Prompt Engineering Cheat Sheet - AI Prompt Engineering in Microsoft 365

Oct 14, 2026

Prompt Engineering Cheat Sheet introduces AI prompt engineering in Microsoft 365, explaining how to craft clear, specific prompts to improve Copilot output. It frames successful AI adoption through user value and a structured communication style, emphasizing Microsoft’s guidance: goal, context, source, and expectations. The guide then surveys common prompt frameworks such as 5W1H, Instructional, Problem-Solution, Contextual, Comparison, Action-Oriented, and RISEN, showing how each supports different business and IT use cases.

LanguageEnglish

CategoryPresentations

FormatPDF

Pages6

Reading Time2 min

AccessFree to download

Prompt Engineering Cheat Sheet

AI Prompt Engineering in Microsoft 365

Microsoft’s prompt guidance: goal, context, source, expectation

Other prompt engineering frameworks

5W1H Framework

Instructional Framework

Problem-Solution Framework

Contextual Framework

Comparison Framework

Action-Oriented Framework

RISEN Framework

PROMPT ENGINEERING CHEAT SHEET AI Prompt Engineering in Microsoft 365 By now, it’s very likely that you’ve begun to adopt AI tools across your organisation. Whether it’s using Copilot in M365 to efficiently manage your workloads, leveraging ChatGPT for content creation, or deploying AI-driven cyber security to detect threats in real time, all sorts of businesses are harnessing AI to work smarter and more efficiently. Of course, successful implementation of AI – just as with any new tool – hinges on user adoption; and this begins when employees see real value in their interactions with it. Throughout this guide, we’ll explore how to maximise the value of AI, with a particular focus on learning how to craft effective prompts to make the most of Microsoft Copilot (and other types of AI) in your daily workflows. Prompt Engineering-M365 Copilot Prompt Engineering in Microsoft 365 Copilot refers to the skill of crafting clear, specific, and effective prompts to get the best results from apps like Word, Excel, Outlook, and Teams. Since Copilot generates responses based on company data and user input, well-structured prompts can significantly improve accuracy, efficiency, and productivity. Remember, although Copilot and other AI tools are trained on human language, they are still not human, and they require us to communicate in a specific format. Therefore, it might help to think of prompt engineering as ‘learning to speak the language of AI’. Microsoft’s guidance You’ll be glad to know that Microsoft haven’t left us high and dry when it comes to prompt engineering and the use of Copilot for M365 . Here, you can access some helpful guidance on how to devise effective prompts from Microsoft itself. In short, however, Microsoft recommends including four elements: goal, context, source, and expectation. 1-Goal Define what you want from Copilot, your objective. 2-Context Explain why you need this information, who is involved, or how you’ll use it. I want to know what our best-selling product is this month, so I can offer a promotion on this item to individuals on our mailing list. Use information from my Sales folder and our inventory data to find this information, also include the top 4 next best-selling products 3-Source Identify data sources or plugins Copilot should use. 4-Expectations Define how Copilot should respond; the tone, audience, length of information, Etc. We can see that, in Microsoft’s framework, including detailed information across four steps means that Copilot will usually retrieve a comprehensive and precise answer. Of course, adding or modifying parameters, clarifying ambiguous instructions, or providing more examples can help guide the AI toward a better understanding of what is required if you don’t receive exactly what you need in the first attempt. What other prompt engineering frameworks are there? There are multiple AI prompt frameworks to make use of because different use-cases require different approaches. AI models, including Microsoft Copilot and other LLMs like ChatGPT, generate responses based on how the prompt is framed. By using specific frameworks, users can structure and optimise AI interactions for their specific business use, gaining accuracy, clarity, and efficiency. Some prompts, for example, ask for step-by-step instructions, while others need problem-solving guidance or side-by-side comparisons. AI can also provide different answers based on role, instruction, or situation (as seen in the RISEN framework below) . Having access to multiple frameworks allows users to tailor their prompts based on their goals, ensuring that AI delivers the most useful output for different business scenarios. Here are a few commonly used prompt frameworks: 5W1H Framework The 5W1H framework is a simple yet powerful method for structuring AI prompts. It helps ensure clarity, completeness, and relevance by focusing on six fundamental questions: Who, What, When, Where, W

FAQs

What does prompt engineering in Microsoft 365 Copilot mean?

It is the skill of crafting clear, specific, effective prompts to get the best results from apps like Word, Excel, Outlook, and Teams. Well-structured prompts improve accuracy, efficiency, and productivity.

What are the four elements Microsoft recommends for effective prompts?

Microsoft recommends including goal, context, source, and expectations. This structured detail helps Copilot retrieve comprehensive and precise answers.

Which prompt frameworks are commonly used besides Microsoft’s guidance?

Common frameworks include 5W1H, Instructional, Problem-Solution, Contextual, Comparison, and Action-Oriented frameworks, plus the RISEN framework. Each supports different business scenarios and interaction styles.

Oct 14, 2026

Upload by วิน