WanderPins · Spring 2026

North Star,
feature backlog,
technical approach.

The plan for the next four months in three parts.

North Star

What WanderPins is, who pays, why it survives.

WanderPins is a personal travel journal with trusted sharing.

In a world where reviews, recommendations, and content are increasingly written by bots and generated by AI, WanderPins is deliberately the opposite — real recommendations from real people you actually know. Not the masses. Not the algorithm. Not a robot. A person whose taste you trust. That's the product. We use AI behind the scenes where it helps, but the core is and stays human.

The core loop

Discover a place

You find somewhere worth remembering — a pullout, a hot spring, a campground off a forest road.

Save it

Drop a pin. Add a photo, a note, the details that matter.

Share with someone you trust

Send it to a friend who'd actually go there — or just let it live on your map for them to find when they plan their trip.

They check in — "that was amazing"

Your friend makes the trip. They message you. That's the moment: you recommended something real, they acted on it, it landed.

They add it to their map

Now it's theirs too — and the next friend they share it with starts the loop again.

↻

The cycle repeats, and reputations compound.

There's an ego to this — and that's not a flaw, that's the engine. It feels good to be the friend who found the spot. It feels better when a friend goes there and confirms you were right. That payoff drives curation, sharing, engagement. Without it, WanderPins is a notebook. With it, it's a reputation.

Who pays

Adventure-curious with disposable income. 3–6 meaningful trips a year. Job, possibly a family. Already has a messy system — Google Maps lists, spreadsheets, pins in chat threads — and is willing to invest in tools that replace the chaos.

Model

Subscription. $4.99/mo · $29.99/yr. Free drives growth (create, follow, share, map, 10 stars, Scout Pins). Pro adds depth (unlimited stars, trips, batch import, export, future AI search). Launch Pro as a supporter tier.

Feature backlog

Nine releases to WanderPins 2.0 by Summer 2026. Each gets its own PRD.

#	Release	Deploy
01	Scout Pins Nearly done. Scraper cleanup + west coast NA seed. App-side complete.	MIGEDGEOTA
—	Web Refresh + Deep Links Can ship alongside any release. Growth mechanism.	WEBFULL
02	Pin Upgrades Manual pins, visibility tiers, smarter clustering, amalgamated card, AI names, data augmentation.	MIGOTA
03	Offline Mode & Export Cached pins, queued writes, GPX/JSON export (Pro). Bundle candidate with R2.	OTAFULL?
04	DB Refactor pin_visits table, linked_pin_id, link_type enum. Foundation for 7a/7b/8.	MIG
05	WanderPins Pro RevenueCat, paywall, Pro badge, starred pins gating. Bundle with R4.	FULLOTA
06	Strava / AllTrails Link external activities to pins. No OAuth — just URL entry + display.	MIGOTA
07a	Check-ins & Tags The social layer. Validation loop, friend tagging, notifications. Depends on R4.	OTAEDGE
07b	Grouped Pins Network-aware map clustering + amalgamated place card. Depends on 7a.	OTAEDGE
08	Trips Beta Link pins into trips. Free feature. Lightweight — no routes, no web, no covers (that's post-2.0).	MIGOTA
09	2.0 Onboarding Polished first-run for launch. Feature tour, invite flow, deep link follows.	OTA

Natural bundles

R4 + R5 — both infrastructure, no user-visible UI until later releases consume them
R2 + R3 — if offline scope stays small
Web Refresh — parallel to anything

Hard dependencies

DB Refactor (R4) → Check-ins (R7a) → Grouped Pins (R7b). Pro (R5) gates starred-pins enforcement and export.

Technical approach

Architecture principles, and an open conversation on how two people ship this.

Principles to preserve

These are the "don't foreclose this" rules as the product grows:

Content and coordinates stay separable. Notes, descriptions, photos are content. Coordinates and timestamps are metadata. They must never structurally entangle — keeps future encryption possible without a rewrite.
Intelligence runs client-side. AI features (search, display names, EXIF) operate on the device. Survives the day content gets encrypted server-side.
Grouping is a query, not an entity. No places table. Clustering at query time via PostGIS ST_ClusterDBSCAN. Pins are always individually owned, never merged.

Stack

React Native 0.81 · Expo SDK 54 · Expo Router 6 · TypeScript · NativeWind · Mapbox · Supabase (Postgres + PostGIS + Edge + Storage + Auth) · React Query · Sentry · RevenueCat (Phase 2) · Next.js web (Phase 3+)

Privacy stance

Starting point: we'll never run ads and we'll never sell user data. Everything beyond that — data export, client-side content encryption, local-only pins — is a direction we can move toward as the product matures. Worth a longer conversation about what a 2-person team can credibly commit to now vs. later.

How we push together — open for conversation

Shipping a mobile app with two people is a different problem than shipping it with five. The usual playbook assumes bandwidth we don't have. Worth figuring out together — not in this doc — what the right default looks like around:

Schema and migrations — hardest to unwind, worth more care
Cross-platform testing — Android regressions keep surfacing late
Branch and release hygiene — what's alive, what's dead
Agent discipline — diffs that touch more than intended, schema drift
Review gates — where a PR is worth it, where it slows us down

Brad's judgment on what's worked in previous teams is the highest-leverage input here.