Quit $250K Job, Switched to Lifetime Deal, Made $80K in Six Months

Executive overview

Praneeth Pike left a $250K/year software engineering role to build apps full-time. His first version of Rabbit Holes AI — a canvas-based AI interaction tool — launched with a subscription model and went nowhere. The single change that unlocked $80K in six months was switching from a subscription to a one-time lifetime deal, converting to a desktop app, and making users bring their own API keys. The move eliminated server and API costs, attracted more serious buyers, and generated immediate validated revenue without outside funding.

The product: canvas-based AI interactions

• Rabbit Holes AI (now Canvas Mode) lets users branch conversations into nodes rather than scrolling a linear chat
• Built for power AI users frustrated by context loss in tools like ChatGPT
• Insight came from personally juggling multiple AI tabs — Claude for coding, GPT for architecture, Perplexity for live data
• Target ICP: users already paying for multiple AI subscriptions

What went wrong with the subscription model

• Launched as a web app with a free trial; conversion to paid was minimal
• No large personal brand to drive organic traffic at launch
• Power-user ICP was already saturated with subscription fatigue
• Freemium requires capital to delay truth-seeking — a luxury bootstrappers don't have
• Result: some interest on Twitter, zero meaningful revenue

The pivot: one-time purchase + desktop + BYOK

• Bring Your Own Keys (BYOK): users supply their own API keys, eliminating LLM costs entirely
• Moved from web app to Electron desktop app — eliminates server costs, no ongoing infrastructure bill
• One-time purchase means ship once, no recurring cost exposure
• Monthly operating cost after pivot: ~$50 (Vercel hosting) + ~$200 (Cursor subscription for ongoing dev)
• Margins: ~$75K profit on ~$80K gross revenue

Lifetime deal pricing strategy

• Start low to generate early signal and build a user base willing to give feedback
• Launched at $29 for the first 100 users
• Raised to $49 after crossing 100 users
• Plan to raise price further and layer in a recurring model now at ~1,200 users
• 60% of ~200 paying users are lifetime; 40% are annual-renewal one-time purchases
• Early buyers get the best deal; price increases reward momentum and validate demand

Why bootstrappers should default to lifetime deals

• An LTD gives an immediate cash injection that validates the idea concretely
• Buyers are more engaged and more willing to give quality feedback than free-trial users
• Prevents the VC trap of delaying product-market fit with runway
• Even if you already have a subscription model, add a limited LTD option alongside it
• Particularly valuable for solo founders who otherwise spiral into building-for-peers rather than building for customers

Growth tactics

• Landing page ICP clarity: headline and copy call out "Power AI users" explicitly — visitors self-select
• Twitter engagement in AI-interaction threads without direct product pitching
• Building in public with targeted replies rather than broadcast announcements
• 38,000 unique site visitors total; conversion driven by specificity of messaging

Tech stack

• Landing page: Next.js hosted on Vercel
• Desktop app: Electron + React + TypeScript
• AI integration: Vercel AI SDK — single abstraction layer connecting to all major LLM providers
• Hosting: Cloudflare (app assets); Vercel (landing page)
• Development tooling: Cursor (AI-assisted coding)

Key founder lesson

Praneeth spent five to six years building products that appealed to fellow developers rather than solving real customer problems. The mindset shift: stop optimising for peer approval and obsess over finding product-market fit — something people will pay for because it makes their lives measurably better.