💡 Product-Level Following & Availability Alerts

[github-actions[bot]]

Summary

Enable customers to follow product categories (e.g., "organic eggs", "honey", "sourdough bread") across all markets and vendors, receiving notifications whenever any vendor with matching products checks in nearby. This creates a fundamentally new discovery mode — product-first rather than vendor-first — that directly serves the core user journey described in the PRD.

Market Signal

DoorDash's Zesty app (piloting in SF Bay Area and NYC) validates AI-powered hyper-local food discovery as a viable pattern. No farmers market app currently supports product-level following — all rely on vendor-level or market-level subscriptions. The PRD explicitly identifies "find item → vendor → market path" as a growth feature, and Customer Journey #1 begins with Rachel searching "organic eggs." The broader trend of product-first discovery (vs. store-first) is well-established in e-commerce and now reaching local food.

User Signal

The existing idea #372 (Reserve-at-Market Intent Signaling) and #370 (What's Fresh Today Broadcasting) both hint at product-level demand, but neither closes the loop of "customer wants X → system finds X → customer gets alerted." This fills the gap between vendor broadcasting and customer discovery. The PRD's primary user journey begins with a product search, not a vendor search — the architecture should support this pattern natively.

Technical Opportunity

The architecture already supports vendor product catalogs (FR12), follow relationships (FR25/FR26), and push notifications via FCM (FR29). Product-level following extends these existing patterns with a curated category taxonomy (15-20 categories) rather than free-text matching, avoiding NLP complexity. The existing event bus (internal/events/) can route check-in events through a product-category filter to reach interested customers. The on-device produce recognition idea (#72) could later enrich this with finer-grained matching.

Assessment

Dimension	Score	Rationale
Feasibility	med	Extends existing follow/notification infrastructure with a new entity (product category). Requires a curated taxonomy and a new notification routing path, but no novel technical challenges.
Impact	high	Directly serves the primary customer value proposition ("I want eggs — tell me when I can get them"). Creates a discovery mode no competitor offers. Strengthens the vendor check-in → customer notification flywheel.
Urgency	med	Most valuable post-MVP when vendor catalog data and check-in flows are established. Can be designed now and implemented in growth phase.

Adversarial Review

Strongest objection: Product taxonomy normalization is hard. Vendors name products inconsistently ("free range eggs" vs "pasture-raised eggs" vs "farm eggs"). Without normalization, product-following produces noisy, unreliable results that erode trust.

Rebuttal: Start with a curated product category taxonomy (15-20 categories: eggs, honey, leafy greens, berries, baked goods, etc.) rather than free-text product matching. Customers follow a CATEGORY, vendors tag their products with categories from a controlled list. This is simple, extensible, and requires no NLP. The on-device produce recognition idea (#72) could later enrich this with finer-grained matching.

Suggested Next Step

Define the initial product category taxonomy (15-20 categories) covering the most common farmers market product types. Create a spike story to validate the notification routing: when a vendor with category X checks in, notify all customers following category X within a configurable radius.

---

Generated by the weekly feature ideation workflow on 2026-08-07.