From BeeDone to Muse Otter: What I Learned Building 3 Structured Reflection Apps
From BeeDone to Muse Otter: What I Learned Building 3 Structured Reflection Apps
Before Muse Otter, there was BeeDone. And Evolum. And Strive.
7 years of building AI companioning products. Some worked. Some didn’t. Here’s what I learned.
The Apps
BeeDone (2019-2021)
The idea: 8 AI companiones with distinct personalities for productivity and habit tracking.
What worked:
- Multiple coaches = 91% engagement rate
- Gamification (streaks, XP) kept users coming back
- Personality-driven AI felt more human
What didn’t:
- Too many features, too little focus
- Monetization came too late
- I burned out maintaining it
Key lesson: Ship with payments from day one.
Evolum (2022-2024)
The idea: AI-personalized daily rituals for personal development.
What worked:
- Focused on one thing: morning routines
- RevenueCat from the start
- Premium positioning ($9.99/month)
Results:
- 30k€ MRR at peak
- 28% free-to-paid conversion
- 4.9★ App Store rating
What didn’t:
- Too niche — hard to grow beyond early adopters
- Rituals are hard to stick to (who knew?)
- Apple policy changes hurt growth
Key lesson: Build for retention, not just acquisition.
Strive (2024-2025)
The idea: AI goal suggestions with step-by-step coaching.
What worked:
- Goal decomposition was genuinely useful
- Users loved the structured methodology
- 60-day build time (getting faster)
What didn’t:
- Goals are too abstract for daily engagement
- Users set goals, then disappeared
- The “coach” felt like a task manager
Key lesson: Daily value > occasional value.
What I Brought to Muse Otter
Every app taught me something. Muse Otter is the synthesis:
| Lesson from | Applied to Muse Otter |
|---|---|
| BeeDone’s multiple coaches | 6 specialist coaches + custom creation |
| Evolum’s monetization | RevenueCat from commit #1 |
| Strive’s methodology | Agent OS structured coaching framework |
| All three: context matters | Know You personal context system |
The Patterns That Actually Work
1. Multiple Personalities > One Generic Assistant
BeeDone proved this: users engage more when coaches have distinct personalities.
In Muse Otter:
- Clarity — Asks questions to help you think
- Focus — Keeps you on track
- Spark — Creative brainstorming partner
- Anchor — Grounds you in values
- Horizon — Long-term vision planning
- Momentum — Action and accountability
Same Gemini model. Different prompts. Completely different experience.
2. Context is the Product
Generic AI gives generic answers. The magic is in the context injection.
Muse Otter’s Know You system builds a profile:
- Current project
- Core values
- Working style
- Recent challenges
Every conversation builds on this. The AI gets better at coaching you specifically.
3. Daily Touchpoints > Big Sessions
Strive failed because goals are weekly/monthly. Users set them and forgot.
Foundation in Muse Otter creates daily value:
- Quick reflections
- Streak gamification
- Context that compounds
4. RevenueCat from Day One
BeeDone taught me this the hard way. Evolum proved it works.
Don’t build for 6 months then “figure out monetization.” Integrate payments first. The free tier is a feature, not a crutch.
The Numbers That Matter
From 7 years of structured reflection apps:
| Metric | What I’ve Seen |
|---|---|
| Free → Paid | 5-8% is good, 20%+ is exceptional |
| Retention D7 | 40%+ means you have something |
| Monthly churn | Under 5% = sustainable |
| LTV | 8-10x acquisition cost minimum |
Muse Otter targets:
- 5%+ conversion
- 70% yearly vs monthly (higher LTV)
- Daily engagement via Foundation
Why This One Feels Different
I’ve shipped apps that made money. I’ve shipped apps that failed.
Muse Otter is different because:
- It’s focused. Not 50 features. Six coaches + personal context.
- It has daily value. Foundation creates reasons to return.
- It builds on everything. Every lesson from the last 7 years is in here.
- It ships fast. 3 weeks from zero to production.
The Real Insight
After 3 apps, here’s what I know:
AI companioning isn’t about the model. It’s about the relationship.
ChatGPT is smart. But it doesn’t know you. It doesn’t remember your last conversation. It doesn’t understand your values.
Muse Otter does.
That’s the difference between a chatbot and a coach.
What’s Next
Muse Otter ships February 12th for the RevenueCat Shipyard contest.
If it works, I’ll keep building. If it doesn’t, I’ll learn something new.
That’s the game.
Try it free — 20 messages/day, no credit card required.
