A lighthearted, social utility app that turns the clinical chore of health-habit tracking into a delightful, close-friends game — designed end-to-end and handed off to development in 3.5 weeks.
01 — Overview
ploop! turns the tedious chore of health-habit tracking into a delightful, shared experience — through playful "poop grading," custom status logs, and a private, close-friends social feed.
The idea came from a real behavior I'd noticed in my own life: my friend group already trades bathroom updates as running jokes in our group chat — latent social energy no existing app had captured. I leaned into a deliberately taboo, low-stakes domain on purpose — a fast, low-risk testbed for a bigger question: could I take a zero-to-one mobile product from concept to dev handoff in weeks instead of months, without sacrificing craft, accessibility, or user trust?
02 — The Problem
Most health-habit tracking is built like a medical chore — clinical forms, lonely charts, and no reason to come back. People log for a few days out of novelty, then quietly quit. The behavior that actually sticks in real life isn't solo and serious; it's social and a little silly.
Keep a daily habit going with close friends — fast, lighthearted, and a little competitive — without charts, streak-shaming, or sharing anything personal in public.
Maya keeps her five-person friend group glued together over text — memes, daily check-ins, inside jokes. She's tried streak apps and a couple of wellness trackers, but always drops them within a week. She doesn't want another health app; she wants something fun to do with her people.
"I'll do almost anything if my friends are doing it too — but it has to be fast, and it has to be funny."
03 — Strategy
Could we pivot a taboo health habit into a low-stakes social game? To move at speed, I ran a dual-AI pipeline — ChatGPT for market research, Claude for structural scoping — while acting as the core product manager and curator.
I prompted ChatGPT to benchmark existing poop-tracking apps and pinpoint the exact gap in the market.
Plop, Happy Poop, Poopify
Strong medical utility (Bristol scale, symptom charts), but sterile and joyless.
Poop Map
Strong social pull (1M+ downloads) via location tags and friend leagues — but thin once the novelty fades.
Number Two
Beautiful UI and lightning-fast logging — proof people crave frictionless logging — though the personality stays muted.
Mapping the named leaders against their categories made the open lane impossible to miss:
| Category | Existing Leaders | Status |
|---|---|---|
| Medical & IBS tracking | Plop, Happy Poop | Saturated |
| Location tracking | Poop Map | Saturated |
| Private / frictionless logging | Number Two | Saturated |
| Social habit game | — | Wide open ← |
The reframe that unlocked everything: almost every poop app asks "How is your digestive health?" — very few ask "How do we make pooping fun?" ChatGPT named the whitespace precisely: "Strava meets Duolingo meets Letterboxd, but for pooping." The bet wasn't a better medical tool — it was a social habit game that happens to be about pooping. The evidence already pointed there: users of existing apps consistently engage with friend leagues more than the health tracking itself.
A market gap is a hypothesis, not a green light. I ran 9 informal interviews with friend groups in their early 20s to pressure-test the social-game premise — and hear which features they'd genuinely use versus dismiss as gimmicks.
"Honestly? I'd download it for the group-chat chaos, not my gut health."
— Maya, 24
"If logging takes more than ten seconds, I'm out. I'm not journaling."
— Devin, 22
"No chance I'm posting that publicly. But to my close friends? That's comedy."
— Priya, 25
A few priorities surfaced again and again — and they became the backbone of the MVP. This is the moment the feature list stopped being my guesses and started being theirs:
| What I heard | How it shaped the MVP |
|---|---|
| "Speed, or I won't bother." | Quick Log became the hero — a center-screen button and an under-30-second flow. |
| "Only with close friends — never publicly." | A friends-only feed with opt-in sharing; the global public feed was cut to V2. |
| "The competition is the whole point." | The weekly "Throne Room" leaderboard earned a permanent spot as a core tab. |
| "I want to see my own streaks." | History & calendar prioritized for private, personal tracking — separate from the social feed. |
With the positioning validated, I translated the priorities into a shippable scope for a lean two-person team. Claude was a fast brainstorming partner for mechanics — but my job was to protect the timeline and stay honest to what users actually asked for, sorting features into launch phases.
