FitCraft · Product Proposal · Plain-English Edition
Our players collect cards that do nothing. The fix isn't 3D assets — it's rules. This is the product case for four features, in the order we'd ship them, each explained the way a player would experience it, with the apps and games that already proved the mechanic. Technical notes are tucked at the bottom of each section.
A FitCraft player's loop right now: work out → earn a pack → open cards → …nothing. The collection tab is a gallery. Duplicates can be burned for Sweat, and Sweat can craft more cards — a loop that goes in a circle and arrives nowhere. There are no daily quests, no seasons, no events, no reason to open the app on a rest day. The only retention mechanic is the workout itself.
The original plan — cards become real 3D gym props and trainer clothing — needs an art team we can't afford. But here's the thing the research made obvious: the most profitable collection games don't give their collectibles any "real" function either. Monopoly GO's stickers are literally pictures that do nothing, and its sticker album helped drive one of the biggest mobile launches in history. What those stickers have that our cards lack isn't utility — it's rules around them: sets, prizes, deadlines, near-misses, trading. Meaning is a game-design problem, not an asset problem. That's a problem a two-person team can afford to solve.
Before designing around assumptions, we pulled 90 days of production workout history from BigQuery (fitcraft_warehouse.workout_history, counted workouts only). Three findings reshape the plan.
Every decision in this proposal traces back to one of these. Each was learned — sometimes expensively — by someone else.
Finch (self-care pet app, ~$30M/year, no investors) has zero punishment: skip a week and your pet is simply happy you're back. It's beloved by exactly the users who quit other apps. Habitica is the cautionary tale: miss a day and your character — and your friends' characters — take damage. Users describe quitting the app entirely the first time life got busy. Duolingo ran 600+ streak experiments and found the counterintuitive truth: more forgiveness produced more daily users, not fewer.
This matters double for fitness: rest days aren't a lapse, they're part of the program. A game that frowns on Thursday's rest is contradicting our own trainers.
Our rule: nothing ever decays or is lost. Rest days quietly bank a bonus for your next workout.
Pokémon GO prices its rarest rewards in kilometers walked — you can't buy your way past 20 km per candy, and the paid item merely halves the distance. WalkScape (a 4-person indie hit) goes further: real steps are the game's only input, and better gear just converts steps more efficiently. Players trust these economies because effort can't be faked or skipped.
Our rule: game resources are minted only by finished workouts. Subscriptions and gear make effort go further — they never replace it.
Marvel Snap lets you "upgrade" a card five times — every upgrade reuses the same illustration with a fancier frame and shine, and players grind happily for it. Genshin Impact's famous artifact sets are literally text: "equip 4 pieces → +20% attack." FIFA Ultimate Team cards are just stat blocks. None of these needed new art to make their collections the core of the game.
Our rule: our existing 388 cards become set pieces, powers, and attack moves through configuration — the art we already shipped does the work.
The math of Panini sticker albums (the World Cup ones) shows collectors spend more completing the last 20% than the first 80% — and enjoy it. Monopoly GO weaponizes this: the UI constantly surfaces the set where you're one sticker short. An almost-complete set is the most powerful "open the app" trigger in collection design.
Our rule: the app should always show you the thing you're one workout away from.
A season that ends in two weeks (Fortnite, Monopoly GO) pulls you in — you come back because something good is expiring. A streak about to break pushes you with anxiety — and rule 1 says anxiety churns fitness users. Same urgency, opposite emotional cost.
Our rule: build urgency from expiring opportunities, never from threatened losses.
Four features, in shipping order. Each one stands alone, and each makes the next stronger.
Limited-time windows where the rewards players already earn get temporarily better — timed to what our data says players actually do, not to a "weekend warrior" assumption.
Ben did Monday's workout — like 74% of our players in a given week, it's his only one so far. Wednesday evening:
Nothing else changes. Same app, same workouts — but now there's a named, dated reason for workout #2. Which happens to be exactly the workout that keeps his streak alive.
