Skip to content
← All work

iOS app · Solo · Aug 2026 to now

Lual: a calendar for friends, built and measured solo

A group calendar that shows when friends are free without showing what they're doing.

Role
Solo: design, mobile, backend, data, App Store releases
When
Aug 2026 to now (v1.0.6)
Stack
React Native (Expo), TypeScript, Supabase, Node.js/Express on Render, Gemini, StoreKit 2
Results
2,944 users · 1,756 monthly active · 36,056 events · ★ 4.9 (15 ratings)
  • React Native
  • TypeScript
  • Supabase
  • Node.js
  • Gemini
  • Lual home calendar showing all friends' plans in one month view
  • Lual groups screen with photo cards for each friend group
  • Lual AI schedule import turning a photo of a work schedule into calendar events
  • Lual plan screen showing the days when everyone in a group is free
  • Lual plans list with upcoming group plans and their status

Screens from the App Store listing. Scroll for more →

The problem

Making plans with friends usually means a group chat full of “when are you free?” Shared calendars fix that, but they show everyone everything. Lual shows friends when you're free, not what you're doing, unless you choose to share.

Part 1: How I built it

iOS appReact Native (Expo), widgets, StoreKit 2
SupabasePostgres + RLS, Auth, Storage, scheduled jobs
Express API on RenderAI import, quotas, receipt checks
Gemini
Apple Push
App Store Server
The app talks to Supabase directly, guarded by row-level security; anything needing secrets or third-party APIs goes through the Express API.

Data model and security

30 tables through 79 migrations, secured by 100+ row-level security policies. Privacy is the product: busy/free is the default, and the database, not the app, decides what a friend can see.

AI schedule import

Snap a photo of a work schedule or type “class MWF 10–11” and get calendar events back. The Express API sends it to Gemini with a fallback chain of models, enforces per-user weekly quotas and rate limits, and keeps the feature inside free-tier limits. 316 imports by 174 users so far.

Debugging story: late notifications

Users reported plan invites arriving late. Querying production data showed the retry job only ran while the Render server was awake, so notifications queued up while it slept. I moved retries into a scheduled database job that runs every 2 minutes, independent of the server.

Shipping

Apple and Google sign-in, encrypted session storage, StoreKit 2 subscriptions verified on the server, six home-screen widgets and iPad support, shipped through releases v1.0.0 to v1.0.6.

Part 2: How I measured it

A nightly SQL rollup writes one row per day to a daily_metrics table: active users, signups, onboarding completion, plans, invites and notifications. A private Next.js dashboard reads it. Everything below comes from aggregate queries against production, as of 2026-11-16.

Growth

Total signups and weekly active users by week. Active-user tracking started Aug 31.

The finding

Users who joined a friend or group within 24 hours came back the next day at 3.2× the rate of users who didn't (43% vs 13%). After week one, half of connected users were still opening the app, versus one in five unconnected.

Connected on day one vs. not

1,471 connected and 623 unconnected signups at least 7 days old.

How that compares

Across all 2,397 tracked signups, Lual keeps 10 points more users on Day 1 than the average iOS Lifestyle app (33.8% vs 23.6%), and stays above it at Day 7 (12.3% vs 9.6%).

Retention vs. the iOS Lifestyle average

Share of signups who opened the app on exactly day N (strict definition, cohorts since tracking began). Day 30: 5.2% vs 4.8% average. Benchmark: Fyresite category benchmarks (opens in a new tab); methods vary between sources, so treat it as a rough reference.

What I changed

I redesigned the first-session onboarding (v1.0.3, Sep 2) to push new users toward adding friends and joining a group before anything else.

What happened

Comparing 30 days of signups before and after the release:

Onboarding metrics before and after v1.0.3
Within 24 hours of signupBefore (520)After (660)p
Connected to a friend or group65.6%67.1%0.59
Added a calendar event49.0%57.1%0.0055
Sent an invite35.6%36.4%0.78

The redesign didn't measurably raise the connection rate, which was its goal. About two-thirds of new users were already connecting on day one before the change, so a new onboarding screen had little room to move them; the remaining third are the real target. First-day event creation did rise, but v1.0.3 also rebuilt AI import, so I can't credit onboarding for that.

Invite funnel

39% of invites were accepted; 589 people joined Lual from one.

A data-quality catch

Open tracking only started reaching users with v1.0.3, so the first two signup days show 0% Day 1 retention. I excluded pre-release cohorts from every retention number instead of reporting a fake “retention doubled.”

What I'd do next

Day 7 retention is flat around 12%. The next experiment targets solo signups: suggest contacts already on Lual during onboarding, and measure the connection rate for users who did not arrive by invite.

What users say

4.9 on the App Store from 15 ratings. Written reviews, unedited:

  • So Easy to Use!!!

    “Incredibly easy and understandable layout, simple enough for anyone to use. Already recommending this app!”
    talismanny · App Store review
  • Helps with planning a lot

    “Me and my friends are able to see eachothers availability an entire month in advance! Very nice”
    Matt BMA · App Store review
  • safe for teens

    “safe for teens! very easy to use and i recommend as a teen!!”
    arielsk9621 · App Store review

See Lual on the App Store (opens in a new tab)