Is the future of indie building a portfolio of small products, not one big startup?
Everyone talks about building the next unicorn. But what if that's the wrong goal entirely?
Here's a different mental model worth considering. Instead of spending 5 years chasing one product to $10M ARR, what if you built 10–20 smaller products (each generating $10k–$50k MRR) managed entirely by AI agents?
The math works out, the risk is distributed. And you don't need to be a monopolist to win.
Pieter Levels has been running this playbook for years, with multiple products each solving a specific problem for a specific audience. But what's changed is the cost of execution. AI agents can now handle customer support, content, marketing workflows, and increasingly even product iteration. The overhead of running multiple products is collapsing.
The case for the portfolio approach:
if one product dies, the others survive
each product can own a small niche completely
agents reduce the operational cost of running multiple products simultaneously
you compound learnings across products faster than doubling down on one
The honest challenge:
focus is a real advantage, spreading across 10 products early can kill all of them
distribution for each product still requires attention
quality tends to drop when you're running too many things at once
So I'm curious:
Do you think the multi-product portfolio model is a real strategy for indie founders in 2026-2027?
Or does focus always win in the end?
And has anyone here actually tried running multiple products simultaneously. What happened?
Replies
minimalist phone: reduce your screentime
In my opinion, it is a good strategy always, because your bets increase the chances of succeeding at least with more projects. More shots = More chances to win. The only important thing is to know when and to which projects say "goodbye".
I think you would have to focus on one thing for it to take off to even make $10k MRR. That alone is a huge challenge. To do that 10-20 times would take incredible effort over years. Unless you are a massive established name with a huge following then the distribution is going to take a long time.
My wife and I have literally been talking through exactly this. For us the draw isn’t swapping a startup for a portfolio - it’s that smaller products spread the risk and throw off extra income toward buying a house.
Because none of the bets are existential, the math feels a lot more forgiving than going all-in on one thing
On focus vs. spread, I’d say it’s less either/or and more about sequencing: build one product to a real footing first, then layer in the next, rather than splitting attention across ten on day one.
the portfolio model works but the sequencing matters more than the math. pieter levels didn't run 10 things in parallel from day one, he compounded one at a time until distribution and ops were repeatable. the part that's actually changed is agents collapse the maintenance cost once something is working, not the build and launch cost. you still need real attention to get each product to first traction. so the honest answer is probably: build one, get it to ramen, then let agents hold it while you start the next. spreading across 10 before any of them have traction is just well-organized failure.
focus wins until distribution is the bottleneck. then portfolio wins.
i watched a founder spend 3 years optimizing one SaaS from $2k mrr to $4k mrr. same founder spent 6 months later shipping 4 small tools. total hit $8k mrr. distribution compounded across the 4. each tool funneled into the next.
the multi-product model breaks when you treat each product like a separate startup. it works when you treat each as a chapter of the same story. shared brand, shared audience, shared distribution rails. that's the levels playbook.
red flag: if you're shipping 5 products and none share an audience you're not running a portfolio. you're running 5 startups poorly. that's how focus reclaims the throne.
I think AI makes the portfolio model much more realistic, but only if every product shares the same ecosystem.
For example, one product brings users, another solves the next problem for the same audience, and they cross-promote each other. That's how distribution compounds.
If every product targets a different market, you're rebuilding trust, marketing, and support from scratch each time. At that point, it's no longer a portfolio—it's multiple independent startups.
For me, the goal isn't 20 random products. It's a small suite of products that naturally grow together.
I like the portfolio idea, but I think timing matters.
If none of your products have found product-market fit yet, splitting your attention across 10 ideas can make it harder to learn what users actually want. I'd rather get one product working first, then use those lessons to build the next.
Once you've got repeatable systems and some revenue, a portfolio starts to make a lot more sense. The products can even help each other instead of competing for your time.
For me, it's less about choosing one approach forever and more about knowing when to switch.
I've built four products solo over the years, so I've lived a version of this. One distinction I'd add: the portfolio model and focus aren't actually opposites. What's worked for me isn't running 10 things at once, it's running them in sequence and letting the learnings compound. Each product taught me something the next one needed.
The thing people underestimate isn't the build cost, which AI really has collapsed, it's distribution. Agents can run support and content, but getting each product in front of the right people still takes human attention, and that doesn't parallelize 20x.
So my honest take: AI lowers the cost of building 20 products, not the cost of getting 20 products noticed. That second cost is the real ceiling, and it's still very human.