The original pitch was "bonus weekends." The data disagrees: Monday is our biggest workout day and the weekend dip is mild, so a weekend-anchored event mostly rewards a minority behavior. The flagship event is the Second Wind window — converting one-workout weeks into two-workout weeks, because that's where 74% of our active population sits, and because two workouts is already the app's own streak-keeping bar. Weekend events remain in the toolkit as a growth play (filling our quietest days), and rarity events ("Legendary Weekend: 3× Legendary odds in every pack") are the reactivation flavor we push to the ~900 lapsed exercisers.
It's the highest return-per-week item in the whole evaluation: no new screens, no new art, no new player education — just a better reward for a dated goal and a push notification. It also starts teaching our audience that FitCraft has "moments" worth showing up for, which every later feature builds on. And it's the cheapest possible experiment factory: Second Wind vs. Legendary Weekend vs. Double Sweat Wednesday are all the same config knob wearing different copy.
Pack odds and reward bands live in server-side config (drop tables / reward schedule); the weekly workout count already exists via the calendar system, so "second counted workout this week" is a condition the server can already evaluate. An event = a scheduled config swap plus a JourneyEngine push campaign with deep links. Zero client code, zero art. ~1–2 person-weeks including scheduling/analytics.
Correction from v1: FitCraft's streak is already compassionately designed — two workouts in any week keep it alive. The problem isn't the mechanic. It's that the mechanic is invisible, and most players live one workout below its safety bar without knowing it.
Maya's home screen, Tuesday (one workout done this week):
And after that second workout lands:
The generous rule we already built finally gets credit for every week it saves — and mid-week, "one more workout" is a concrete goal instead of vague guilt.
74% of active weeks contain exactly one workout — one short of the streak-keeping bar. So for most of our players, most weeks, the streak is silently at risk, and the app never says so until it's gone. Meanwhile the weeks where the rule does save someone pass without acknowledgment. Both moments are free retention beats we're currently discarding. Note the compounding with Step 0a: "one more workout by Sunday" is simultaneously the streak bar and the Second Wind event goal — one behavior, two rewards, one push.
The streak calculation is already server-side; what's missing is a weekly streak state (safe / needs-one-more / at-risk) exposed to the client and a home-screen treatment plus one push template. ~1–2 person-weeks, mostly UI and copy. Later, Power Sets can grant Streak Insurance — one light week forgiven per month — as an earned upgrade, connecting the card game to the habit loop.
An 8-week themed "season" with two faces: a progress track (the pass) and a themed collection challenge (the album). This becomes FitCraft's reward calendar — the heartbeat every later feature syncs to.
It's week 3 of "Summer Shred." Maya finishes leg day and watches her season track tick from tier 11 to tier 12 — a Silver pack drops. The album tab shows this season's themed sets built from the cards she's been collecting all along: Gym Classics — 5 of 6. One short.
The season ends in 5 weeks. Free members can reach tier 25; her subscription unlocks the full track to tier 35, ending in a guaranteed-Legendary chest. She's never once asked "why am I collecting these?" since the season started.
The pass is the mechanic Fortnite made universal: do the thing you already do (for us: work out) → fill a visible tier track → claim rewards along the way. Free track for everyone; premium track for subscribers. It's the most-copied subscription converter of the past five years because it turns engagement itself into the purchase reason — by tier 20 you can see exactly which rewards premium would have paid you, and the season deadline gives the decision a date.
The album is Monopoly GO's playbook: group the collection into themed sets, pay an instant prize per completed set, a grand prize for the full album, and re-theme everything each season. Our twist — and our unfair advantage — is that players' existing cards count. The album is a fresh lens over the collection they already built, so a lapsed collector reopens the app to find themselves 60% done and three sets one-card-short. That's rule 4 doing the work from day one.
FitCraft sells a subscription; there are no consumable purchases, and this plan doesn't add any. That's an advantage, not a constraint: no dice bundles, no gem store, no pay-to-win accusations — the premium track is simply part of the subscription. What the season changes is the shape of the subscribe decision: instead of a one-time paywall moment, every eight weeks there's a dated, visible reason to subscribe (or re-subscribe, or not cancel): "the Summer Shred premium track ends in 3 weeks and I'm at tier 19." Zombies, Run! and Duolingo both monetize this way — content and forgiveness through the sub, never through consumables. Our existing countdown-timer paywall finally gets a real deadline to count down to.
