roadmap

This replaces the March 2026 roadmap, which described a deleted architecture (Node.js microservices/MongoDB/Redis), missed every milestone (beta April, launch June), and targeted 100k users by Dec 2027. Old version: git history of this file.

Reality check — 2026-07-22 (verified, not aspirational)

• Users: 0. App never distributed — applicationId is still com.example.flow_app, debug signing, no keystore, no store/TestFlight presence.
• Waitlist: 1 signup since the site went live (~May 30).
• Instagram: 0 posts ever published. The digest tool works (11 PNGs generated) but the channel never started.
• Nobody noticed production was down. Supabase auto-paused (~Jun 29) and stayed down ~3.5 weeks — no monitoring, no traffic. Restored + keep-alive cron added 2026-07-22.
• What IS strong: clean architecture post-migration (analyze: 0 issues), 66-migration ledger in sync, 8 Edge Functions live, editorial design system shipped, l10n 39/43 screens, marketing site live, IG strategy documented, automated digest tool ready.

Diagnosis — where we were going wrong

1. Building depth-first on product, zero-first on distribution. ~6 months of feature work; the distribution channel (IG) that the strategy itself names as the wedge was never switched on.
2. Surface area >> team capacity. Tickets/payments, gamification, PR dashboards, matchmaking, event series, chat receipts… for zero users. Every ⚠️ Partial in the tracker is capacity spent without a user to feel it.
3. No validation loop. No analytics SDK, no distribution, no uptime monitoring. Nothing built can be learned from.
4. Planning against fiction. The two most strategic docs described a dead architecture; sessions kept “polishing” (5+ redesign passes) because no external signal ever forced prioritization.

Strategy (from Target Study)

Audience-first, one city, monetize transactions later: grow the IG channel now (zero cold-start), convert followers to a small real beta, build venue relationships via the /partners funnel — and only expand surface area when weekly retention in RE proves the loop.

North-star metric: weekly active users in Reggio Emilia. Not features, not followers.

90-day roadmap

Phase 0 — Switch on the channel (this week) — Elia

• Add SUPABASE_ANON_KEY secret in flow-platform repo (keep-alive goes live) + UptimeRobot free on site & REST endpoint
• Merge fix/ci-test-hygiene → develop (CI resurrection, ready & pushed)
• Create IG Business account + Meta token (guide in .env.example) → publish first weekend digest
• Cadence: 2 digests/week (Wed teaser, Fri full) — tool is automated, cost ≈ 0

Phase 1 — Make it distributable (weeks 2–4) — Federico mobile, Elia growth

• applicationId → it.chasetheflow.flow + release keystore + Firebase app re-registration
• Firebase Analytics (free) — events: open, feed_view, event_detail, rsvp
• Play Internal Testing + TestFlight with 20 real RE-based testers recruited from IG
• Fix only what testers hit; finish l10n gaps (15 notifier error strings, series_detail, 18 ES keys)
• IG: 1k followers RE-area target; every digest CTA → waitlist

Phase 2 — 100 real users (months 2–3) — both

• Contact 10 venues/organizers from the /partners page pitch: free listing + free PR list tool concept
• Event supply: ERT + IG Graph nightlife sources + manual curation of the top weekend picks
• Public beta in RE when W1 retention of testers ≥ 30%
• Measure: WAU, D7/W1 retention, RSVPs/user — decide next quarter from data, not vibes

Kill list (explicitly NOT now)

Payments/Stripe/QR tickets · badges-leaderboard expansion · PR & vendor dashboards data-wiring · matchmaking v2 · multi-city · any further redesign pass without user feedback.

Success gates (replace the old 10k/100k fantasies)

Gate	Threshold	Unlocks
IG channel	1k RE-area followers, ≥5% engagement	Beta recruiting works
Beta	20 testers, W1 retention ≥30%	Public beta
Public	100 WAU in RE	Venue monetization conversations

Related

Target Study · Feature Tracker · Instagram Strategy · 2026-07-22 audit changelog