The bet
B2B sales teams spend thousands of dollars sending reps to conferences, trade shows, and industry events without knowing if their Ideal Customer Profile (ICP) will actually be there. If you can score the event before they buy the flight, you save them $5k per rep.
What we ship
EventScout is an AI event intelligence tool for B2B teams.
- ICP Event Matching: Give it your ideal customer profile, and EventScout finds the events where those personas congregate.
- ROI Analysis: Predicts the return on investment based on historical data, attendee personas, and ticket prices.
- CFO-Ready Reports: Generates a PDF business case that the sales rep can hand directly to their CFO to get the travel budget approved.
Architecture
- Frontend: Next.js (App Router), Tailwind CSS
- Backend & Auth: Supabase (Postgres)
- AI/LLM: Google Gemini for parsing event descriptions, extracting attendee profiles, and generating ROI narratives
- Payments: Stripe integration for Starter ($49/mo), Pro ($99/mo), and Team ($249/mo) tiers
The 50-in-50 Challenge
I launched this with a strict "50-in-50" constraint: 50 paying customers in 50 days at an $80 Average Revenue Per User (ARPU). This forced ruthless prioritisation of features. If a feature didn't directly help a sales rep justify an event ticket to their CFO, it got cut.
What this taught me about building AI SaaS
- AI is just a feature. The core value prop isn't the LLM—it's the PDF report that gets the budget approved. The LLM is just the fastest way to generate that PDF.
- Speed over architecture. When you have 50 days to get 50 paying customers, you don't over-engineer the database. Supabase and Next.js allowed for shipping a fully authenticated, monetized product in a week.
- Solving a painful niche. "Event intelligence" sounds boring to a consumer, but it's a massive pain point for a B2B sales director managing a $100k travel budget.
Active: 2025 – present. Solo technical build.