Of all 18 concepts evaluated, this scored highest on monetization and tied for most-buildable — and every judge picked it. But the strategic reason it goes first: it creates the calendar. Event windows get scheduled inside seasons. Power Sets rotate with seasons. Boss raids theme to seasons. Once FitCraft has a season, live content stops being ad hoc.
Season points are the workout reward points the server already grants — which are already effort-scored and capped per day, so the pass inherits anti-cheat for free. The album is a read-only view over the existing collection data. Art per season: ~10–15 flat images (banner, set badges, track frame) — a morning's work in the fcadmin image pipeline. Seasons are config documents, so season 2+ is authoring, not engineering. MVP 4–6 person-weeks (pass first, album in the following release).
Completing a themed card set grants a permanent perk. This is the moment the sticker book becomes equipment.
Maya finally pulls the last card of Iron Foundations (five strength-training cards she mostly already owned). A celebration plays, and something new happens:
She scrolls the set list: Lucky Lifter (5/6 owned) improves rare-card odds in every pack. Recovery Ritual (4/5) grants Streak Insurance — one light week a month can't break her streak. Season Sprinter boosts season-track points. She has duplicates to burn and just enough Sweat to craft the missing card — the first time crafting has ever felt like a decision.
Fair pushback from v1: "+10% XP" is weak because XP is currently just a number — levels don't gate anything players feel. Powers only work if they touch things players demonstrably care about. In FitCraft today that list is short and precise: which pack drops after a workout (reward points → pack tier), what's inside it (rarity odds), Sweat, the streak, and — once Step 1 ships — season progress. So the power catalog draws from exactly those: weekly-bonus-pack upgrades, pack-luck, Sweat multipliers, Streak Insurance, season-point boosts, an extra season wildcard. If we ever want XP to matter, levels can later gate Titan tiers or cosmetic gym backdrops — optional, not load-bearing.
Three compounding effects. First, it's retroactive: the day this ships, every player's existing collection partially completes dozens of sets — instant "almost there" hooks everywhere (rule 4). Second, every future pack open has stakes, which makes the Seasons pass rewards more exciting, which makes workouts pay more. Third, the powers themselves point back at working out: better packs, faster season progress, protected streaks. The collection stops being a side room and becomes the engine room.
Sets are server configuration over card metadata we already have (body sections, category, rarity, trainer variants) — no catalog migration. Powers hook existing server calculations: the reward-points grant (pack tier), the pack-odds roll, the streak check, and the season-points hook from Step 1. Art: ~30 flat set-badge/power icons (~$50 in generation). The one real risk is economy tuning — powers touch live reward rates, so they launch with conservative caps and analytics. MVP: 5 hand-designed sets, 3 power types.
A weekly boss monster that the player's real workouts damage — with their card collection as the moveset. This is the feature that earns the word "game."
Ben's Tuesday workout charges 340 Power. Afterward, a 30-second beat: he picks three cards from his collection as attacks. His Legendary squat card is leg-type — and he actually did squats today, so it hits for double. The Titan's health bar drops to 61%.
Sunday night it dies. A chest arrives. Next Monday: a new Titan, weak to core. Ben, who has skipped core day for a month, looks at his ab-exercise cards differently.
Fair concern: FitCraft is premium and clean — realistic 3D trainers, metal-finish cards, dark cinematic pack openings. A goofy monster would wreck that. The direction is "premium monolith," not creature: each Titan is a monumental sculptural form — the training obstacle it's named for, rendered like a high-end product shot. The Couch is a brutalist throne in black marble; The Plateau a massive hovering slab; The Snooze a cracked obsidian bell. Think perfume-ad lighting on a dark backdrop, not Saturday-morning cartoon.
