Optimize anything with AI-assistanceCopy
This AI workflow turns any initiative (from service delivery and fundraising campaigns to operational processes) into a dynamic learning system. It helps you move beyond basic reporting to systematically analyze stakeholder feedback, connect it with performance data, and generate actionable insights for continuous improvement.
By acting as your on-demand data analyst and innovation consultant, AI helps you make smarter, evidence-based decisions, ensuring your organization evolves to meet the real changing needs of your community.
Benefits
- Make data-driven decisions by uncovering the “why” behind your performance numbers.
- Foster a culture of continuous improvement by creating a structured process for acting on feedback.
- Increase stakeholder satisfaction by demonstrating that you listen to and act on input from clients, staff, volunteers, and donors.
- Identify new opportunities by systematically analyzing unmet needs, operational gaps, and emerging trends.
- Accelerate innovation by moving quickly from an idea to a testable plan.
Resources
Time:
2-4 hours for a full analysis and innovation cycle.
People:
1-2 people, including one manager and maybe a junior assistant or volunteer (to gather data, prepare surveys and tests, etc.).
Tools:
- Required:
- ChatGPT (Plus/Pro subscription with data analysis capabilities is strongly recommended).
- Optional:
- Survey tools (Google Forms, SurveyMonkey, Typeform) to collect feedback.
- Perplexity or other AI research tools for exploring best practices.
- Social listening tools (e.g., Hootsuite, Sprout Social, or simple manual searches) for monitoring online conversations.
- A/B testing tools (e.g., VWO, features within Mailchimp or other platforms).
Data:
- Qualitative feedback: Anonymized survey responses, interview notes, focus group transcripts, staff meeting minutes, social media comments, online reviews, or CRM notes.
- Quantitative data: Key performance metrics for the same period as the feedback (e.g., program attendance, donation conversion rates, volunteer retention, website engagement).
Workflow
This workflow is best performed in a single, continuous AI chat conversation. This allows the AI to build a deep, contextual understanding of your data and the connections between different steps.
STEP 1: Gather feedback with AI-Assisted surveys
Before you can analyze feedback, you need to collect it. AI can act as an expert survey designer to help you ask the right questions.
If you already have surveys (currently active or finished recently) with a good volume of useful responses, you can skip this step and use those responses in the next step.
Start a new chat to design your survey:
Act as a nonprofit survey design specialist. I need your help creating a short, effective survey to gather feedback.
# CONTEXT:
- Topic:
**Briefly describe the program, process, or initiative you want feedback on. e.g., "Our monthly donor newsletter," "Our volunteer onboarding process," "A youth mentoring program."**
- Goal:
**What is the single most important thing you want to learn? e.g., "Understand why newsletter open rates have declined," "Identify friction points for new volunteers," "Assess program impact from the perspective of participants."**
- Audience:
**Who will be taking this survey? e.g., "Current monthly donors," "Volunteers who joined in the last 6 months," "Youth participants aged 14-18."**
# REQUEST:
Please draft a 5-7 question survey that is clear, unbiased, and easy to answer. Include:
1. A brief, friendly introduction for the survey.
2. A mix of question types (e.g., multiple choice, rating scale, open-ended).
3. Questions that directly address our stated goal.
4. A concluding thank-you message.
Ensure the survey is optimized for a high completion rate and getting really useful feedback (actionable responses).
STEP 2: Analyze qualitative feedback
In addition to direct surveys, you can also gather valuable qualitative data by searching social media platforms, community forums, and review sites for mentions of your organization or relevant keywords. Sometimes you get more candid and useful feedback from this (social listening) than from surveys.
Once you have your feedback (from surveys, social listening, interviews, etc.), feed it to the AI. It will act as a qualitative researcher, organizing unstructured comments into actionable themes.
Continue the chat (or start a new one) and upload your data:
Act as a qualitative data analyst for a nonprofit. I need you to analyze a collection of raw feedback to identify key themes and insights.
# DATA:
**Upload your anonymized feedback files (e.g., survey exports, interview transcripts, compiled social media comments) OR paste the raw text here.**
# REQUEST:
Please analyze all the provided feedback and produce the following:
1. Thematic Analysis: Identify the top 5-7 recurring themes in the feedback. For each theme, provide a brief description and 2-3 direct (but anonymized) quotes that exemplify it.
2. Sentiment Analysis: Categorize the overall sentiment of the feedback as Positive, Negative, or Mixed. Provide a rough percentage for each category.
3. Actionable Insights: List the top 3 most critical issues raised by the feedback and the top 3 biggest strengths highlighted.
STEP 3: Connect feedback to performance data
Now, bridge the gap between what people are saying (qualitative) and what your numbers show (quantitative). This crucial step helps uncover the root causes of your performance.
Continue in the same chat:
OK, now let's cross-reference those qualitative themes with our quantitative performance data to look for connections.
