Home Open the app
Case study

Aftershow.

Letterboxd for concerts: your full lifetime show history, no proof needed, now with the people you went with. Conceived at a Black Queen show at Thalia Hall. Built as a portfolio-grade product from ideation to a working app.

The problem

Concertgoers have no home for their history.

Fans who go to shows have nowhere dedicated to log setlists, rate performances, and build a lasting record of every concert they’ve attended. Generic social apps don’t serve this niche, and the closest direct competitor, Concert Archives, is data-rich and soul-poor: deep fields, zero identity, nothing that feels like it’s for music fans.

Three gaps stood out early. There was no assisted logging, so every entry is manual grunt work. There was no sense of the archive as identity, because competitors treat history as a spreadsheet, not something worth showing off. And there was no private-first social layer: everything defaults to public with no middle ground.

Insight

The archive is the product. The feed is the reward.

The North Star user isn’t a casual show-goer. It’s the die-hard fan with 100+ shows behind them, the type who becomes a community’s backbone once they have somewhere worth putting that history. Everything downstream follows from designing for that person first.

It started at home. My wife’s concert tracker, Concert Archives, felt more like a list than an experience: you read your history instead of reliving it. Aftershow is the version where the night comes back, with the setlist, the photos and the people you went with.

Three decisions

What shaped everything else.

Architecture

My Shows is Tab 1, not the feed

Every competitor leads with a feed or upcoming shows. Aftershow leads with your own archive, because the archive has to be seeded before the feed has any value. The feed unlocks after your first logged show.

Onboarding

One-show minimum, not three

A three-show threshold alienates people early in their concert journey. The bar is deliberately low: encourage more logging, never gate the product behind it.

Stack

Pivoted from React Native to a web-first build

The original plan was a native prototype. Rebuilt mid-project as a web app on Supabase and Cloudflare Pages: faster to ship a real, data-backed product than a native shell with no backend.

Four iterations

The fights.

Fight 1

Sako’s teardown

iOS developer Sako H. reviewed the first screens for buildability, not just looks. He flagged illegible field labels, elements floating on the background with no containers, and headings too weak to read as headings. Every fix became a hard rule, not a one-off patch.

Fight 2

Fourteen screens, one shell

The screens were built as fourteen standalone files, three of them megabytes from embedded images. Consolidating into one shell meant killing the bloat and building shared components once instead of copy-pasting them fourteen times.

Fight 3

Public vs. private-first

“Private-first social” was a named differentiator. The backend work forced the real call: fully public reads are simpler to ship, but quietly abandon that promise if chosen by default instead of on purpose. The answer was public by default with a one-switch private mode.

Fight 4

It looked like a prototype

The first complete build worked, but a phone-width column on a desktop screen and a heavy retro skin made it feel like a database with a costume. The redesign started from what the best social apps get right and wrong: a photo-first feed, a poster-grid archive, shareable recaps, and no ads or algorithm. Rebuilt around that, in one design system.

Product

What shipped.

Fourteen screens across five tabs, plus activity, comments and import, in one dark-primary, light-alternate app. A shared design system: Playfair Display Italic for artist names and headlines, Inter for UI, Crimson Pro Italic for what people write, DM Mono for dates and counts, and a single accent that shifts from gold in the dark theme to rust in the light one. The theme switch is an amp dial.

From-memory logging

Add any show ever attended: no ticket, no proof, no location check. Artist, venue, date, setlist as best remembered, and a rating, with a live preview of how it will look in your archive.

Archive-first profile

A poster grid with a diary view, four pinned favourite shows, a month-by-month map of your life in concerts, and ranked artists and venues.

A calm, chronological feed

Friends, everyone, and Tonight for live-logging. Photo-first posts with likes and comments. No algorithm, no ads.

Share cards and photo voting

Turn any show or your whole recap into an image for stories, and vote for the best fan photo of each show.

HTML / CSS / JavaScript modulesSupabase (Postgres, Auth, RLS, Storage), shared project, own schemaCloudflare PagesDemo mode (localStorage)
Status: what runs todayThe “Try it with sample data” demo runs Aftershow on sample data kept in your own browser, so nothing you do there is sent to a server. The real backend is connected to a live Supabase project (its own database schema, sharing a project with another app of mine) and was tested end to end with throwaway accounts: sign-up, privacy rules (a private profile’s shows stay hidden until someone follows), photo upload, likes, comments, photo voting and account deletion, which removes everything without touching the other app. It has not yet had a security review or any real-user testing, so real sign-ups are switched off on this site for now and I am not calling it production-ready.