Design for social innovation
Client: Good Academy AI
Good Academy is an AI education provider focused on empowering nonprofits and mission-driven organizations. Built on the core values of empowerment, accessibility, practicality, ethics, and curiosity - Good AI Academy is currently in its alpha stage and is focused on understanding client needs and exploring program structure. While currently offering three-week online synchronous bootcamps with live instructors, the organization's long-term vision is centered on developing an asynchronous online learning product in partnership with Charming Robot, a NYC-based design agency.
Consulting Team
Challenge
Client Needs
- Curriculum development: Create adaptable AI training programs for various non-profit departments and transitioning to asynchronous online learning
- Business growth: Explore strategies to deepen reach within the non-profit sector and refine offerings for diverse budgets
- Vertical expansion: Extend curriculum programs into new markets like healthcare and law
- Marketing & thought leadership: Position as a trusted authority in AI education for non-profits
- Measuring impact & outcomes: Demonstrate tangible benefits of training programs
- Business growth: Explore strategies to deepen reach within the non-profit sector and refine offerings for diverse budgets
- Vertical expansion: Extend curriculum programs into new markets like healthcare and law
- Marketing & thought leadership: Position as a trusted authority in AI education for non-profits
- Measuring impact & outcomes: Demonstrate tangible benefits of training programs
Project Objectives
Analyze Pain points: Conduct in-depth research to identify specific AI-related challenges faced by public interest organizations across various sectors.
Refine the Value proposition: Develop a compelling value proposition that articulates Good Academy’s differentiators in the AI education landscape
Assess Product-market fit: Evaluate the alignment between Good Academy’s current and planned offerings with the identified pain points and market needs, informing strategic decisions on curriculum development and delivery methods
Research
Key Client Hypotheses
1. Public Interest professionals are very nervous about AI and integrating it into their workflows.
2. We can develop an AI literacy curriculum that can be generalized across all social-impact sectors.
3. There is a need for role-based curriculum for CEOs vs. Managers vs. Outreach workers
2. We can develop an AI literacy curriculum that can be generalized across all social-impact sectors.
3. There is a need for role-based curriculum for CEOs vs. Managers vs. Outreach workers
Research methods
Stakeholder Mapping of 5 key stakeholder groups + Empathy Mapping from 20 preliminary interviews
Market Analysis of 16 competitors + Benchmarking alongside 10 products in the current market
Product Analysis of a full bootcamp offering (3 lectures)
Market Analysis of 16 competitors + Benchmarking alongside 10 products in the current market
Product Analysis of a full bootcamp offering (3 lectures)
Stakeholder Mapping
Primary Axes of Comparison:
- Openness to Tech vs. Resistance to Tech
- High Power/ Influence over the Product vs. Low Power/ Influence over the Product
- Openness to Tech vs. Resistance to Tech
- High Power/ Influence over the Product vs. Low Power/ Influence over the Product
Secondary Axes of Comparison:
- People who work with People vs. People who work with Systems
- Low resources available vs. High resources available
- People who work with People vs. People who work with Systems
- Low resources available vs. High resources available
Stakeholder Categories and Key Themes of Investigation
We consolidated all the various stakeholders into 4 key categories: Client, User, Product-Market Fit, and User Acquisition Channel. For each category, we developed a unique set of research questions to explore their pain points, motivations, behaviors, interests, and incentives.
1. Good Academy Staff (Client): Daily Workflow, Product-Market Fit, Value Proposition, Curriculum, Client Onboarding, Bootcamp Outcomes
2. Good Academy Alumni (Current User): Involvement in Product Development, Literacy gaps in AI, Client/ Bootcamp experiences
3. Public-interest professionals (Target User): Daily Workflow, AI use in personal/ professional life, Need for AI skills, Problem-solving with AI
4. AI Experts (Product-Market Fit): AI applications in public-interest work, Trends in gaps in AI literacy, Barriers to AI literacy, Strategies to empower people
5. Nonprofits (User Acquisiton Channel): Overall concerns, pain-points in the workflows, Opportunities to optimize, Budget/ interest to upskill
2. Good Academy Alumni (Current User): Involvement in Product Development, Literacy gaps in AI, Client/ Bootcamp experiences
3. Public-interest professionals (Target User): Daily Workflow, AI use in personal/ professional life, Need for AI skills, Problem-solving with AI
4. AI Experts (Product-Market Fit): AI applications in public-interest work, Trends in gaps in AI literacy, Barriers to AI literacy, Strategies to empower people
5. Nonprofits (User Acquisiton Channel): Overall concerns, pain-points in the workflows, Opportunities to optimize, Budget/ interest to upskill
Product Analysis
Key Observations and Areas for Improvement:
- The current lecture content establishes a good foundation but requires additional resources to enhance learning
- The session on AI privacy and ethics do not address the breadth or depth of concerns adequately.
