💡 Customer Pre-Market Demand Signals for Vendor Preparation

[github-actions[bot]]

Summary

Let customers tap a lightweight "I want this" button on vendor product listings before market day, giving vendors aggregated demand signals to inform what and how much to bring. This is NOT e-commerce — no payment, no fulfillment obligation. Think Instagram "Remind Me" rather than a shopping cart.

Market Signal

Pre-ordering is the fastest-growing feature in local food platforms (Homegrown at $10/mo, FarmStand, Cottage CMS all launched or expanded in 2025-2026). Research shows pre-orders reduce vendor waste by letting them "know exactly how many items to bring." However, full e-commerce is intentionally deferred in Markets' PRD. This demand-signal approach captures the core value — vendors know what customers want — without the operational complexity of order fulfillment, payment processing, or inventory management.

User Signal

The PRD identifies Brad (vendor persona) needing tools that "increase his discoverability and repeat customer flow." Rachel (customer persona) wants confidence before traveling. Demand signals serve both: vendors prepare better, customers see aggregated interest as social proof. Discussion #372 (Reserve-at-Market Intent Signaling) touches on at-market reservation, but this addresses pre-market demand forecasting — a distinct and complementary capability.

Technical Opportunity

Architecture already includes Apollo Client with optimistic UI, Firebase Realtime Database for status propagation, and FCM push notifications. Demand signals would use the same event-driven pipeline: customer taps "interested" → domain event → vendor sees aggregated demand before market day. No new infrastructure required — fits cleanly into the existing internal/events/ → internal/notify/ → internal/realtime/ pattern. Data model: (customer_id, vendor_id, product_category, market_date, created_at).

Assessment

Dimension	Score	Rationale
Feasibility	high	Reuses existing event-driven architecture; simple data model; no payment integration
Impact	high	Directly serves both personas (Brad gets demand intel, Rachel gets social proof); reduces vendor waste
Urgency	med	Not blocking MVP, but high strategic value as a differentiation from competitors who jumped to full e-commerce

Adversarial Review

Strongest objection: Low-commitment signals may generate noise without actionable data. If 50 people tap "interested" in strawberries but only 5 show up, vendors lose trust in the signal.

Rebuttal: The signal is explicitly framed as interest, not commitment. Vendors see it as "demand temperature" alongside their own experience. Start with aggregated counts only (e.g., "12 people interested in strawberries this week") — no individual identity exposed. Over time, the system can calibrate by correlating signals with actual attendance, building a conversion baseline. Even imprecise demand signals beat the current zero-information baseline.

Suggested Next Step

Design the demand signal data model (customer_id, vendor_id, product_category, market_date) and the vendor-facing aggregation view. Prototype as a single "Interested" button on the vendor/product detail screen.