Skip to content
Phase 4: Program Design

AI Prompt to Turn a Nonprofit Idea Into a Fundable Program Plan

By Angel Reyes, DrPH(c), PMP, PMI-ACP, MPH, MCHES · Updated · Free to copy and use
Quick Answer

This AI prompt helps you turn a nonprofit idea into a fundable program plan. The AI asks clarifying questions first, then organizes your answers into a one-page program design brief: who is served, the specific problem, the activities and how often they happen, staffing, partners, timeline, expected results, cost drivers and risks. It finishes with the design decisions that are still open. It uses only what you tell it.

The prompt

Copy and paste
You are a program design facilitator for a nonprofit. Help me turn a rough idea into a one-page program design brief that a funder could evaluate.

Use only my starting information and my answers. Do not invent facts, activities, partners, evidence, costs, results, timelines or staffing. Label anything I have not decided "Not yet determined."

Organization and mission: [ORGANIZATION AND MISSION]
The rough idea: [PROGRAM IDEA]
Community or location: [COMMUNITY OR LOCATION]
Who might take part: [POSSIBLE PARTICIPANTS]
What we already have, and our limits: [ASSETS AND CONSTRAINTS]
Where the idea came from: [PARTICIPANT REQUESTS, STAFF OBSERVATION, DATA OR OTHER]

First, ask me focused clarifying questions, grouped under these topics, and wait for my answers before drafting:
- who is served and how they would be reached
- the specific problem and local context
- the change we intend
- activities, with dose (how often and how intensive) and duration (how long participation lasts)
- staffing and capacity
- partners
- timeline
- expected results and how we would measure them
- what drives the cost
- risks
Ask no more than three questions per topic.

After I answer, write the brief under these headings:
1. Who is served
2. The specific problem
3. Program goal
4. Activities, dose and duration
5. Staffing and partners
6. Timeline
7. Expected results and measurement
8. Cost drivers
9. Risks
10. Open design decisions

In the brief:
- Separate confirmed partners from possible ones.
- Separate planned results from results already achieved.
- List cost drivers without dollar amounts unless I supplied them.
- Include a risk response only if I gave you one.

Do not describe the design as evidence-based unless I supply the evidence. Do not write a grant narrative or a budget. End with a short checklist of points for program staff and community members to validate.

Works in ChatGPT, Claude, Gemini and other AI chat tools. Replace each bracketed placeholder with your own details before you send it.

What to fill in

[PROGRAM IDEA]
Whatever you have, in plain words. Partial ideas are fine. The questions are there to fill them out.
[POSSIBLE PARTICIPANTS]
Who you think the program is for, described generally. Do not include names or details that identify individuals.
[ASSETS AND CONSTRAINTS]
Staff time, space, existing relationships and funding you can count on, plus limits such as geography, schedule or licensing. Keep confirmed resources separate from possibilities.
[PARTICIPANT REQUESTS, STAFF OBSERVATION, DATA OR OTHER]
Be honest about the origin of the idea. A program that participants asked for and one that staff proposed are both legitimate starting points, but they are not the same, and a funder may ask.

Why this prompt works

An idea becomes fundable when a reader can tell who is served, what will happen, who will do it and how anyone would know it worked. Most early ideas are missing several of those. The prompt finds out which.

  • Questions before drafting. If you ask an AI to write up a rough idea, it fills the gaps with plausible details and hands back something that looks finished. Making it ask first puts the gaps in front of you.
  • A facilitator role. The AI is asked to draw out and organize your thinking. It is not asked to design the program or judge it.
  • Grouped, limited questions. Ten topics with a cap of three questions each keeps the conversation manageable and covers what funders look for.
  • Dose and duration. "Ongoing support" is not a design. How often, how intensive and for how long are the details that determine both cost and plausibility, so the prompt asks for them by name.
  • "Not yet determined." Undecided items are labeled and collected under open design decisions. That list is often the most useful part of the output.
  • Named separations. Confirmed versus possible partners, and planned versus achieved results, are the places where early plans overstate.

Why it is appropriate to use

This is the first step of program design: deciding what the program would actually do. A one-page brief gives staff, community members, partners and leadership the same document to react to before anyone writes a logic model, a budget or a proposal.

AI is appropriate here as an interviewer and note-taker. It is quick at turning a scattered conversation into a structured page. It should not be making the design choices, because it knows nothing about your community and has no stake in the result.

The brief cannot assess community need without real engagement, verify that an approach works, establish what a funder requires or tell you whether a funder would be interested. A coherent design is not the same as funder fit. That takes a specific opportunity and its guidelines.

"Fundable" here means clear enough for a funder to evaluate. It is not a prediction that the program will be funded.

Keep names, case histories, client records and internal financial documents out of the chat. When the brief is settled, the logic model prompt is the natural next step.

How to check and revise the output

  1. Check it against what you said. Compare every statement with your answers. Remove invented details and restore "not yet determined" where the AI smoothed over a gap.
  2. Test delivery. Do the activities, frequency, duration and staffing fit the capacity you actually have?
  3. Look at results. Are services delivered kept apart from the changes participants should experience? Is each measure practical to collect?
  4. Review it with program staff and people from the community served. They decide whether the design is right, and their changes go in before anything else.
  5. Work the open decisions list. Settle each one with the right people, then rerun the prompt with the decisions made.
  6. Apply the reader test. Can someone outside your organization tell who is served, what happens, who delivers it and how progress is judged? If not, go back to the open decisions. More persuasive wording will not fix it.

Frequently asked questions

What if I cannot answer all the clarifying questions?
Say which answers you do not know. The brief should mark them as not yet determined and list them among the open design decisions. Do not let the AI choose services, staffing or schedules for you.
Is the brief ready to paste into a grant application?
No. First validate the design with staff and the community, confirm every factual claim and check the funder's policy on AI use. Then adapt the approved design to the questions in the application. The brief is not a grant narrative or a budget.
What makes a program idea fundable?
A funder needs to see a defined population, a specific problem, activities that plausibly address it, the capacity to deliver, results that can be measured and costs that match the work. It also has to fit that funder's priorities, which is a separate question from whether the design is sound.

More prompts for program design

Related prompts and guides

← Back to all AI prompts for grant writing

Ready to Master Grant Writing?

The Complete Grant Architect is a free 16-week course that transforms you from grant writer to strategic grant professional. Learn proposal engineering, federal compliance, budgeting, evaluation design, and AI-powered workflows.

Enroll Free in The Complete Grant Architect