- Group exercises have inconsistent timing, engagement, and relevance as they do not cater to diverse levels of learning.
- Lack of 1on1 instruction, guidance, or support.
- The session on AI privacy and ethics do not address the breadth or depth of concerns adequately.
- Group exercises have inconsistent timing, engagement, and relevance as they do not cater to diverse levels of learning.
- Lack of 1on1 instruction, guidance, or support.
Market Analysis
A landscape analysis of 16 competing organizations and a benchmarking analysis of 10 competing products in the current market. Given that Good AI Academy current offering is a live bootcamp series on foundational AI literacy, the primary axes of comparison for the market analysis were the "type of classes" (operationalized as live vs. asynchronous courses) and the "Level of detail/ nuance" (operationalized as "general" vs. "tailored/ personalized for public-interest"). The Benchmarking analysis delved further into the quantity and quality of features that could either provide Good Academy a competitive edge or place it at a position of disadvantage.
Empathy Mapping
How the users are feeling about AI:
- "How can you trust something that doesn't feel real"
- "People built these tools - so the tools aren't perfect. AI isn't god."
- "How can you trust something that doesn't feel real"
- "People built these tools - so the tools aren't perfect. AI isn't god."
What the users are saying about AI:
- "We need examples of good AI outputs and bad AI outputs to be critical users"
- "Some tech bro comes in to talk to a group of clinicians and public-health workers (with no context) about how to do their job better"
- "We need examples of good AI outputs and bad AI outputs to be critical users"
- "Some tech bro comes in to talk to a group of clinicians and public-health workers (with no context) about how to do their job better"
What the users are thinking about AI:
- If AI will rely on current databases, then it will produce results with the same biases
- How can AI be a collaborator and not a replacement?
- If AI will rely on current databases, then it will produce results with the same biases
- How can AI be a collaborator and not a replacement?
What are users are doing with AI:
- Create meeting agendas, draft emails, and write social media posts.
- Find sources, analyze data, summarize findings, and write reports.
- Create meeting agendas, draft emails, and write social media posts.
- Find sources, analyze data, summarize findings, and write reports.
key findings
Stakeholder Insights
Hypothesis Testing
1. Intentionality: Need to establish rapport, trust, and credibility
- Public-interest professionals are aware of AI, already use it/ open to using it, and have specific safety/ ethics/ data privacy concerns
- Nonprofits felt disconnected and didn't trust "tech people" to understand their sector and unique issues
- Public-interest professionals are aware of AI, already use it/ open to using it, and have specific safety/ ethics/ data privacy concerns
- Nonprofits felt disconnected and didn't trust "tech people" to understand their sector and unique issues
2. Heterogeneity: Nonprofits cannot be generalized for AI up-skilling
- Nonprofits have unique applications, concerns, and levels of familiarity with AI tools and literacy.
- Nonprofits have unique applications, concerns, and levels of familiarity with AI tools and literacy.
3. Community: Need for peers and lasting connections
- Bootcamp attendees wanted long-term engagement with Good Academy, and to have shared experience beyond individual up-skilling
- Bootcamp attendees wanted long-term engagement with Good Academy, and to have shared experience beyond individual up-skilling
Human-Centered Design Principles
Guiding Design
How might Good Academy be inclusive of all learning levels?
How might Good Academy make AI adoption quick and easy?
How might Good Academy make AI adoption quick and easy?
Guiding Impact
How might we make Good Academy a community of Impact Makers?
How might Good Academy care about what the nonprofits care about?
How might Good Academy care about what the nonprofits care about?
Guiding Acquisition
How might we make Good Academy the Go-To-Place to learn Good Tech?
How might we make Good Academy the ultimate sidekick for public-interest changemakers?
How might we make Good Academy the ultimate sidekick for public-interest changemakers?
Sketching
Rapid idea generation
Our team collectively generated a variety of "solutions", in 60 second increments, that could potentially solve our client's pain points. The parameters for this exercise ranged from "0 budget solutions" to "infinite budget solutions" to "Disney-inspired solutions", etc. While there was a spectrum of mundane and outlandish ideas because of the varied prompts – once we started organizing these ideas under common categories (i.e. food, creative, policy, etc) we could see trends across ideas – and how they could become potential solutions in more ‘realistic’ categories (Ex. marketing, design, financial, etc).
