Solo Project

0 —> 1 Product Build (Vibecoded)

Jul 2026

Family group chats have spent years optimizing how we report our lives to each other. Updates, photos, "what time are you coming home" — the messages keep everyone informed, but they rarely change whether a family actually spends meaningful time together.

This case study covers how I identified a pain point in my own family, designed a game mechanic around it, vibecoded a working native app, and what shipping it to four real users taught me about building products.

—> Kept running on iOS — try the web demo (no signup required)

How I turned my family's group chat guilt into a game we actually want to play

My family's group chat had a rhythm I suspect many families recognize: I'd forget to post updates, get nagged for it, post something obligatory, repeat. The chat kept us informed but not connected.

The deeper problem showed up whenever we actually gathered. Holidays and breaks kept falling flat — everyone arrived with individual plans already made, and we could never reach consensus on what to do together.

Then one day I realized I couldn't name my sisters' current hobbies. Or my parents'.

We hadn't grown apart. We'd just stopped having a reason to overlap.

That observation led me to a simple question:

What would make a family look forward to time together, instead of just scheduling it?

Why did being in a family group chat feel like a chore?

PAIN POINT & INSPIRATION

Kept is a shared bingo board for families. Instead of coordinating logistics like a calendar or reporting status like a group chat, it introduces a new objective: filling a board built from each family member's real hobbies, trip ideas, and bucket-list moments — together.

Each family builds its own board. A square might be "Sunday hike with Dad," "Try Yena's pottery class," or "Beach day before summer ends." When the family does the thing, the square is completed — with a date, and eventually a photo — and the board slowly becomes a record of memories made.

Although the bingo mechanic defines the core experience, Kept is deliberately simple: create a family, build a board, complete squares. I built v1 with nothing else, on purpose.

What is Kept?

THE GAP

Add your hobbies

Each family member contributes what they love

Build the board

Squares are generated from everyone's interests + shared trip ideas

Gather & Play

At the next family hangout, pick a square

Golf 18 holes

with Dad

Karaoke Night at Alyssa's favorite bar

Group workout at Mom's pilates studio

Go to Alice's new favorite brunch spot in Hannam-dong

Complete a square

Mark it done, with the date and a memory

Watch the board fill

The family record grows; plan the next one

PRODUCT PSYCHOLOGY

Why Kept works: How a bingo board changes what family time means

At first glance, Kept looks like a checklist with a grid.

But the mechanic produces several subtle shifts in how a family behaves:

Reporting → Anticipating

A group chat asks "what did you do?" A board asks "what should we do next?" The direction of attention flips from past-tense obligation to future-tense anticipation.

Individual plans → A shared menu

Gatherings used to fail at the consensus step — everyone proposing from scratch, nobody agreeing. The board pre-loads consensus: every square was contributed by a family member, so anything you pick is something someone already loves.

Strangers' hobbies → Shared experiences

Because squares come from each person's actual interests, completing the board requires learning each other's hobbies and trying them. Reconnection isn't a feature; it's the win condition.

Forgettable hangouts → A kept record

Completed squares accumulate into something a family can look back on. The name is the thesis: memories, kept.

Vibecoding v1 — and the decisions the AI didn't make

HOW I BUILT IT

I built Kept on Replit using AI agents, going from idea to working native app in days. The AI wrote most of the code. The product decisions were mine:

I scoped v1 to a single family: my own. Five real users whose gatherings I could directly observe. No accounts system for the public, no friend groups, no features for hypothetical users — I wanted real usage evidence before building anything else.

I hand-wrote the seed content. The squares are the product. "Sunday hike at [trail] with Dad" motivates in a way "outdoor activity" never will, so I treated board content as a product decision, not filler data.

I overrode the AI where gamification defaults pointed the wrong way. When I described a "family bingo game," the agent reached for the standard gamification kit: individual progress, points, a leaderboard. I stripped that out and rebuilt it as a single shared board with no per-person scores — because the moment family time becomes a competition, someone loses, and the entire point of Kept is that the family wins together. It was a small technical change and the most important product decision in the app.

WHAT REAL USAGE TAUGHT ME

Four users, one board, one summer

I seeded the first board with things my family had actually done — and wanted to do — this summer. Watching them use it taught me two things I couldn't have predicted from design alone:

Proximity beat ambition.

The first squares to complete weren't the exciting far-away trips. They were the local ones — the squares my parents contributed: a nearby hike, a neighborhood spot, a Sunday afternoon plan. The big travel squares sat untouched, not because nobody wanted them, but because they required planning that a casual weekend couldn't absorb. The board's momentum came from squares a family could complete this week.

My family wanted authorship, not suggestions.

The squares that motivated people were the specific, hand-written ones tied to a real person's real hobby. Anything vague — the kind of generic square an AI would generate from a list of hobby tags — created no pull. Being able to edit and write squares themselves wasn't a convenience feature; it turned out to be the product. A square works because someone you love wrote it.

The change: I rebalanced how boards get built.

Based on watching local squares drive all the early momentum while travel squares stalled, I restructured the board template — the default board now weights toward nearby, low-planning squares that can be completed on an ordinary weekend, with just a few big trips kept as long-term "anchor" squares. And based on the vagueness finding, I committed to hand-written squares as a design principle: Kept will never auto-generate board content, even though AI generation would be the easy vibecoded default.

Success Metrics

Squares completed per month — is the board driving real activity?
Gatherings initiated by the board — did a hangout happen because of Kept, rather than despite everyone's schedules?
Squares contributed by family members other than me — is this a product, or my project that my family tolerates?

LONG-TERM STRATEGY

Beyond one family: Can Kept grow without losing what makes it work?

Kept works for my family because the board is deeply personal. Scaling introduces trade-offs worth designing for deliberately:

Expanding to friend groups without diluting intimacy. Families have built-in commitment; friend groups don't. The mechanic may need lighter stakes — seasonal boards, smaller grids — to survive lower-frequency gatherings.
Keeping boards personal at scale. Hand-written squares are the magic. If growth means templated boards, Kept becomes a generic checklist. Any onboarding for new families has to protect the "everyone contributes their real hobbies" step.

CONCLUSION

When I started Kept, I assumed I was solving a coordination problem — my family just needed a better tool for planning time together. Building and shipping it taught me the problem was motivational, not logistical: we had plenty of ways to schedule time, and no reason to look forward to it.

This project reinforced a lesson for me as an aspiring product manager: shipping to five real users teaches you more than speculating about five thousand. AI tools collapsed the time from idea to prototype — but they made the product decisions more important, not less, because building became cheap and knowing what to build stayed hard. Kept exists because I watched a real problem in my own life closely enough to design for it — and the most honest success metric I have is whether my family asks to play again.