Case Study: MomentumDIY From Idea to Publishable MVP
- Hillary McMullen
- Mar 24
- 4 min read
Executive Summary
I designed and built MomentumDIY, a SaaS platform designed to reduce cognitive load for small business owners by turning abstract marketing strategies into linear, actionable "tracks." This project served as a technical proof-of-concept for AI-agentic development, moving from initial architecture to a production-ready MVP in four months.
1. The Problem
Small business owners are paralyzed by marketing overwhelm. They have the tools, but they lack the System or knowledge to use them effectively. Most marketing platforms are built for experts, leaving Main Street owners drowning in features they don’t understand. The existing platforms make small businesses' marketing wheels constantly spin without actually making any forward progress.
2. The Solution
From interacting with so many small businesses in my line of work, I really wanted to develop a solution for this gap in the market. I thought there must be a better way to give small businesses the tools and knowledge they need, without too much effort or time commitment. With the assistance of an AI IDE, I developed a simplified, task-oriented SaaS that turns complex marketing strategies into a linear, gamified 90-day "Track" that helps to keep small business owners focused on their marketing tasks at hand.
3. The Technical Infrastructure
Instead of a simple "no-code" builder, I architected a decoupled system focused on high performance and data integrity. Check out my GitHub repository if you're interested in diving deeper into the code infrastructure.
The Backend (Supabase & PLpgSQL): I utilized Supabase for authentication and data persistence. To ensure data consistency across complex marketing tracks, I wrote custom PLpgSQL functions. This allowed me to handle logic at the database level, reducing latency and ensuring that user progress was immutable and accurately tracked.
The Frontend (TypeScript): Choosing TypeScript (92.6% of the codebase) was a strategic decision to enforce strict typing across the application. This minimized runtime errors and allowed for a much smoother integration of complex APIs, making the system "self-documenting" for future engineering handoffs.
Communications (Resend): I integrated Resend to handle transactional email notifications. I designed these triggers to be state-dependent, ensuring that users receive "nudge" reminders only when specific task thresholds weren't met, effectively reducing "notification fatigue."
4. User Experience & Friction Reduction
I really wanted to step into the shoes of these small business owners. Understand what tools and information were already out there, and think about what an actual helpful tool might look like for them. Upon completing my research of the existing tools and developing my user personas, I applied first principles thinking to the user journey, focusing deeply on the psychology of small business owners.
Cognitive Load Management: Small business owners are often non-technical. I designed the UI to be Information-Light. I used a "progressive disclosure" strategy where users only see the current task and the immediate next step, preventing the "paralysis of choice."
The Gamification Logic: I implemented an "Interaction Design System" that includes confetti triggers and progress bars. These aren't just aesthetic; they serve as behavioral hooks that reward consistency—a critical factor in marketing success.
Automated Onboarding: I built a logic-driven onboarding flow that categorizes the user's business type and automatically assigns the relevant "Content Track." This reduces the "Time to Value" from minutes to seconds.
Results & Technical Takeaways
1. Performance & Deployment Velocity
The primary goal was to validate a Design-to-Code pipeline. By utilizing an AI-agentic supervisor model, I moved from initial conceptual wireframes to a production-ready, full-stack MVP in under four months.
Outcome: Successfully deployed a 92.6% TypeScript codebase on Vercel, achieving near-instant load times and 100% uptime during the initial pilot phase.
Infrastructure: The decoupled architecture (Supabase + Resend + Vercel) proved that a single "Systems Designer" can manage enterprise-level infrastructure without the overhead of a traditional DevOps team.
2. UX Validation: Reducing Cognitive Load
The central hypothesis of MomentumDIY was that linear pathing reduces user churn.
Result: My small sample of pilot users reported a decrease in "decision fatigue." By stripping away the traditional "dashboard" view in favor of a "Next-Action" track, the Time to Value (TTV) for a new user was reduced to under 2 minutes post-onboarding.
Interaction Success: The implementation of state-dependent UI feedback (progress bars and confetti triggers) increased task completion rates by providing immediate visual "rewards" for manual administrative work.
3. Technical Learnings: The Power of Type Safety
Building 92.6% of the application in TypeScript provided a massive lesson in long-term maintainability.
Key Takeaway: Enforcing strict types from the start allowed for seamless integration of the Supabase backend and Resend API. It turned the codebase into a "self-documenting" system, significantly reducing the "debugging tax" and allowing for rapid feature iteration without breaking core functionality.
Database Logic: Moving complex user-state logic into PLpgSQL at the database level was a strategic win. It ensured that "Content Tracks" remained immutable and consistent, regardless of frontend state changes.
4. Security & Operational Rigor
A major takeaway from this project was the importance of Secret Management and DevOps hygiene.
The Lesson: Managing a public-facing repository requires a strict "Security-First" approach. I successfully implemented a robust .env and .gitignore workflow to protect sensitive API keys while maintaining a public-facing code audit trail for transparency and collaboration.
5. Future Scalability: The Path to V2
While the MVP is a functional "shell" of the vision, the architectural foundation is ready for scale.
Next Steps: The next iteration would involve deepening the AI-Agentic integration, allowing the app to not just track marketing tasks, but to draft them using the user’s unique brand voice stored within the Supabase database.
Comments