Write down the release plan: stores, money, social, paid ads

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Author: Kjubikstronk

CLAUDE.md

@@ -159,6 +159,9 @@ device testing.
 3. The Maps budget is €2. It would be exhausted fast at scale, and the map then
    silently dies for everyone.
 
+**Full release plan** (store packaging, costs, organic social, TikTok and
+Instagram ads): `docs/release-plan.md`.
+
 Decided since: reports stay readable by everyone for now; later they go to
 a developer-only inbox. Worth doing at release:
 - **Maps (decided):** Google while private (free allowance covers it). At

docs/release-plan.md

@@ -0,0 +1,153 @@
+# Release plan
+
+Planning notes for taking "our dates" public. Nothing here is built yet.
+Figures are rough estimates from 2026-09; check current prices and terms
+before spending anything.
+
+---
+
+## 1. Order of work
+
+1. **Sign-up + partner invite flow, privacy page, in-app account deletion.**
+   Needed for both stores and for strangers at all. The invite screen is the
+   most important screen in the launch: every user must bring a partner.
+2. **Own domain** (~€10/year). Required for the Android wrapper, and the
+   landing page lives there.
+3. **Android first** via a TWA (cheap, fast, learn from real users).
+4. **iOS** via Capacitor, together with native push notifications.
+5. **Free maps** (see CLAUDE.md, "Maps (decided)") before any real marketing
+   push — Google's bill scales with every user.
+6. **Organic social**, then a small paid test. Paid ads last, and only once
+   people stick around.
+
+---
+
+## 2. Packaging the web app
+
+**Android — Trusted Web Activity.** The PWA runs full screen inside a Play
+Store app. Generate with PWABuilder or Bubblewrap. Needs Digital Asset Links
+(`/.well-known/assetlinks.json`) at the domain root, which is why the GitHub
+Pages subpath won't do. Play developer account: $25 once.
+
+**iOS — Capacitor.** Apple rejects thin web wrappers (guideline 4.2), so the
+app must use real native features: push notifications (also better than web
+push, which only works from a home-screen install on iOS), share sheet,
+haptics, later a "next date in 3 days" widget. Apple developer account: $99
+a year. Builds need a Mac, or a cloud Mac (Codemagic, GitHub macOS runners).
+
+Capacitor can produce the Android build too, if one path for both is simpler.
+
+**Store requirements:** privacy policy URL, in-app account deletion (Apple),
+Data safety form (Google), age rating, screenshots, a demo account for
+reviewers. Check the name "our dates" isn't taken in either store.
+
+**Old iPhones:** Tailwind v4 needs Safari 16.4+. Older iOS renders broken.
+
+---
+
+## 3. Money
+
+| Item | Cost |
+|---|---|
+| Google Play account | $25 once |
+| Apple developer account | $99 / year |
+| Domain | ~€10 / year |
+| Free maps | €0 until thousands of users, ~€25–50 / month at tens of thousands |
+| Firebase | Free plan until roughly tens of thousands of users, then pay-as-you-go |
+
+**Income:** no in-app ads (they'd wreck the pixel design, pay very little at
+small scale, and drag in tracking-consent work). Instead, **themes as a
+supporter perk**: pink stays free; seasonal and extra themes unlock for
+€1–2/month or one-off. Roughly 1 in 500 users paying covers the maps.
+
+---
+
+## 4. Organic social (TikTok + Instagram) — start here
+
+Costs €0 and shows which angle works before any money is spent.
+
+**Setup:** one handle on both platforms, ideally the app's name. Bio link to
+the landing page (which links to both stores).
+
+**Cadence:** 3–5 posts a week for 4–6 weeks. Post the same video to TikTok
+and Instagram Reels.
+
+**Format:** vertical 9:16, 7–15 seconds, hook in the first second, text on
+screen (most watch muted), a trending sound, the app shown on a real phone
+in hand rather than a flat screen recording.
+
+**Content pillars:**
+
+1. **The story.** "I built my girlfriend an app for our dates." Building
+   something for someone is one of the strongest formats there is; it gives
+   the account a face and a reason to follow.
+2. **The blind rating reveal.** Each of you rated the date separately;
+   show both scores appearing at once. The app's most unusual feature.
+3. **Date idea lists.** "5 cheap dates in Amsterdam" with the pixel map
+   filling up with pins. Useful, so people save and share it.
+4. **Theme drops.** The Halloween switch with the bats and the greeting.
+   Every new theme is a new post, and a reason for old users to come back.
+5. **Relatable couple humour.** "Where do you want to eat?" / "I don't
+   know" → opens the app.
+
+**What to measure:** saves and shares matter more than views; profile visits
+and link clicks show real interest. Keep a simple list of which posts did
+best, since those are the ones to pay to boost.
+
+**Micro-creators:** couple accounts with 5k–50k followers often beat ads.
+Offer a free supporter subscription, or roughly €50–200 per post. Paid
+posts must be labelled as ads (EU rules and both platforms' terms).
+
+---
+
+## 5. Paid ads — TikTok + Instagram
+
+**Don't spend a cent until retention is proven.** If most couples still use
+the app after 30 days, ads multiply that. If not, ads buy people who leave.
+
+### Measuring without trackers
+
+Ad-platform tracking SDKs and pixels need consent under GDPR (and Apple's
+tracking prompt on iOS). Start without them: use **App Store Connect
+campaign links** and **Google Play UTM links**, one per ad. The store
+reports installs per link with nothing added to the app. Count "couples
+where both partners joined" and "still active after 7 / 30 days" from our
+own database.
+
+### Phase 1 — test (2 weeks, ~€100)
+
+- **What:** boost the 3 best organic posts, not new ad creatives. TikTok
+  **Spark Ads**; Instagram **boosted posts**. Proven content is the cheapest
+  reach you can buy.
+- **Who:** ages 18–34, only countries where the app works (start NL, AT,
+  DE), interests around relationships, dating, date ideas, going out.
+- **Spend:** €3–5 a day per platform.
+- **Goal:** app installs (or landing-page visits before the store launch).
+
+### Decide
+
+Rough cost per install for lifestyle apps in Western Europe: €1–5, more on
+iOS than Android.
+
+- **Stop** if cost per install is over ~€4, or fewer than ~30% of couples
+  are still active on day 7. Fix the product or the videos first.
+- **Continue** if cost per couple is low and they stay.
+
+### Phase 2 — scale (only if the test worked)
+
+- Raise budget on winning ads by 20–30% a week; big jumps reset the
+  platforms' learning.
+- Make new videos every 2–3 weeks — the same ad wears out.
+- Cap total spend until income covers it: a user must bring in more than
+  they cost to acquire. Remember each paid install brings a partner free.
+
+### Seasonal pushes
+
+| Moment | Organic from | Paid push |
+|---|---|---|
+| **Valentine's Day (14 Feb)** — the best launch moment | early January | 1–14 Feb |
+| Halloween theme drop | late September | October |
+| Christmas / winter theme | mid November | December |
+
+Launching the store apps just before Valentine's Day gives a natural hook:
+"plan your Valentine's date together".