# Strategic Review - 2026-04-08

## 1. PRODUCT BLIND SPOTS
- **Availability Decay**: The voice profile builder is slick, but worker availability in construction changes daily. If a contractor pays $50 and the matched worker is already on another site because their profile is 3 days stale, trust is instantly destroyed. We have no "available tomorrow" ping system.
- **Dispute Resolution is Offline**: We deployed "Report Issue" buttons, but Cog (Ops) is shut down for budget. If a Day 1 match fails, the complaint goes into a void. The contractor will chargeback the $50 on Stripe and never return.
- **WhatsApp Gateway Fragility**: The 440 session conflict error isn't just an inconvenience; it completely breaks the core communication loop (as seen with Jo's timesheet). If we rely on WhatsApp for worker/contractor notifications, we are dead in the water at scale.

## 2. BUSINESS MODEL STRESS TEST
- **The One-and-Done Flaw**: The $50/hire model with "unlimited duration" explicitly encourages disintermediation. Once a contractor pays $50, they have the worker's number. They will never pay us again for *that* worker. To survive, we need contractors to constantly experience high turnover or scaling events, or we need to pivot to a subscription for pipeline access. 
- **Scale Breaking Point**: The current reliance on WhatsApp Web sessions tied to individual agents cannot scale beyond a few dozen concurrent conversations. Ghost processes causing crash loops will take down the platform under moderate load.

## 3. COMPETITIVE LANDSCAPE
- **Yakka is Here**: "Labour Jobs + Hire | Yakka" is operating nationwide in Australia (urban and regional) doing exact on-demand construction hiring with verified workers, ratings, and smart notifications.
- **Our Edge vs Theirs**: We compete on the $50 flat fee. But Yakka has liquidity and reliability. If our WhatsApp gateway crashes or our workers don't show up because of stale availability, the $50 discount won't matter. Contractors pay for certainty.

## 4. LAUNCH READINESS (HARD BLOCKERS)
1. **Susan's Crash Loop**: Contractor outreach is halted. Susan is stuck in a MEMORY.md truncation and WhatsApp 440 error loop. 
2. **Stripe Test Mode**: MASTER-STATE-OF-PLAY confirms Stripe is still in test mode. We literally cannot accept real money yet.
3. **Google Workspace Auth**: Cold email execution requires local GWS bridging that Michael hasn't pushed. Susan has no outbound email capability.
4. **Domain Name**: The production app is still on `rivet.rateright.com.au` instead of the root domain.

## 5. OPPORTUNITIES
- **Rainy Day Outreach**: "Heavy rain days wash out sites, meaning contractors are stuck in offices." This is a massive tactical advantage. We should tie Susan's cold outreach volumes directly to the BOM (Bureau of Meteorology) weather forecast. 
- **Physical Liquidity First**: Cold email is blocked technically, but the worker QR voice funnel is QA passed. We should aggressively execute the physical hostel poster strategy *now* to build worker supply while Builder fixes the contractor demand channels.

## 6. OVERNIGHT WORK PLAN
**Builder (Priority 1):**
- Kill the ghost WhatsApp processes causing the 440 crash loops (requires root/systemctl override if possible, or script a safe kill switch).
- Fix Susan's MEMORY.md truncation error.
- Prepare the Stripe Live Mode keys and DNS records (root domain swap) for Michael's 90-minute evening window to execute.

**Rivet (Priority 2):**
- Draft a BOM-weather-triggered cron job specification for Susan's outreach.
- Prepare the exact GWS connection commands Michael needs to run locally to unblock email outreach.
- Queue Harper to format the 3 alternative grant options for the morning brief.