal results.
By implementing these rules, you’ll unlock the full potential of powerful AI tools like ChatGPT, Gemini, Claude, and others—achieving better outcomes, faster.
Stay updated on the latest AI prompting research and best practices by following resources like Ultimate Guide to Writing Effective AI Prompts.
