7.5 KiB
🌟 NORTH STAR
One Company. One Brain. One North Star. Every decision, every feature, every research activity, every implementation is evaluated against this single objective.
The North Star
First paying SaaS customer onboarded and retained for 90 days.
Current Stage
Pre-revenue. Product is feature-complete for launch (155 tables, 159 API routes, 40+ modules, proven multi-tenant architecture with tenant provisioning tool). One user (Brendan's own shop, Cabinet HQ). The system has been built, tested on staging, and proven with real production data from Brendan's operations (~$78.4K monthly revenue).
What has been proven:
- Multi-tenant architecture with shared-schema + business_id + RLS — works
- Tenant provisioning tool — 16 verification checks pass
- PDF compliance service (all 8 AU states) — built, not deployed to fix template defaults
- Financial planning engine — live on Brendan's real P&L data
- Email intake with AI parsing — operational
- SMS notifications (Twilio) — working
What has NOT been proven:
- Anyone else will pay for it
- Onboarding pipeline end-to-end (SaaS Admin built, not deployed)
- Support model for non-Brendan users
- Multi-tenant performance under load
Why It Matters
Without a paying customer, BoundHQ is Brendan's custom JMS — not a business. A first customer:
- Validates the product-market fit hypothesis
- Opens revenue generation ($39-199/mo per business)
- Provides real feedback to shape the product
- Generates case study material for customer acquisition
- Proves the multi-tenant architecture works in production
- Validates the pricing model
BoundHQ targets 2,000+ AU cabinet and joinery shops. Customer #1 proves the model works.
Current Bottleneck
All three P0 blockers require Brendan action — not AI action.
| Blocker | Age | Impact |
|---|---|---|
| Login bounce (SameSite=Lax on auth cookie) — needs Brendan to test on staging | 14 days | Brendan cannot test — all dev slower |
| Xero scopes for Financial Stages 1-4 — Brendan enables in Developer Console | 11 days | Blocks SaaS Premium tier revenue |
| VPS drift — requires VPS access for git operations | 9 days | 13 modified files at risk |
The AI has completed or is blocked on every task that doesn't need Brendan. Progress depends on Brendan clearing these three blockers.
Success Metrics
| Metric | Target | Current | Owner |
|---|---|---|---|
| SaaS Admin deployed | Done | Not deployed | AI (unblocked) |
| Login bounce fixed | Verified on staging | Not tested | Brendan |
| VPS .env permissions set | chmod 600 | World-readable | AI (needs VPS access) |
| Git parity sweep complete | All changes committed | 13+4 files at risk | AI (needs VPS access) |
| PDF deposit clause compliant | All 8 states | 4 states illegal | AI (done, needs deploy) |
| VPS monitoring active | UptimeRobot + rate-limit | Flying blind | AI (unblocked) |
| First paying customer | $39+/mo | $0 SaaS revenue | Brendan |
| Customer retained at 90 days | Active + paying | N/A | Joint |
Current Priorities
Unblock the North Star
- Brendan: Test login fix on staging (~5 min)
- Brendan: Enable Xero scopes in Developer Console (~2 min)
- AI: Deploy SaaS Admin module (2 hours — ready to go)
- AI: VPS security sweep — .env permissions, git parity, monitoring (3 hours — needs VPS access)
- AI: Deploy PDF compliance template defaults (30 min)
- Brendan: Identify and approach first paying customer
Protect What Exists
- VPS backup verification — confirm backups are restorable (30 min)
- Login bounce — deploy fix, Brendan verifies (20 min)
- Credential rotation — rotate all exposed secrets (1 hour)
Things We Are NOT Doing
The following are explicitly deprioritised. They are not contributing to the North Star:
- ❌ New feature development — 155 tables, 40+ modules. Feature-complete for launch. New features delay customer #1.
- ❌ Mozaik CNC expansion — Only one potential user (Brendan's shop). Not revenue-critical.
- ❌ AI memory system expansion — v1 complete (37 files, 14 directories). Use it, don't expand it.
- ❌ Content marketing, social media, SEO — Zero customers to market to.
- ❌ Mobile apps — Desktop-first, mobile-compatible. Native apps are not the North Star.
- ❌ NZ/UK expansion — Australia first. 2,000+ AU shops unserved.
- ❌ Enterprise tier features — Target is 2-20 staff shops, not enterprises.
- ❌ AI model experimentation — Current stack works. New models don't move the North Star.
- ❌ Documentation perfection — "Good enough" documentation ships. Perfect documentation doesn't.
Founder Focus
Brendan's highest-leverage activities for the North Star:
- Test login fix on staging — unblocks all feature verification (5 min)
- Enable Xero scopes — unblocks Premium tier revenue (2 min)
- Grant VPS access for security sweep — unblocks 3 P0 security items
- Identify first paying customer — who is the ideal first customer?
- Review first customer onboarding — walk through SaaS Admin flow
- Decide on PDF compliance priority — legal risk vs. deploy delay
- Weekly North Star review — 2-minute priority check
Brendan's role: Think, decide, validate, sell. AI does everything else.
AI Focus
AI's highest-leverage activities for the North Star:
- Deploy SaaS Admin module — enables first customer onboarding
- Fix VPS .env permissions — secures credential exposure (needs access)
- Git parity sweep — commits all changes to Gitea (needs access)
- Deploy PDF compliance defaults — makes deposit clauses legal
- Set up VPS monitoring — UptimeRobot, rate-limiting
- Deploy login bounce fix — unblocks Brendan's testing
- Close completed handbacks — reduce noise from 21 active items
AI's role: Execute, remember, inform, maintain, document. Surface problems with solutions.
Review Cycle
| Frequency | Review | Participants | Output |
|---|---|---|---|
| Weekly (Mon) | North Star progress check | AI produces report, Brendan reviews 2 min | Blockers cleared? Customer pipeline? |
| Monthly | Customer health review | Brendan + AI | Revenue, churn indicators, feedback analysis |
| Quarterly | North Star itself | Brendan | Is "first customer" still the right goal? |
AI Decision Filter
Before beginning significant work, ask:
"Does this move the North Star?"
If No:
- Question whether the work should be done at all
- Explain why the work is being performed if proceeding anyway
If Yes:
- Make the relationship to the North Star explicit
- Prioritise above non-North-Star work
What Success Looks Like
30 days:
- Login bounce fixed, Brendan testing independently
- VPS security sweep complete (env perms, git parity, monitoring)
- PDF compliance deployed (all 8 states legal)
- SaaS Admin deployed, first tenant provisioned
- First paying customer identified and approached
60 days:
- Customer #1 onboarded and active
- Revenue: $39-99/mo recurring
- Feedback loop open — product shaped by real customer needs
- Premium AI tier ($99/mo) operational
90 days:
- Customer #1 retained (90-day active)
- Revenue trend established
- Acquisition channel validated
- Case study material ready for customer #2
This document is maintained by AI. Updated when the North Star changes or when significant progress shifts priorities. Never changed without Brendan approval.