# Hard Questions for Michael *Generated from deep synthesis of USER.md, SOUL.md, TODO.md, COMPANY.md, work logs, and daily memory files.* --- ## Strategic (The Big Stuff) ### 1. You're building RateRight while working 12-hour construction days — but what's the actual timeline for leaving the tools behind? You've been at this for months. The platform is 85% done but not launched. Your father is funding this. At what point do you draw the line and say "I'm all-in on RateRight" or "this isn't working"? Because right now you're paying the price of both lifestyles without the upside of either. ### 2. You sacked the team until Growth Engine is fixed — but is the problem the team, or that you don't actually know what you're building? The team was messy, sure. But you've gone through multiple developers, multiple approaches, and still don't have a working growth system. At some point you have to ask: do I actually understand what I need well enough to hire for it? Or am I hoping someone else will figure out my business for me? ### 3. You've got 231 leads in Growth Engine and 91.8% are untouched — why are you still trying to find MORE leads instead of converting the ones you have? This is classic avoidance. New leads feel like progress. Following up on old ones feels like rejection. But math doesn't lie: 212 untapped leads > 0 new leads. What's the real reason you're not working the existing pipeline? ### 4. The $50 fee model sounds great to contractors — but have you actually validated that workers will sign up for this? You're optimizing for the customer (contractors) but workers are the supply side. If workers don't trust the platform, don't check it, or don't see value — you have no marketplace. When's the last time you talked to an actual worker who isn't Rory? ### 5. You're waiting on ABR GUID approval (1-5 days), Stripe webhook secret, Supabase avatars bucket, key rotation, git history cleanup — at what point does "waiting on X" become an excuse for not shipping? Every founder has a list of blockers. But looking at your TODO.md, half the "blockers" are 30-minute tasks you've been sitting on for days or weeks. Are these really blockers, or are they comfortable reasons to not face the harder question: "What if I launch and nobody cares?" --- ## Operational (The Day-to-Day) ### 6. You told Rivet to "forget about hot leads" because the team is sacked — but the leads are still there, still hot, still uncontacted. Who's actually responsible for revenue right now? There's no team. There's no sales process. There's just you, on a construction site, hoping the website will magically generate revenue. Do you have a plan for getting your first paying customer, or are you hoping it'll happen organically? ### 7. Your voice calls with Rivet "ring but don't say anything" — you've known this for hours, but you haven't fixed it. Why is operational brokenness acceptable to you? This is a pattern. Things break. You note them. You move on. But the voice call is your primary interface with your AI COO. If you tolerate broken tools, what does that say about how you'll tolerate a broken business? ### 8. You want Rivet to ask "hard questions" — but will you actually answer them honestly, or will you deflect like you do with the timesheet and lead follow-ups? You asked for this. But every time Rivet pushes on something real (the stale brief, the hot leads, the voice calls), you either give a partial answer or change the subject. Are you actually open to being challenged, or do you want the illusion of accountability? ### 9. You've got two websites (the $50 app + Growth Engine), two different tech stacks, two different teams — are you building one business or two hobbies? RateRight should be one thing. But you're maintaining two separate codebases with different deployment processes, different auth systems, different databases. Is this strategic separation or just the result of not deciding what the business actually is? --- ## Personal (Your Patterns) ### 10. You work 5:30am to 6pm on site, then 6:30pm to 11pm coding — but when's the last time you took a full day off without guilt? This schedule is unsustainable. You know it. Everyone knows it. But you keep pushing because stopping feels like failure. What are you actually trying to prove, and to whom? ### 11. You keep asking for "natural conversation scripts" and better voice tones — but you haven't defined what success looks like for Rivet. You want her to sound less gruff. You want her to ask hard questions. You want transcripts. But what's the actual goal? Are you building an AI COO that helps you run the business, or are you tinkering with a fun side project because the real work (sales, marketing, hiring) feels harder? ### 12. You told the team they were underperforming and sacked them — but have you looked at your own output this week? Site work: 7.5 hours (heat stoppage). RateRight work: scattered voice messages, one signup bug fix, some cron updates. Are you holding yourself to the same standard you're holding others to? --- ## Technical (The Platform) ### 13. The signup bug Rory hit was a double-authentication error — but why didn't you catch this in testing before he did? You're about to launch (supposedly) but your onboarding flow breaks for new users. This is basic QA. Do you have a launch checklist, or are you just hoping things work when real users show up? ### 14. You've got 7 open tasks marked "blocked on Michael" — Stripe keys, Supabase bucket, .next ownership, cron fixes. These are 10-minute jobs. Why are they still open? You're waiting on yourself. That's the blocker. What's the fear that's keeping you from just doing them? Is it that once they're done, you'll have to face the actual hard work of getting customers? ### 15. You want OpenRouter so Rivet can "use whatever model is best" — but you haven't optimized what you already have. You've got Kimi, Claude, DeepSeek, and ElevenLabs configured. But half the time Rivet spawns the wrong model for the task, or Claude Code fails, or the voice calls don't work. More tools won't fix broken workflows. Have you actually mapped out what should happen when, or are you hoping more options will solve prioritization for you? --- ## The Meta Question ### 16. If RateRight fails, what will be the reason? And are you doing anything to prevent that today? Be honest. Is it lack of technical execution? (No — the platform is nearly done.) Is it lack of market need? (Unlikely — construction hiring is huge.) Is it that you never actually launched because you kept finding reasons to wait? The pattern is clear: you build, you refine, you audit, you fix. But you don't ship. You don't sell. You don't put it in front of real customers who might say no. **What's the real blocker?** --- *These questions are designed to be uncomfortable. A good COO asks them. What you do with them is up to you.*