Resources

Mastering Prompt Engineering for Nonprofits: A Comprehensive Guide to Using ChatGPT and Other AI Tools

by
Alfredo Ramirez

Whether we embrace it or hide from it, artificial intelligence (AI) tools like ChatGPT are increasingly essential in the day-to-day work of most professionals, especially in the nonprofit space. Well-known tools like DonorSearch and Blackbaud, as well as up-and-comers like Causewriter, are taking the world by storm. 

Understanding and effectively utilizing these tools begins with mastering the art of prompt engineering—the skill of crafting precise prompts to guide generative AI. This comprehensive guide dives deep into the elements of effective prompt engineering, ensuring your nonprofit can harness AI's full potential for various tasks, from drafting compelling grant proposals to personalizing donor communications. We also include a downloadable resource for all your prompt engineering needs!

Understanding the Basics

Before you dive into the simple intricacies of prompt engineering, it is essential to understand the language that underlies this skill. Here are some common terms you may have heard before and will come across in developing your prompt engineering skills:

  • Generative AI: This AI technology creates new content, including text, images, and videos. It's transformative for content creation, offering a way to generate fresh, engaging material quickly.
  • Language Learning Model (LLM): At the heart of ChatGPT and similar tools, LLMs are advanced AI systems trained on vast amounts of text data, enabling them to understand and generate human-like text.
  • Prompt: The instructions or questions you provide to AI. A well-crafted prompt results in more accurate and valuable AI-generated content.
  • Prompt Engineering: The practice of designing effective prompts to achieve specific outcomes from AI tools. It's a critical skill for anyone looking to leverage AI in their communications strategy.
  • Input and Output: The 'input' is what you feed into the AI (your prompt), and the 'output' is what the AI produces in response.

Download the Cheat Sheet


Crafting the Perfect Prompt

Any time you communicate with ChatGPT or any other AI tool, you create a prompt. This prompt can take many forms, such as a question (e.g., "What is a theory of change?") or a command (e.g., "Write an email to my boss").

Prompt engineering is the process of crafting text that can be accurately interpreted and understood by a generative AI model. In other words, it involves intentionally phrasing a query, specifying a style, providing relevant context, and assigning a role to the AI to achieve a specific outcome. A well-crafted prompt can even include examples for the model to learn from.

Prompt engineering applies not only to ChatGPT but also to most other AI tools. In fact, if you have recently used an AI-powered tool that claims to reply to emails or review text, chances are it's utilizing a sophisticated prompt that has been automated to deliver efficient and effective results.

A good prompt is clear, concise, and specific. It should clearly state what you want the AI to do, in what style or tone, and with what information. Here are key components to consider when crafting your prompt:

  1. Establish Intent: Clearly define what you want the AI to accomplish. Are you drafting an email, a blog post, or a grant proposal? The clearer the intent, the better the output.
  2. Choose the Right Tone: The tone of the AI-generated content should align with your nonprofit's voice. Whether formal, persuasive, or humorous, the tone will set the stage for the message, whether formal, persuasive, or humorous.
  3. Provide Context and Background: Offering detailed background information helps the AI understand the subject better, leading to more relevant and targeted content.
  4. Define the Desired Format and Style: Whether it's a newsletter, website copy, or social media post, specifying the format helps the AI tailor its output accordingly.
  5. Use Examples: Examples of past successful communications can guide the AI to replicate similar successes.
  6. Specify the Outcome: What's the goal? The desired outcome should be clear, whether it's engaging donors, informing supporters, or rallying for a cause.

Prompts can have different effects depending on the model you use. Models are different versions of the tools you are using, such as ChatGPT 3.5 or ChatGPT 4, as well as similar tools from other companies, such as Claude and Gemini. Because new models are emerging and existing ones are updated, your output can vary as the landscape changes.

Chain Prompting & Refining your Output

One of the benefits of working with high-capacity tools like ChatGPT is that you can have a conversation with the tool, providing feedback and updates to arrive at a better output. Here are some things you can do to improve the first iteration of your output:

  • Provide Feedback: Use follow-up prompts after the initial output to communicate with AI tools where they did well and where they need improvement. Like most tools, these get better with feedback, and yours will only improve your future efforts.
  • Be Specific About Certain Changes: If certain aspects of the output need improvement, be specific about what they are. You can isolate and ask for updates to specific sections, titles, or formats so they resemble your desired output.
  • Offer Additional Context or Background: If you haven’t already, share background information or examples of what you’d like to see. This can focus the output so that it relies more and more on the information you have shared.
  • Ask AI for Help: Tell the model to train and improve. One of my favorite prompts is to ask, “How would you update the prompt so you get it right on the first try next time.”

If you build multiple prompts on top of one another, you might engage in chain prompting.

Chain prompting takes the output of one prompt as the input for another. This approach can be particularly useful for asking a tool to follow a logical reasoning process or tackling complex problems. The process of chain prompting involves presenting an input question, followed by a series of intermediate steps that lead to the final answer. For instance, you might start by saying something like, "Let's think step by step" ahead of the first prompt.

Practical Applications in the Nonprofit Sector

AI has become ubiquitous and has the potential to enhance speed and efficiency in various domains. For nonprofits, ChatGPT is an excellent tool for improving their communication strategies efficiently. Here are some ways AI can be helpful in this context:

  • Drafting Grant Proposals and Reports: AI can help structure and draft compelling narratives for your funding applications.
  • Personalizing Communications: Tailor messages to donors and supporters, making email and SMS interactions or calls to action feel personal and genuine.
  • Translating Content: Break language barriers by translating communications into multiple languages, broadening your reach.
  • Creating Event Materials: From invitations to event plans, AI can streamline the creation process.
  • Formatting Engaging Newsletters: Keep your community informed with AI-assisted newsletters that resonate with your audience.
  • Inspire Design: Using tools like Dall-E-2 or Midjourney 

Our motto is to treat AI tools like ChatGPT as an intern: it will create good work and has high output capacity, but it doesn’t mean the output will be the finished product. More often than not, you will use it as a first draft or well of inspiration to arrive at your final draft.

Embracing AI in Nonprofit Communications

The journey into prompt engineering and AI utilization offers nonprofits unprecedented opportunities to enhance their communication strategies. By mastering the art of crafting effective prompts, your organization can unlock new potentials in efficiency, creativity, and engagement. As AI continues to evolve, so will how nonprofits can leverage this powerful tool to advance their missions and connect with communities worldwide.

Remember, the power of AI in your nonprofit communications strategy lies not just in the technology itself but in how creatively and effectively you can direct it through your prompts. Happy prompting!

ABOUT THE AUTHOR

Alfredo Ramirez

Alfredo Ramirez is the Cofounder and CMO of Prosal.

We also think you’ll like:

Resources

Write a Winning Proposal: Insider Tips to Master the RFP Process
Write a Winning Proposal: Insider Tips to Master the RFP Process

Resources

Prosal Partners & Resources
Alfredo Ramirez

Respondent Spotlight

Mixte: From Bikes to Digital Communications
Alfredo Ramirez

General

7 Problems with the RFP
Nick Lopez