Testerly

The only platform that teaches you to validate products by practicing on others, then connects you with pro testers when you scale

Comment

The idea for Testerly was born from a frustrating pattern I kept witnessing in the builder community. Every week, I'd see talented creators in Bolt.new Discord channels, IndieHackers forums, and hackathon Slack groups sharing the same story:

"I built this amazing app in 3 days using Bolt.new and AI, but I can't get anyone to try it. How do I know if it's actually good?"

The irony was striking. We've democratized creation - anyone can now build sophisticated web apps with tools like Bolt.new, Supabase, and ChatGPT. But we've completely ignored the validation problem. These builders had gone from idea to MVP in days, but were stuck for months trying to get their first 10 real users. What I Learned: The Validation Gap Through conversations with 50+ novice builders, I discovered a painful reality:

90% fail not because of technical issues, but because of product validation They have $0 marketing budgets (students, side projects, first ventures) Existing solutions like BetaTesting.com cost $2,500+ - completely inaccessible Friends & family give biased feedback; Reddit posts get ignored Most critically: They don't know how to validate properly

But here's what surprised me: They were hungry to learn. When I showed them successful app onboarding flows or pricing strategies, their eyes lit up. They wanted to understand what made products successful, not just build features. That's when the "Learn by Testing" model clicked. What if validation became education? How I Built Testerly: From Complex to Simple Initial Architecture (The Over-Engineering Phase) My first instinct was to build a sophisticated marketplace with:

Complex AI processing pipelines Automated tester matching algorithms Advanced reputation systems Multiple payment tiers Real-time chat systems

I spent weeks designing database schemas with 15+ tables, planning microservices, and researching machine learning models for review quality scoring. Then reality hit: I was solving the wrong problem. The Pivot: Manual + Automated Hybrid The breakthrough came when I realized I was falling into the same trap as my target users - over-engineering before validation. I needed to prove demand first, automate second. The new approach: Phase 1: Manual AI reviews (me + ChatGPT) → Validate demand Phase 2: Semi-automated processing → Scale gradually
Phase 3: Full automation → Scale efficiently Technical Implementation Journey Stack Evolution:

Started with: Complex Next.js + tRPC + Prisma + Redis + multiple AI APIs Ended with: Bolt.new + Supabase + Stripe (3x faster development)

Key Technical Decisions:

Bolt.new as MVP Platform typescript// Why Bolt.new worked perfectly: // - Rapid prototyping (2-3 weeks vs 2-3 months) // - Built-in deployment // - Perfect for validation phase // - Easy migration path when ready to scale

Simplified Database Schema sql-- From 15 complex tables to 5 essential ones: users, projects, peer_reviews, paid_reviews, payments -- Focus: Core functionality first, optimization later

Manual Processing First typescript// Admin dashboard for me to process reviews manually // Generate AI prompts → Copy to ChatGPT → Format responses // 20 minutes per review vs weeks building automation

The Challenges That Shaped the Product Challenge 1: Legal Compliance 🚨 The Problem: My initial idea involved review exchanges, which violate App Store policies. The Solution: Pivoted to educational framework:

"Learn by analyzing successful products" (legitimate) "Practice validation methodology" (valuable) "Get feedback from fellow learners" (community-driven)

This wasn't just a legal fix - it became our competitive advantage. Education creates stickier users than transactions. Challenge 2: Cold Start Problem The Problem: Marketplaces need both sides (reviewers + projects) to launch. The Solution: Curated learning content first: javascript// Phase 1: Curated analysis targets const learningProjects = [ { name: "Notion", category: "productivity", lessons: ["onboarding", "freemium"] }, { name: "Duolingo", category: "education", lessons: ["gamification", "retention"] } ]; // Users learn by analyzing these, then submit their own Challenge 3: Proving Value Without Scale The Problem: How do you demonstrate a marketplace's value with 0 users? The Solution: Started with the learning experience:

Created detailed analysis templates Wrote example reviews showing the depth of insights Built the full UX flow with demo data Showed exactly what users would learn and receive

Challenge 4: Pricing Psychology The Problem: How do you price education vs. services? The Learning: Through user interviews, I discovered:

$0-9: 85% willing (students, side projects) $4-8: Sweet spot for "professional but accessible" $25+: Only for funded startups

The Solution: Freemium with micro-transactions:

Free: Peer learning and reviews $4: AI simple analysis $8: AI comprehensive review Future: $200 influencer reviews

What I Learned About Building for Builders

  1. Builders Want to Learn, Not Just Launch The biggest insight: Novice builders are more motivated by learning than by quick wins. They want to understand why successful products work, not just get generic feedback.
  2. Community Beats Automation (Initially) My peer review system creates more value than I expected:

Users learn by reviewing others They build relationships with fellow builders Quality feedback emerges from shared learning goals Network effects grow organically

  1. Mobile-First for a Desktop-Heavy Audience Counterintuitive discovery: Even though builders work on desktop, they consume learning content on mobile. The review interfaces needed to work perfectly on phones.
  2. Timing Trumps Features Rather than building every feature, I focused on timing alignment:

Bolt.new hackathons happening monthly IndieHackers community growth No-code adoption acceleration AI democratizing development

The Current State: 95% Complete MVP What's Working:

✅ Landing page with 4.9/5 conversion messaging ✅ Clerk authentication with session persistence ✅ 3-step onboarding (project type selection) ✅ Learning hub with analysis templates ✅ Peer review submission system ✅ Dashboard with conversion CTAs ✅ Pricing flow for AI upgrades ($4/$8)

What's Still Manual:

AI review processing (me + ChatGPT for now) Project curation for learning hub Quality control for peer reviews Customer support and onboarding

Lessons for Other Builders

  1. Start Manual, Automate Later Don't build automation until you prove demand. My manual approach let me:

Validate the business model in weeks vs months Understand user needs deeply through direct interaction Iterate on pricing and positioning quickly Build cash flow before building complexity

  1. Leverage Your Constraints Bolt.new's limitations forced better decisions:

Simple architecture = faster iteration Memory constraints = focus on core features JavaScript-only = creative solutions over complex ones

  1. Community-First Go-to-Market Instead of paid acquisition, I'm targeting:

Bolt.new Discord community (50K active) IndieHackers (direct value alignment) Hackathon organizers (timing synergy) No-code Twitter (existing relationships)

  1. Legal as Competitive Advantage The App Store compliance "constraint" became our differentiator. While competitors risk policy violations, our educational approach is:

Completely compliant More valuable for users Harder to replicate Better for long-term brand building

What's Next: The Growth Plan Immediate (Next 30 days):

Launch in Bolt.new community Get first 50 users through peer reviews Process first 10 paid AI reviews manually Iterate based on user feedback

Short-term (90 days):

Prove unit economics ($4-8 AI reviews) Build waiting list of 500+ users Develop relationships with hackathon organizers Create viral learning content

Long-term (12 months):

Semi-automate AI review processing Add influencer review tier ($200) Expand to mobile app validation Launch referral and community programs

The Vision: Transforming Builder Success Today, 90% of novice builders fail at validation. My goal is to flip that to 60% success through practical education and accessible professional feedback. Testerly isn't just a service - it's a methodology. We're teaching builders to think like product people, validate like entrepreneurs, and iterate like professionals. The future I see: A world where every builder has access to the validation skills and feedback that used to be exclusive to funded startups and enterprise teams. That's the impact I'm building toward. One review, one lesson, one successful product at a time.

Built With

Share this project:

Updates