Leveraging AI Technology at Your Nonprofit: A Starter Guide
This article was written primarily by Adam Wolford with the assistance of ChatGPT for copyediting and the images.
Introduction
Lions and tigers and bears, AI!
In the world today, there is a never-ending slew of conversations around technology and leveraging AI. This is even more prevalent in the nonprofit space, where leveraging these tools can help to alleviate the administrative burden that exists for every member of the staff and act as a thought partner for nonprofit leaders when they feel there is no one to turn to.
However, staring at a blank chat screen can be…well…scary! Like Dorothy in the Wizard of Oz, it can feel like walking through a dark forest.
This image was generated using ChatGPT
Fear not! This article is meant to serve as a light for you to not only understand where you might use AI, but how you can leverage it in your nonprofit today. I know how busy you are; this is meant to be practical and get you started.
The Set Up: Picking Your Tools
There is a lot of talk about all the different models and tools that exist in the world. While it is helpful to stay up to speed on the latest, it can be overwhelming to know where to get started and how. My advice is to assess the tools on three categories:
Specific to Nonprofits: there are a LOT of tools out there that tote benefits such as increasing responses to solicitation, helping with marketing, and supporting your other operations such as HR and legal. However, many of these models are trained on and built for the for-profit sector. While this does not rule them out completely, you need to make sure you contextualize responses for the nonprofit sector, or find a tool specific to your sector!
Cost: You are just getting started in your AI journey. This is still an experiment. Unless there is a clear and immediate benefit of the tool, you should consider the cost. There are many free models available for you to dip your toe in before you commit!
Adoption / Support: Since you are new-ish, it’s always best to find a model or platform that provides hands-on support or is widely adopted so you can self-study.
With this in mind, if you need a place to start, go with ChatGPT (GPT stands for Generative Pre-trained Transformer), which is a product made by the company OpenAI. It is the most widely adopted so that means that it has the most publicly available information on best-practices and uses. Others to consider are Anthropic’s Claude, Google’s Gemini, and Perplexity (made by Perplexity AI), each with their own pros and cons. At this stage, it’s a matter of personal preference and picking one and getting started. This beats spending days / weeks testing different models for optimization.
With this in mind, I am going to provide examples in this article focused on using OpenAI’s ChatGPT. You can of course take these tips to your favorite models and extrapolate how they might help you there!
Prompt Engineering 101
Your job as a user of any large language model is to ensure that you are providing the right information for it to work for you (and not make stuff up while doing it–an unfortunate habit called “hallucination”). The six steps below are key to beginning your prompts. In this case, a prompt is telling Chat GPT what you want it to do, and in particular, what role you want it to play for you. Do you want it to be a consultant? A coach? A copyeditor?
You may not always need to use all of these steps, or you might need more depending on what you are trying to accomplish, but these will get you started!
Once you have opened Chat GPT, you will want to:
Set the Role: Always start by telling the model who it should act as.
Example: “You are a fundraising consultant advising a small nonprofit.”Provide Clear Context: Briefly describe the situation, goal, audience, and any relevant background info.
Example: “We are writing an email to donors who gave between $250–$1,000 last year.”Give Specific Instructions: Be direct about what you want (tone, length, format, perspective).
Example: “Write 2 paragraphs in a warm, professional tone with a call-to-action at the end.”Include Examples: Add 1-2 examples of desired output style or structure if consistency is important.
Example: “Here’s how we wrote it last quarter...”Refine with Follow-Ups: Ask the model to review or revise: tone, structure, clarity, or alignment to mission.
Example: “Make this sound more aligned with our brand voice and mission” (and give it an example of your voice, by cutting and pasting a letter that you’ve sent before or a link to your website).
With your model and prompting in mind, let’s dive in!
Six Use Cases for AI
A recent article released by OpenAI highlights the summary from over 600 use cases of their most successful customers. While the customers varied in size, sector and tech maturity, all of their uses fell into just six categories: content creation, automation, research, coding, data analysis, and ideation & strategy.
This image was generated using ChatGPT
I will walk you through each of these, what they mean, and their immediate and long-term implementations in your nonprofit.
1. Content Creation
Definition
Use AI to write, edit, or improve content such as newsletters, board updates, donor emails, grant narratives, or blog posts.
Level of Difficulty: Low
Why: Requires no technical setup—just open your model and paste your prompt.
How to Start (Simple Steps)
Decide on the type of content (e.g., newsletter, social media post).
Write down a few facts or updates you want included.
Open your preferred model and paste your prompt using the template below.
You are the communications lead at a nonprofit focused on [insert mission or cause area]. Please write a [type of content] that highlights the following:
[Insert #] individuals served
[Insert #] new volunteers, donors, or partners
[Insert 1-2 other key highlights or milestones]
Use a warm, appreciative tone. Keep it to 2-3 paragraphs. Conclude with a thank-you message to supporters and a soft reminder that their continued support makes this work possible.
Interact with the model to refine the messaging.
Review and edit the output as needed to match your voice.
Prompt Example (Executive Director)
You are a communications lead at a nonprofit that provides free after-school tutoring. Please draft a donor newsletter update highlighting that we served 80 students and added 10 new volunteers last month. Use a warm, appreciative tone and include a thank-you message to supporters.
Advanced Use
Create a custom bot trained on your nonprofit’s tone, mission, and previous content. Use it to generate newsletters, social media posts, and donor letters with a single prompt.
2. Automation
Definition
Use AI to automate repetitive communication tasks such as reminders, thank-you emails, and follow-ups.
Level of Difficulty: Medium
Why: Writing prompts is easy, but automating the sending of messages often requires additional tools like Zapier or Make.com.