# DATA:
Qualitative Themes:
Get them from your previous message.
Quantitative Data:
**Upload a spreadsheet with your key performance indicators for the same period OR paste a summary here**
# REQUEST:
Please analyze both sets of data and identify potential correlations. Answer the following questions:
1. Can you find a connection between certain negative feedback themes and a decline in our performance metrics?
2. Can you find a connection between certain positive feedback themes and an increase in our performance metrics?
3. Based on this cross-analysis, what are the 2-3 most likely "levers" for improving our performance? (i.e., The issues that, if addressed, would have the biggest impact on our results).
STEP 3: Generate improvement hypotheses
Turn your insights into specific, testable ideas. This moves you from analysis to action.
OK, now let's turn those "levers" into concrete improvement hypotheses.
# DATA:
Key Levers for Improvement:
**Paste the best levers identified in the previous step (don't copy the ones that you don't find relevant)**
# REQUEST:
For each lever, formulate a specific, measurable, achievable, relevant, and time-bound (SMART) hypothesis for an improvement.
Frame each one like this: "If we [specific action], then [expected measurable outcome] will occur within [time frame] because [reason based on the data]."
Example: "If we offer childcare during our morning workshops, then attendance among parents of young children will increase by 30% within the next quarter because lack of childcare was a major barrier identified in the feedback."
STEP 5: Brainstorm innovative solutions
Move beyond incremental fixes and use the AI as a creative partner to think about transformative new ideas.
Those hypotheses are great for incremental improvements. Now, let's think bigger. Act as a social innovation consultant.
# DATA:
Consider everything in the current conversation.
And also the following context:
**Provide more info about your organization or this specific process/initiative. You can use text, files and/or URLs. This is optional, but may help you get more relevant/personalized ideas**
# REQUEST:
Give me at least 3 bold and innovative ideas. Think deeply about this before answering. If necessary, do online research to find great ideas, brainstorming frameworks and/or relevant data. The ideas should go beyond simple tweaks to our current operations.
For each idea:
1. Give it a catchy name.
2. Provide a one-paragraph concept description.
3. Explain how it leverages our core strengths and directly addresses an unmet need.
STEP 6: Develop a Pilot Plan or a Simple Test
Take your most promising idea and create a plan to test it. Depending on the idea’s complexity, this could be a multi-week pilot project or a simple A/B test that runs for a few days.
Option A: For complex ideas, create a pilot plan
I like the idea for **Name of the most promising idea**
Help me sketch out a plan for a small-scale pilot project to test it.
# REQUEST:
Create a pilot project plan for this idea. Include the following sections:
1. Pilot Objective: What is the single most important question we are trying to answer with this pilot?
2. Key Activities: List the core activities we need to perform for the pilot.
3. Success Metrics: What key numbers will tell us if the pilot is successful?
4. Timeline: A simple 3-month timeline (Month 1, Month 2, Month 3) with key milestones.
5. Budget Estimate: A list of the main potential costs (e.g. staffing, materials, marketing).
Option B: For simple changes, design a direct test
OK, now act as a digital growth specialist and help me design a simple A/B test.
# CONTEXT:
I want to test one of your hypothesis:
**Paste hypothesis here (e.g. simplifying our online donation form will increase completion rates)**
We have the following restrictions:
**Mention here if you have to use a certain tool or platform for the test, have specific time or budget limits, don't have a big audience to test, don't want to change certain things, etc.**
# REQUEST:
Outline a plan for a simple A/B test. Include:
1. Control (Version A): Describe the current version.
2. Variation (Version B): Describe the specific change to be tested.
3. Primary Metric: What single number will determine the winner? (e.g. Click-through rate, donation conversion rate, form submission rate...).
4. Setup recommendations: If you recommend specific settings, tools, duration, etc.
Recommendations
- Anonymize all feedback thoroughly. Before uploading any stakeholder feedback, rigorously scrub it of all personally identifiable information (PII).
- Garbage in, garbage out. The quality of your AI-generated insights depends on the quality of the data you provide. Ensure your feedback is from a representative sample and your performance data is accurate.
- Leverage social listening. Don’t just rely on surveys. Monitor social media, online forums, and review sites for honest feedback about your organization and its work. This can reveal insights you wouldn’t get otherwise.
- Combine AI insights with human wisdom. AI is a powerful pattern-recognition tool, but it lacks the on-the-ground context and lived experience of your team. Use the AI’s analysis as a starting point for a deeper strategic conversation with staff and stakeholders.
- Close the feedback loop. When you make a change based on feedback, communicate it back to the people who provided the input. This shows your community they’ve been heard and encourages them to share more in the future.
- Start small. Whether it’s a pilot or an A/B test, the goal is to learn quickly and cheaply. Resist the urge to launch a full-scale initiative before you have data that validates your idea.