Skip to content
Phase 4: Program Design

AI Prompt for a Grant Evaluation Plan

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

This AI prompt for a grant evaluation plan turns your objectives and your real capacity into a first-draft evaluation matrix. For each objective it gives the evaluation question, indicator, data source, collection method, timing and the role responsible. It then checks whether the plan is feasible with the staff, tools and budget you described, and lists the decisions that need program staff or an evaluator.

The prompt

Copy and paste
You are helping me draft an evaluation plan for a grant proposal. Use only the information I supply below. Do not invent baselines, targets, instruments, staffing, budget details or funder requirements. Mark anything unknown as "needs confirmation."

Program summary: [PROGRAM SUMMARY]
Population served, described without identifying anyone: [POPULATION]
Objectives: [OBJECTIVES]
Activities and expected outputs: [ACTIVITIES AND OUTPUTS]
Expected outcomes: [OUTCOMES]
Staff available for evaluation work, by role and time: [EVALUATION STAFF AND TIME]
Tools and systems we already use to collect data: [TOOLS AND SYSTEMS]
Evaluation budget or limits: [EVALUATION BUDGET]
Project timeline: [TIMELINE]
Funder's evaluation and reporting requirements: [PASTE FUNDER REQUIREMENTS]
Data we already have, in summary: [EXISTING DATA SUMMARY]
Our privacy and consent requirements: [PRIVACY REQUIREMENTS]

Part 1. Create an evaluation matrix with one row per objective and these columns:
Objective | Evaluation question | Type (process or outcome) | Indicator | Baseline (have, need or not applicable) | Data source | Collection method | When collected | Role responsible | How the finding will be used

Make every outcome indicator observable and specific about who and what is measured. Flag any objective that is too vague to measure as written, and say what is missing.

Part 2. For each objective, write a feasibility note. Given the staff time, tools and budget I described, can this data realistically be collected, entered and analyzed? Name any task that exceeds the capacity I described.

Part 3. List your assumptions and the details I did not provide.

Part 4. Under the heading "Decisions for program staff or an evaluator," list the open choices about measures, methods, sampling, analysis and interpretation.

Do not name specific survey instruments. Do not state or imply that measuring a change shows the program caused it. Keep the plan proportionate to the size of the project.

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

[OBJECTIVES]
Your objectives as currently written, ideally stating who will change, what will change and by when. If they are still rough, the prompt will flag which ones cannot be measured yet.
[EVALUATION STAFF AND TIME]
Who could actually collect and analyze information, by role, and how many hours they realistically have. This is what makes the feasibility check meaningful, so be honest.
[TOOLS AND SYSTEMS]
What you use now: attendance sheets, a database, intake forms, survey software. The best evaluation plans build on data you already collect.
[PASTE FUNDER REQUIREMENTS]
What the funder says about evaluation, required measures and reporting, copied from the announcement. If they require specific indicators, those take priority.
[EXISTING DATA SUMMARY]
Summaries only, such as last year's enrollment or completion numbers. Never paste participant records or anything that could identify a person.
[PRIVACY REQUIREMENTS]
Your organization's rules on consent and data handling, and any review process that applies to collecting information from people.

Why this prompt works

Evaluation sections fail in two ways. Either they are vague ("we will track outcomes"), or they promise more measurement than the organization can deliver. The prompt addresses both.

  • A fixed matrix. Ten columns force each objective to be connected to a question, a measure, a source, a schedule and a person. An empty cell is a visible gap.
  • Capacity is an input. Because you state staff time, tools and budget, the AI's suggestions are bounded by what you have instead of by what an ideal evaluation would include.
  • A feasibility pass. Part 2 makes the AI test its own plan against your capacity. Treat its notes as questions to check, not as confirmation that the work is achievable.
  • Process and outcome are labeled. Marking each row keeps a count of sessions from being presented as evidence of change.
  • Baselines are surfaced. The baseline column shows where you need a starting measurement before services begin, which is far easier to arrange now than after the award.
  • A decisions list. Method, sampling and analysis choices are handed back to people qualified to make them.
  • Two guardrails. No named instruments, because AI tools can cite instruments that do not exist or do not fit your population. No causal claims, because measuring change is not the same as showing your program caused it.

Why it is appropriate to use

Draft the evaluation plan during program design, while objectives, staffing and data collection can still be adjusted. A matrix often reveals that an objective cannot be measured as written, or that nobody has been assigned to collect the data. It is better to learn that now than when the first report is due.

AI is appropriate for this first draft because it is structuring information you supplied and pointing out gaps. The evaluation plan is also a commitment. A funder may hold you to what it says, so the people who will do the work have to agree to it.

It is not a substitute for evaluation expertise. Choosing instruments, judging whether a measure is valid for your population, designing a sample and interpreting sensitive findings are decisions for program staff or an evaluator. Collecting information from people may also require review under your organization's or your funder's rules, so confirm what applies with the appropriate staff.

Never paste participant data into an AI tool. For the methods themselves, see the guide to evaluation methods for grants.

How to check and revise the output

  1. Check alignment. Does each question address its objective? Does each indicator answer the question, or does it only count activity?
  2. Define each measure precisely. What exactly is counted or assessed, for whom and over what period?
  3. Confirm the source exists. Make sure the data is already collected or can reasonably be collected with the tools you have.
  4. Add up the workload. Consider collection, data entry, analysis and reporting together, for every objective at once.
  5. Get agreement on ownership. Replace role labels with people who have accepted the task and the schedule.
  6. Take the decisions list to an evaluator or experienced program staff before the plan goes into the proposal.
  7. Check it against the funder's requirements and your budget. Evaluation costs in the plan should appear in the budget.

Frequently asked questions

What is the difference between process and outcome evaluation?
Process evaluation looks at whether activities happened as planned, including participation, reach and delivery. Outcome evaluation looks at the changes associated with the program, such as knowledge, behavior or conditions. Observing a change does not by itself show that the program caused it.
What if we do not have baseline data?
Say so in the prompt. It will flag where a baseline is needed, and your staff or an evaluator can decide whether and when collecting one is feasible. Do not let the draft invent a starting value.
Can I paste the matrix directly into a proposal?
Only after reviewing and revising it. Confirm the measures, responsibilities, workload, privacy requirements and funder instructions first. The prompt creates structure and exposes gaps. Program knowledge and professional judgment finish the plan.

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