How to Start (Simple Steps)
Identify a repetitive task (e.g., sending reminders to volunteers).
Create a spreadsheet with names, emails, and otherrelevant info.
Paste the table and a prompt into your preferred model.
You are an operations assistant at a nonprofit. Use the following list of volunteers and their scheduled shifts to write polite reminder emails. Format them so I can copy/paste easily into Gmail.
[Paste list with Name, Email, Date]
Copy/paste the results into your email platform.
Prompt Example (Executive Director)
You are an operations assistant at a nonprofit. Use the following list of volunteers and their scheduled shifts to write polite reminder emails. Format them so I can copy/paste easily into Gmail.
Jamie Smith, jamie@email.com, June 20, 10 AM, Food Pantry
Pat Lee, pat.lee@email.com, June 21, 1 PM, Warehouse
Advanced Use
Connect your spreadsheet to Mailchimp or Gmail using Zapier. Use AI to automatically generate and send personalized emails weekly.
3. Research
Definition
Use AI to summarize long or complex documents like RFPs, legislation, or program evaluations.
Level of Difficulty: Low
Why: You just copy and paste content into your preferred model—no software setup required.
How to Start (Simple Steps)
Copy the full text of a document you need summarized.
Paste it into your model of choice with a clear question or instruction.
Ask for action items or red flags based on the summary.
Prompt Example (Executive Director)
You are a grant writer for a nonprofit. Please summarize this grant RFP in plain language. Include eligibility, funding amount, key deadlines, and any red flags I should watch out for.
[Paste RFP text]
Advanced Use
Use a document Q&A tool or Custom GPT to upload and search across dozens of RFPs, asking: “Which of these could support our food access program?”
4. Coding
Definition
Use AI to generate scripts that automate workflows, like syncing data between your CRM and email platform or formatting messy spreadsheets.
Level of Difficulty: High
Why: The prompt is easy, but running code often requires developer or tech volunteer help.
How to Start (Simple Steps)
Identify a task (e.g., syncing data from Google Sheets to Mailchimp).
List the tools and what you want to automate.
Paste the goal into your model of choice and ask for code.
Share the code with a developer or volunteer.
Prompt Example (Executive Director)
You are a technical support employee at a nonprofit. Write a Python script that reads names and emails from a Google Sheet and adds them to a Mailchimp audience list using their API. Assume I have API keys.
Advanced Use
Automate real-time data flows between donor platforms, CRMs, and dashboards. Add logic to track tags, segment donors, or send automated thank-you emails.
5. Data Analysis
Definition
Use AI to analyze donor, attendance, or program data from spreadsheets. Identify patterns, outliers, or summaries without needing formulas.
Level of Difficulty: Medium
Why: Simple to use with clean data; large or messy data may need formatting first.
How to Start (Simple Steps)
Export your data to Google Sheets or Excel.
Copy a manageable portion of the table (too much data and AI gets confused).
Paste it into your model of choice with a prompt asking for trends or summaries.
Ask follow-up questions like “How could I present this to the board?”
Prompt Example (Executive Director)
You are a fundraising analyst at a nonprofit. Here's our donation data from last year. Tell me the best months for giving, any seasonal trends, and any other interesting information you notice.
[Paste simplified table]
Advanced Use
Use a live dashboard (e.g., Google Data Studio or Airtable) and connect it to GPT to answer questions like, “What’s our average gift size by donor type this year?”
6. Ideation & Strategy
Definition
Use AI to brainstorm ideas for fundraising campaigns, programs, events, or strategic initiatives.
Level of Difficulty: Medium
Why: No setup needed. Great for kickstarting planning sessions or board materials. However, if you are using a publicly available model, you must ensure that you review all outputs before taking action.
How to Start (Simple Steps)
Identify the challenge (e.g., need a new campaign idea).
Describe your team’s size and constraints (budget, staff, timeline).
Ask your model of choice for 3-5 ideas using the below template.
You are a nonprofit strategist helping a small team plan an upcoming fundraising campaign. The organization has:
[Insert #] full-time staff
[Insert #] active volunteers
[Insert seasonal timing or event context]
A goal to raise [Insert $ amount]
Limited budget and capacity (specify virtual, in-person, or hybrid preference)
Suggest 5 creative, realistic fundraising ideas that fit within these constraints. Each idea should be actionable and achievable with minimal spend.
Follow up to turn one idea into a draft plan or outline.
Prompt Example (Executive Director)
You are a nonprofit strategist helping a small team with 2 staff and 10 volunteers. We need a fall fundraiser that’s low-cost and easy to run either virtually or in-person. Suggest 5 creative ideas that could raise $5,000 (net).
Advanced Use
Ask AI to build a full 6-week campaign plan with tasks by role, sample emails, board involvement ideas, and outreach scripts.
Getting Started Today: Your Next Steps
You don’t need a team of engineers or a six-figure budget to begin using AI effectively in your nonprofit. The smartest way to start is small, focused, and fast.
Here’s a simple system to get going:
Pick one of the 6 AI use case categories (content creation, automation, research, coding, data analysis, or strategy).
List three tasks you regularly do that fall under that category.
Choose the most annoying or time-consuming one.
Search for a language model and write a prompt that solves that specific task.
Test it for two weeks—push through initial glitches, try variations, and see what sticks.
If it works, great. If it doesn’t, move on—no sunk cost.
Revisit the other two tasks in that category.
Then move to the next category.
Some tools will surprise you. Others will overpromise and underdeliver. That’s fine. You only need one or two great use cases to save hours a week, improve consistency, or unlock capacity on your team.
Want to see how PlanPerfect can support your organization’s strategy through AI? Book a demo here.
Author:
Questions or comments?
Reach out to us at founders@planperfect.co!