How to master AI prompt writing in the NFP sector
Generative AI can create text, video, images, and sound by using vast amounts of information. However, to help create these responses, we need to provide a guide for the AI to follow. That’s where prompt writing comes in. Think of prompt writing as your key to using generative AI. It’s hard to build a house without a blueprint to guide you, and likewise, it’s hard for generative AI to create without an effective prompt.
Prompt writing is the method by which AI generates content. By crafting specific instructions via text, you can guide an artificial intelligence model to generate responses. This can be in the form of questions, statements, or commands that provide clear and concise instructions. This helps the AI understand the context and deliver accurate, relevant, and coherent outputs. The more detail that is provided in the prompt, the more detailed and relevant of an answer you will receive.
Whether you're crafting a volunteer recruitment post, drafting a grant proposal, or summarising meeting notes, knowing how to write a good prompt can mean the difference between a frustrating AI experience and a game-changing productivity boost.
So, let’s break it down—what makes a good AI prompt, and how can you write one that delivers the results you actually need?
Why prompt writing matters
Think of AI as a well-intentioned intern. It’s smart, eager to help, and full of potential—but it can’t read your mind. If you give it vague or unclear instructions, you’ll likely end up with something generic, confusing, or just plain wrong.
For example, let’s say you want to recruit new volunteers for your community program. You could type:
Poor prompt: “Write a social media post to recruit volunteers.”
AI will generate something, but it might miss key details—who the volunteers are, what they'll be doing, and why they should care. Instead, a well-crafted prompt gives AI the right guidance:
Stronger prompt:
"Create a warm and engaging Facebook post encouraging local community members to volunteer for our food relief program. Highlight that volunteers will be assisting in packing and distributing food hampers on Saturdays at our community centre. Emphasise that no experience is needed, just a willingness to help. Keep it short and inspiring.”
Notice the difference? The improved prompt provides clear instructions, context, and a desired tone, which helps AI generate something far more useful.
The PROMPT framework: A simple guide for better results
To make AI work for you, we recommend using the PROMPT framework—a structured way to craft effective instructions.
P – Purpose definition
Start with clarity: What exactly do you need AI to do? Define the task in a straightforward way.
Example: Instead of “Write a report”, try “Summarise our latest volunteer impact report into three key takeaways for a donor newsletter.”
R – Role setting
Tell the AI who it should “be” when responding. This helps refine its style and expertise.
Example: “Act as an experienced volunteer coordinator and draft a heartfelt thank-you message for our long-term volunteers.”
O – Outcome specifications
Define the format or structure you need. Should the response be a list, a paragraph, or bullet points? Set word limits or exclusions.
Example: “Provide five short bullet points highlighting the benefits of volunteering with us.” “Keep it under 100 words and avoid using jargon.”
M – Modality & tone
Specify how you want it to sound—formal, friendly, persuasive?
Example: “Use an enthusiastic and conversational tone to invite supporters to our upcoming community event.”
P – Precise detail
If you require specific detail or information to be part of the outcome, ensure you include this in the prompt.
Example: “Ensure it is clear registrations close on Monday 4 February 2025.”
T – Testing & iterating
Always review the AI’s response. If it’s not quite right, tweak your prompt and try again.
Navigating common AI challenges
Even with well-written prompts, AI isn’t perfect. Here are some common pitfalls to watch out for:
AI hallucinations – AI sometimes generates incorrect or made-up information. Always fact-check responses.
Lack of context – If the AI's response seems off, try adding more background details.
Bias in outputs – AI reflects biases from its training data. Always review responses for unintended language or assumptions.
Ignoring instructions – If AI overlooks part of your request (like a word limit), gently rephrase: “Rewrite this in 100 words or fewer.”
Understanding these challenges will help you use AI effectively while staying critical of its outputs. For more on how to best navigate these common challenges, we have a self-paced learning course, CARE'ing about AI & ethics, available.
Your turn: Try it yourself!
Now it’s time to put this into practice. Open an AI tool (like ChatGPT or Microsoft Copilot) and choose a simple task—perhaps responding to an email that’s been sitting in your inbox.
Start with a rough prompt, then refine it using the PROMPT framework. See how small tweaks improve the AI’s response and keep iterating until it feels just right.
The more you practice, the more natural it becomes. And soon, AI will feel less like a confusing technology—and more like a helpful assistant supporting your not-for-profit’s mission.
Want to learn more?
You can find more tips, tricks and information on how to make AI work for you at our webinar catalogue. If you're looking to further your AI expertise, we recommend:
From prompts to picture-perfect creations: How to create stunning images from text
Research smarter, not harder: AI tools for writers & analysts
Status message
Thanks for rating this guide.