For the past 3 years, I’ve been running 1811 Labs - a micro products studio.
We’ve built 15+ products, and exited 6 of them.
Unlike most stories you hear on X, I’m not a solo builder. I have a small but incredible team, and we’ve learned this lesson the hard way - by building, shipping, selling, and sometimes getting lucky.
This is not a victory lap. This is an honest post-mortem.
Because the Micro SaaS playbook that worked a few years ago is quietly dying.
I’m an engineer by training, but I had never coded properly before the vibe-coding era.
My early career was in Business & Strategy - eCommerce roles, followed by a short stint at a venture capital fund. In 2021, during a sabbatical, I was a Founder in Residence at EF, binge-watching YC videos and trying to make sense of startups.
That’s when I stumbled upon Micro SaaS.
And honestly? I was in awe.
Small problems.
Niche users.
Real money.
No VC pressure.
Financial independence.
You could build:
Big companies wouldn’t care - the markets were too small.
VCs wouldn’t care - not venture scalable.
But users would And some of them would pay.
Then I read about PushOwl doing millions in revenue with a tiny team.
Then I discovered levelsio. That was it. I was hooked.
At that point, I couldn’t really build.
So I did what everyone did:
No-code. Low-code.
Bubble. Airtable. Softr. Everything.
I went back to a job, kept building “stupid” products with friends, made tiny money - but one thing became clear:
The playbook was simple.
Pick a small problem → Build an MVP → Post on X / Reddit → Launch on Product Hunt → Do some SEO → Pray.
And sometimes… it worked.
Most products barely made ramen profitability.
But the idea was intoxicating.
You just had to BUILD.
Mid-2022 changed everything.
I read a generative AI note by Sequoia Capital.
Then I played with Stable Diffusion.
My brain melted.
Around the same time, Pieter Levels went viral again with avatar AI.
I thought - This is it. My time has come.
I was working full-time in VC but started building aggressively on the side. The idea was obvious - build something like Avatar AI, but for India.
We built Avatarize - a hacky AI avatar generator. Somehow got it working. (The project is now deadpooled)
At the same time, on a random Friday night, a friend and I built a pickup lines generator in a couple of hours. We launched it on Product Hunt and posted on Twitter.
The stupid product went viral.
100K+ visitors in no time.
I randomly listed it on Acquire.
We made a tiny 5-figure exit.
That was the final nail in the coffin.
I quit my job. Again.
I didn’t have a concrete idea.
VC had also given me a bias against non-venture-scale ideas.
So I defaulted to the old playbook:
Over 3 years:
And here’s the uncomfortable truth:
The playbook stopped working the way it used to.
Yes, we still:
But the cracks were obvious.
Over the last 6–12 months, everything accelerated.
With vibe coding tools, Cursor, Claude, and agents — building is trivial.
Anyone smart, with clarity and context, can build:
The supply of “Micro SaaS ideas” has exploded.
Old channels are dying:
Right now, TikTok UGC is one of the few channels still working at scale.
Users will pay. But they will not tolerate garbage MVPs anymore.
They want:
“MVP” as an excuse is dead.
If your plan is:
The window is extremely limited.
Most niche users are now vibe-coding their own solutions.
Many “wins” you see on X are:
And the number of builders has exploded.
We saw this early.
We exited most of our portfolio.
We currently have 3 products on sale.
And we’re focusing on just two.
We moved from:
Web SaaS → Consumer / Cross-platform AI apps
Our hero product Audionotes is going through the most intense build cycle of its life.
Our lead engineer, backed by a small army of Claude code agents, and our entire team is now obsessed with one thing -
Building the best voice-first note-taking app in the world.
Our 100K+ users give us data, feedback, and iteration speed - a temporary moat.
But even that isn’t enough.
We cannot be another app.
We have to be a complete product.
The new play looks like this:
The days of:
“I’ll ship a quick Micro SaaS and see what happens”
are mostly over.
There’s a paradox here that’s worth calling out.
While Micro SaaS as a business model is getting harder…
building itself has never been more powerful.
At 1811 Labs today, we are:
Not because we suddenly became 10x engineers overnight -
but because the tooling changed the ceiling of ambition.
We’re now doing proper engineering:
Earlier, our constraint was:
“Can we even build this?”
Now the constraint is:
“Is this good enough to win?”
That’s a massive shift.
AI copilots, agents, and modern infra haven’t just made building faster =
they’ve made taste and engineering judgment the bottleneck.
And that’s exactly why half-baked Micro SaaS struggles.
Micro SaaS isn’t dead because demand disappeared.
It’s dying because the barriers collapsed.
When everyone can build,
taste, distribution, and retention become the game.
Build - but only if you can acquire users profitably and serve them better than anyone else. But rememeber - you still may not get micro-acquired.
The playbook is changing.
Cursor. Claude. SOTAs are here.
And they’re not waiting for anyone.