Game 01 — Kimosabe Beta Team Competition
Codename: First Hundred. Duration: 90 days from Cohort 001 invite-send. Stakes: Founders' badge, leaderboard glory, and a vote on Phase 2 priority. No money. Money corrupts beta signal.
This is the playbook for turning the 90-day Phase 1 freeze into a game that beta testers actually want to play — and that produces better data than any survey ever could.
Why a game?
Beta testers ghost. Surveys lie. Bug reports are sparse and shallow. But people who feel like they're on a team, with a scoreboard, will:
- Use the product more often (we need real usage data).
- Report more bugs (we need real bugs, not theoretical ones).
- Tell their friends in a way that's authentic (we need a tiny, honest growth signal).
The trick is to gamify the right behaviors — the ones that produce signal — without gamifying vanity metrics that produce noise.
How you score points
| Action | Points | Cap | Why |
|---|---|---|---|
| Daily active session (≥1 message sent or read) | 1 | 1/day | Real usage > installs |
| Send a real message to another Cohort 001 member | 2 | 10/wk | Network effect inside the cohort |
| File a bug with repro steps | 10 | — | This is the gold |
| File a bug we ship a fix for | +25 | — | Reward signal, not noise |
| Ship a UX paper-cut nit (text/contrast/layout) | 5 | — | A11y + polish |
| Win a "Feature Wish of the Week" vote | 25 | — | Steers Phase 2 priority |
| Refer a person who completes onboarding | 15 | 5 total | Capped — quality > quantity |
Help another beta user in #help (founder-verified) |
5 | 3/wk | Community is force-multiplier |
Anti-patterns (penalties):
- Spam-sending to inflate message count: messages reset to 0 for that day and a polite warning. Repeat = removed from the cohort.
- Vague bug reports ("it's broken") get no points. We reply with the template and re-grant points if updated.
The leaderboard
A simple ranked list, refreshed nightly, visible at /cohort (auth-walled
to Cohort 001 members). Three views:
- All-time — total points since the cohort opened.
- This week — fresh start every Monday; encourages re-engagement.
- Bug hunters — bug-fix points only; the highest-signal subset.
Display: rank, @handle, points, top-3 badges. No profile pictures, no
streaks pressure, no notifications about other people's scores. Privacy
software cannot copy the dark patterns of attention software.
Tiers & badges
Lifetime tiers (persist past beta, displayed as a small badge in the user's profile forever):
| Tier | Threshold | Badge name | Visual |
|---|---|---|---|
| Bronze | 50 pts | Founder | gold dot |
| Silver | 200 pts | Cartographer | gold ring |
| Gold | 500 pts | Architect | filled gold octagon |
| Mythic | Top 5 overall at end of beta | Kimosabe | Metatron-cube glyph |
Names matter. Founder is intentionally generous — anyone who shows up gets it. Kimosabe is intentionally rare — five humans, ever.
Weekly rhythm
Mon 09:00 ET → Weekly leaderboard reset
Mon 10:00 ET → "Last week's wins" digest in #announcements
Wed 18:00 ET → Founder office hours (voice in Discord)
Fri 17:00 ET → "Feature Wish of the Week" vote opens (24h)
Sun 17:00 ET → Vote closes, top wish goes into ROADMAP backlog
Predictability is the feature. People show up to things they expect.
Special events (use sparingly)
- Bug-Bash Saturday (once a month, 2h window): 2× points on every
bug. Founders are live in
#help. Catered for the founders, not the testers — keeps it human. - Stress-Test Hour: announced 24h ahead. Everyone tries to break the same flow at the same time. Reveals real concurrency / realtime issues no synthetic test would catch.
- Final Week Sprint: last 7 days of the 90-day program, all categories worth 1.5×. Decides the Mythic tier.
What the founders commit to
If we're asking testers to play, we owe them:
- A reply to every bug within 48h. (Same SLA as
SUPPORT.md.) - A weekly digest that names the top contributors and the bugs they
found, by
@handle(with consent). - A vote that actually steers the roadmap. The Feature Wish of the
Week winner gets added to the
ROADMAP.mdbacklog within 7 days, with a written reason if we deprioritize it. - No dark patterns. No streak guilt. No "you're falling behind" notifications. No leaderboard push notifications. Ever.
Judging criteria for the founder vs. founder challenge
When we both ship a version of (landing + onboarding + manual + support + this game) and one of us judges, the rubric is:
| Criterion | Weight | What "good" looks like |
|---|---|---|
| Honesty | 25% | Says what the product is and isn't. No marketing fog. |
| Brand fidelity | 15% | Looks and feels like Kimosabe — purple/gold/cream, Inter, Metatron, Curate Your Brain. |
| Beta-tester experience | 25% | Could a real human actually use it? Did onboarding teach the one thing that matters (recovery key)? |
| Gameplay design | 20% | Does the game reward signal, not vanity? Are anti-patterns penalized? |
| Operational realism | 15% | Could 2 founders actually run this on day one without burning out? |
Tie-breaker: which version would the next hundred users want to be part of?
Why this matters more than it looks
A beta is a contract: testers give us their time and attention, we give them the feeling of building something with us. If we run it like a broadcast (here is the build, please file tickets), we get broadcast results (silence, then churn). If we run it like a game — with rules, rhythm, and stakes that match what we actually want to learn — we get a hundred co-founders for the price of a Discord server.
That's the game. Let's play.