Interestingly, Brainstorming under diverse constraints in short bursts meant that we could step out of perfectionist tendencies – and have the courage to bring “bad” ideas onto paper (instead of self-rejecting them).
Low-tech protoype: An AI Literacy Convening
High-tech prototype: A Slack Community for Good Academy
A dedicated Slack ecosystem for the Good AI Academy community members that features various forums for questions, collaborations, advice, and innovation. The Workspace is moderated by a Good AI Community Liaison as well as a Good AI chatbot. Between the human and AI-assisted support, community members have access to trending AI articles, resources, curriculum advice, networking support, and personalized lesson plans.
testing
Run a Demo Day with various cohorts of aspiring public-interest professionals (target users)
Interactive voting through 4 in-person focus groups, 10 user interviews, and online surveys
Key Learnings
iterating
Reframing user feedback into actionable takeaways
1. Participants responded more to messaging that highlights what they will gain (skills, certifications, connections) >> Make Event Outcomes Clear
2. Sustained Slack engagement is higher if participants first experience it in a low-barrier setting (live demo) >> Slack intros/ demos at Events
3. People care about credibility and social proof, and want opportunities to connect >> Networking at Events + Plug in for Slack
2. Sustained Slack engagement is higher if participants first experience it in a low-barrier setting (live demo) >> Slack intros/ demos at Events
3. People care about credibility and social proof, and want opportunities to connect >> Networking at Events + Plug in for Slack
Developing User Personas and Journey Maps
After the preliminary research, sketching, and testing - patterns around user motivations, behaviors and incentives were becoming clearer. Given the need to rapidly develop, test, refine, and retest our prototype on a short timeline - we decided to focus solely on the target user stakeholder group for further design research. The target user group spanned both current working public-interest professionals as well as aspiring public-interest professionals (i.e. students). While there were many similarities in terms of interest in connection and social impact - there were noted differences in acquisition channels and engaging behaviors. Therefore, we developed user persons and possible journey maps for both kinds of users to further understand how to refine both the low-tech as well as high-tech prototypes.
User Journey Mapping for Events
prototyping
Insight: Consolidate low tech and high tech prototypes into one seamless ecosystem
Prototype: Good Academy Connect
A low-resource, high-impact ecosystem that gives nonprofits what they need: community and hyperspecific insights. Drawing inspiration from existing tech communities like All Tech is Human, we’re aiming to build a a self-reinforcing system/cycle that turns AI skeptics into advocates.
Insight: Integrate Good Academy's LYRA Chatbot into Slack
Prototype: Establish Good Academy's Slack as an AI Impact Ecosystem for all changemakers
Insight: Create a feedback loop where customers stay engaged with Good Academy over time
Prototype: Incorporate the human-centered principles to design a virtual community experience that makes customers want to stay connected to Good Academy
Our initial hypothesis testing revealed that the target users for Good Academy honor the following core values: intentionality, respect for diversity of thought (i.e. honoring the heterogeneity inherent in their work), as well as a deep love for community. While it is pivotal for Good Academy to incorporate these values in product design - it is equally important for them to embody these values in how they design space for their community outside their classes (and offline from their product). Therefore, we designed virtual facilitation guides, digital community moderation guides, and a cheery branding guide to create spaces that would not only welcome like-minded individuals - but also help them thrive in their impact journey.
Outcomes and Lessons Learned
Outcome: Good Academy launched LYRA
LYRA is an AI-driven learning platform that replaces static, one-size-fits-all online courses with a dynamic, conversational AI tutor that adapts content, feedback, and practice to each learner's role, skill level, and real-life nonprofit workflow.
How LYRA incorporates findings from our research:
1. You emphasized the need to move beyond “AI 101” and focus on task-specific, horizontal applications. LYRA’s curriculum is modular and workflow-driven—from content creation to data analysis, ethics, and decision-making.
2. You highlighted the heterogeneity of nonprofit professionals and the limits of a one-size-fits-all experience. LYRA adapts dynamically to user roles, skill levels, and goals through an AI tutor that personalizes feedback, examples, and support in real time.
3. You pointed out the challenges of scaling the bootcamp model—especially regarding individualized guidance, privacy concerns, and ROI for busy professionals. LYRA is designed with self-paced, high-impact learning in mind, offering personalized coaching, safe data practices, and time-saving use cases that directly support users’ work.
Outcome: Good Academy is the ultimate AI learning sidekick for Impact Leaders
