I'm a developer. I built an AI-powered manifestation app. If that sounds like an oxymoron, you're not alone — I thought so too at first.
But here's the thing: 28,547 people signed up. The longest streak is 93 days. And the most surprising part isn't the numbers — it's what they taught me about why people quit things, what actually changes behavior, and why the hardest part of building solo isn't the code.
The Accidental Niche
I didn't set out to build a "manifestation" app. I was building a daily ritual engine — a system that generates personalized, AI-crafted morning practices. Think affirmations, visualization prompts, small action steps, and daily signs to look for.
The idea came from my own struggle. I'd read every productivity book, tried every habit tracker, and built enough side projects to fill a GitHub graveyard. Nothing stuck. Not because the systems were bad, but because they were generic. A habit tracker doesn't care that I hate mornings. A meditation app doesn't know I process things through writing.
So I built one that adapts. Using DeepSeek V4 Pro, each user gets a ritual that evolves with them — shifting based on their goals, their streak, and which phase of the journey they're in.
I called it Wishyze. Mostly because I liked how it sounded.
The Phase Model (and Why Most People Quit)
Early on, I noticed a pattern in the data that changed everything.
Users follow a predictable arc:
Spark (days 1-7): Everyone's excited. Downloads, sign-ups, first rituals completed. Conversion from free to premium is decent. Morale is high.
Void (weeks 2-6): This is where 730f people disappear. The novelty fades. The ritual feels like a chore. The streak breaks once, and then it's over.
Alignment (weeks 6-12): The people who make it through the Void start showing real behavioral shifts. They're not just doing the ritual — they're living it.
Manifestation: Things start happening. Not because of magic, but because consistent small actions compound.
I called these the four phases, and I built the entire product around them. The ritual you get on day 3 is deliberately different from day 21. The prompts get harder, more specific, more challenging. Because that's what the data said people needed.
What Nobody Tells You About Solo Building
The code was the easy part. Next.js 14, TypeScript, Tailwind, Supabase, Vercel — I know this stack. I can ship fast.
The hard part was the loneliness.
Not the "I work from home and don't see people" kind of lonely. The "I made a decision at 2 AM about payment processing and there's literally no one to sanity-check it" kind.
Choosing Paddle over Stripe took me three weeks. Not because the integration was hard, but because I had no one to bounce the decision off of. Every choice felt permanent. Every bug felt personal. Every day with zero signups felt like proof that the idea was bad.
Here's what helped:
Ship something small every day. Even a one-line fix. Momentum beats motivation.
Talk to users, not founders. Founder Twitter will make you feel like everyone's doing $100K MRR. Your actual users will tell you what matters.
Set metrics you can control. I stopped tracking revenue and started tracking "rituals completed today." That number told me more about product health than MRR ever did.
The Thing That Surprised Me Most
I expected the technical challenges to be the hardest part. Getting AI to generate genuinely useful affirmations instead of cringey platitudes. Making streaks work across timezones. Handling the $9.90/month premium tier without losing users who can't afford it.
All of that was solvable.
What surprised me was the emotional weight of building something that people rely on. Not "rely on" like a to-do list. Rely on like a morning anchor. I started getting emails — not support tickets, actual emails — from people telling me that their daily ritual was the only consistent thing in their week.
That's when the project stopped being a side hustle and started being a responsibility.
Lessons for Other Solo Builders
If you're building something alone, here's what I wish someone had told me:
Your niche is probably weirder than you think, and that's fine. I spent months worrying that "AI manifestation" was too niche. Turns out, the people who need it really need it. Don't dilute your product to appeal to everyone.
The free tier isn't a loss leader — it's your best product. 3 free daily rituals. That's it. But those 3 rituals are genuinely useful. The conversion to premium happens when people realize they want more, not because you locked basic features behind a paywall.
Retention is the only metric that matters early on. I'd rather have 1,000 users who come back every day than 100,000 who sign up and forget. The Phase Model exists because I obsess over the Void — that 2-6 week window where people decide whether your product is part of their life or not.
Solo doesn't mean alone. I lean heavily on AI (obviously), but also on communities like Indie Hackers, dev Twitter, and the occasional 3 AM conversation with ChatGPT about whether Paddle or Stripe is better for international payments. (Paddle, by the way. If you sell to a global audience, Paddle handles all the tax compliance for you.)
What's Next
We're past the 28K mark now. The streak record keeps climbing. I'm building new features — more personalization, better ritual generation, deeper analytics for users who want to understand their own patterns.
But honestly? The thing I'm most proud of isn't a feature. It's that people come back. Day after day, week after week. They do their ritual, they look for their sign, they take their small action. And slowly, things change.
Building solo is hard. Building something people actually need is harder. But if you get the second part right, the first part becomes worth it.
If you're curious what a 28K-user AI ritual engine looks like, you can try it at wishyze.com. And if you're building something solo and feeling the weight of it — you're not alone in being alone. Ship small, talk to users, and don't let the Void win.
Top comments (0)