The interaction language comes from what the app already renders beautifully: weak points are glowing fault lines in the stone (color-coded to body sections); damage is fractures spreading across the surface; the kill is a slow-motion shatter into particles using the same shader vocabulary as the card foils. The attack moment is your actual 3D card prefab — the premium object players already own — slamming into the monolith with camera shake. Visually, the raid screen is the pack-opening screen's dramatic sibling, not a new art world. And because fcadmin's Scenario pipeline supports custom-trained style models, we train one on this exact render look so every monthly Titan matches. If even sculptural "Titans" feel too fantastical for the brand, the identical mechanics reskin as Weekly Trials — a monument the gym demolishes together — with zero design change.
Everything before this step makes collecting meaningful. This step makes working out meaningful inside the game — the workout finally has an in-fiction consequence beyond points. It scored highest overall of all 18 concepts, with the judges' "makes FitCraft feel like an actual game" differentiation. It also finally employs the 272 exercise cards (the album/sets lean mostly on item cards), and the typed weaknesses gently nudge balanced training: a leg-weak Titan is the best leg-day advertisement we will ever run.
Power is minted server-side from the same scored, daily-capped workout results as everything else. Attack presentation reuses the existing 3D card prefabs animating into the boss — no character animation needed. MVP art: 3 Titan renders (intact / fractured / shattered states) in the locked photoreal style via the fcadmin Scenario LoRA; a year of monthly Titans is ~12 renders, well under $100 of generation. Personal-HP bosses first (no player-population risk); community bosses, damage leaderboards, and seasonal Titan tiers come later.
The original dream — your cards become a gym you actually build — rescued at 2D prices.
Picture the collection as a side-view illustrated gym: every equipment card you own becomes a placeable machine, decoration cards raise its style rating, environment cards re-theme the walls, and tiny animated members wander in and train on gear you earned by sweating. This is the most marketable screenshot FitCraft could ever produce, and it scored highest of all 18 concepts on differentiation.
So why wait? Because even at ~150 sprites it's a second product: simulated members, placement UX, an idle economy to balance. Built today it would eat two quarters and sit on top of an economy that doesn't exist yet. Built in year 2, it sits on Seasons, Sets, and Titans — and every card a player collected along the way is already waiting to be placed.
Eighteen concepts were designed and scored by three independent judges. The full scorecards live in the technical edition. The notable passes, in product terms:
| Idea | The pitch | Why we passed (for now) |
|---|---|---|
| Companion mascot | A Finch-style creature fed by your workouts, dressed in your avatar cards | Proven psychology, but the entire bet rides on one AI-generated character being genuinely lovable — and Finch's real moat is thousands of hand-written reactions. Revisit after our art pipeline proves it can produce a character with charm. |
| Idle expeditions | Send squads of cards on hours-long adventures that resolve while you rest | Fun loop, but the timers make the game about waiting rather than training — it risks decoupling the game from the gym. The Titan's banked Power keeps the same "rest is productive" feeling without a second economy. |
| PvP dojo ladder | Your training builds a fighter who auto-battles other players weekly | Great fantasy, wrong stage: ladders need a large player population or they're arenas full of ghosts. Shelved until DAU supports it. |
| Renovation story | Each workout funds the next before/after transformation of a derelict gym | Charming and cheap per chapter (Merge Mansion's meta), but retention dies the week the hand-authored content runs out — a treadmill a 2-person team shouldn't sign up for. Its best mechanic (spending duplicate cards on projects) gets folded into Power Sets. |
| In-workout coach widget | A chibi cheerleader reacting to your sets during the workout | It would compete with our own 3D trainer at the product's most important moment, and mid-set attention is the one thing we shouldn't gamify. |
The reason this plan is affordable at all — and we're further along than we thought.
FCAdmin is already an AI art factory. The Image Editing pipeline built for program character images does, today: prompt generation (GPT), image generation through Scenario with a custom-trained style model (which solves the "AI art looks inconsistent" problem), editing passes via Gemini/Nano Banana/Flux, automatic background removal to transparent PNGs, and publishing to the CDN that delivers content to the app without an app-store release. This proposal doesn't require building an asset pipeline — it requires pointing the existing one at a second asset type.
The one-sentence version: we stop asking "how do we afford to make the cards real?" and start asking "what rules make the cards matter?" — because that's the question Monopoly GO, Marvel Snap, and Duolingo already answered, on budgets per-mechanic smaller than ours.