For Grant Writers ·
What you'll accomplish
By the end of this guide, you'll have a repeatable process for drafting any grant proposal section in ChatGPT — from needs statements to evaluation plans — with specific prompting techniques that produce grant-quality output rather than generic text, cutting your per-proposal writing time by 50–70%.
What you'll need
Go to chat.openai.com and sign in (or create a free account). Click New chat.
Before writing any proposal content, start your session with a context-setting message. This "primes" ChatGPT with your organizational details so every response is accurate to your organization:
Paste this and fill in your details:
I'm a grant writer for [Organization Name], a nonprofit in [city, state]. Here's our context for this session:
Mission: [paste mission statement]
Programs: [list your 2-3 main programs with one-sentence descriptions]
Population served: [describe]
Key outcomes (most recent year): [list 3-5 data points — numbers served, completion rates, etc.]
Annual budget: approximately $[X]
Please use this information accurately in all grant writing you help me with today. Never invent statistics beyond what I've provided. Flag if you need additional information.
Press Enter. ChatGPT will confirm it understands your context.
What you should see: A brief acknowledgment that ChatGPT has your organization details and is ready to help.
Troubleshooting: If ChatGPT produces generic content that ignores your context, paste the context again at the start of your next message as a reminder.
Once context is set, request the section you need. The more specific your request, the better the output.
Bad prompt (too vague): "Write a needs statement for my grant."
Good prompt (specific): "Write a 400-word needs statement for a community foundation grant application. The funder's priority: addressing youth academic achievement gaps. Our program: after-school tutoring. Key data I'm providing: [paste your specific statistics]. Structure it as: national problem → local conditions → specific gap → our program as the solution."
What you should see: A structured 400-word draft that uses your actual data points woven into the narrative, not invented statistics.
Read the output. Don't start over if it's close — refine it with specific follow-up instructions:
ChatGPT will revise and return an improved version. Typically 1–2 rounds of refinement produces a draft ready for final review.
Before using any ChatGPT output in a real proposal:
Keep a running Google Doc or Word file of your best ChatGPT-generated sections, organized by type (needs statement, evaluation plan, etc.) and program area. Over time, this becomes a high-quality boilerplate library you can draw from.
1. Needs statement:
"Write a [X]-word needs statement for [funder priority]. Structure: national problem → local data → specific gap → our solution. My data: [paste stats]."
2. Program description:
"Write a [X]-word program description for our [program name]. Include: problem addressed, service model (what/how often/with whom), evidence base, and outcomes. Data: [paste yours]."
3. Organizational capacity:
"Write a [X]-word organizational capacity section. Include: years operating, staff size, budget, track record, partnerships, and relevant experience for this grant. Facts: [paste yours]."
4. Evaluation plan:
"Write an evaluation plan for our [program]. Include: data collection methods, specific instruments, responsible staff, analysis approach, reporting timeline. Our objectives: [list]."
5. Sustainability:
"Write a [X]-word sustainability section explaining how we'll sustain this program after the grant period. Our funding mix: [describe]. Our strategy: [describe]."