Claude generated 15 feature vectors; my job was to curate them through UX feasibility and empathy. A few representative calls:
Highly invasive and causes immediate social friction — the kind of feature that gets an app deleted within a day. Cut entirely.
Too clinical — it turns the app into an exam. I pivoted it to a subjective, playful "Poop Grade" (A+, B−, "The Masterpiece") that users self-select or let a fun prompt decide. Kept the engagement hook, dropped the judgment.
Exactly right — it minimizes social anxiety and guarantees a psychologically safe space. This became the backbone of the entire feed model.
The biggest UX risk was the privacy barrier. To make users feel safe in a social app about bowel movements, I set two non-negotiable guardrails:
Within 48 hours I synthesized our AI conversations into a product spec anchored by a 5-tab architecture — Feed, Leaderboard ("Throne Room"), Log Entry, History, and Profile — with a prominent center action button for logging. Each is shown in full in the final prototype below.
04 — Visual Identity
Setting the right tone was a tightrope walk. Too sterile and it reads as a medical app; too loud and it looks childish. I aimed for a "sophisticated cute" aesthetic — structural typography (Fredoka One display, Nunito body) carries the playful energy so the palette and UI stay clean and modern. I built a rough system in Figma first, then handed it to Claude Design as a reference so its output matched my direction from the start.
The rough system I built in Figma to steer the AI — a Fredoka/Nunito type scale, a grounded dark-brown base, a crisp cream canvas, and saturated accents. From here I built out the full system in Claude Design, art-directed toward a flat, high-end look where type, color, and tone carry the personality, not decoration.
05 — Wireframing
With the visual system locked, I used Claude Design to explore layouts in parallel rather than sketching screen by screen — three distinct directions for each core screen, each weighed for scannability and speed of logging, then merged into a deliberate composite. Shown here: the hero "Log a ploop" flow.
Three directions for the Log flow — each annotated with pros, cons, and the rationale behind my pick.
06 — Deep Dive
While the core flows fell into place quickly, the History (Calendar View) tab posed a real challenge: a high-level monthly overview had to track three data layers at once without cluttering the screen.
The initial concept used colored dots where circle size was proportional to logging frequency. I quickly spotted the flaws: it was subjective, added visual noise, and forced users to guess the mathematical ratio between one log and three.
I challenged Claude to bypass the proportional-dot logic and explore a broader spectrum of data-visualization styles. It returned a matrix of alternatives:
Seven directions for encoding three data layers in one calendar cell — per-log pips, dot + numeral, tally bar, corner badge, hand-tally ticks, and the two tinted-cell finalists. Option F (tinted cell + number) won: the background tint turns the whole month into a readable color map, while the count pill stays quiet until it's actually needed.
07 — Hi-Fi & Accessibility
With the structure in place, I moved into high-fidelity prototyping — balancing design goals against engineering constraints through a continuous feedback loop with my developer partner.
AI could generate options, but only real people could tell me whether a social app about poop would actually feel safe and fun. So before hardening anything, I ran a scrappy round of guerrilla testing — a clickable prototype put in front of 7 people in the target demographic (early-20s, all active in group chats). It wasn't a formal study, but it changed real decisions:
The polished build, prototyped in Claude Design — every earlier decision made visible: the tinted-cell calendar, the accessible color tokens, the Bristol scale softened into plain-language labels, and a voice that hypes you up instead of charting your health.
08 — Status & Takeaways
ploop! is in active development. Because the design system and tokens were defined and validated up front, the handoff has been clean — my engineering partner is spinning up the iOS app with minimal layout friction.
Pre-launch, I defined the metrics we'll judge it on: activation, D7 retention, share rate (posted vs. kept private), and median time-to-log (under 30s). Share rate is the one I'm watching hardest — it's what tells us whether the privacy-and-delight balance actually works in the wild.