The Hyper Niche Is Wide Open

Fable 5.1 was released today. Every time one of these models drops, the same thing happens: everyone posts about it, argues about the benchmarks, and then goes back to their day. Almost nobody actually builds anything with it.

I think that's the mistake. Ever since Opus 4.5, anyone with a cutting-edge lab subscription — Anthropic, OpenAI, whichever — has had a genuine opportunity in front of them. Not a theoretical one. A real one, sitting there, largely unclaimed. And the best version of that opportunity for a normal person is not the next big platform. It's hyper niche SaaS.

Why small markets are suddenly worth chasing

A hyper niche SaaS might have a total market of under a million a year. That's the whole thing. Before, chasing a market that small made no sense. You'd need a developer, a designer, someone to answer support tickets, someone to do SEO, someone to run ads, someone to sell. By the time you paid all of them, the market wasn't big enough to cover the cost of serving it. So those problems just sat there, unsolved, because nobody could afford to solve them.

That math has changed. The website, the software development, the support, the sales, the SEO, the marketing, the advertising — all of it can now be done autonomously. Which means the small market is finally worth entering, because your cost of entering it has collapsed.

And it means you can do it while keeping your day job. You can do it with your kids in the room. You talk to your AI and let it build. It will do a better job than you, and it will be faster than you. I say that without any ego about it — it's just true, and once you accept it, you stop being the bottleneck in your own projects.

What happened this morning

Overnight I got a paying customer for narccount.ca. My ARR is now 3K.

I haven't spent a dime on marketing. I haven't started SEO. I built the website and forgot about it. I told Claude — 4.8 at the time — to just go, agreed with every suggestion it made, and the whole thing was done in maybe an hour.

And it's working.

The reason it's working isn't clever. It's hyper niche, it's easy to Google, and it solves a real problem for pharmacies. That's the entire formula. When the niche is narrow enough, the people who have that problem search for exactly that problem, and there you are. You don't need to outrank the world. You need to exist in a place where almost nobody bothered to show up.

What I do with it now

Now I go back and actually build on it.

I'll do the SEO properly. Then I'll reinvest the revenue — into Google Ads, into mail ads, into faxes. Pharmacies are a specific audience and you reach them where they actually are, not where it's fashionable to advertise.

More importantly, I'll keep developing the product. My goal is to provide so much value that a pharmacy can't let go of it — as tightly woven into their day as the PMS software itself. That's the standard. Not "nice tool," not "we use it sometimes." Something that becomes part of how the pharmacy runs, so that removing it would be a disruption.

The target is 100K ARR by year end. The pricing is public, so anyone can do the arithmetic: it works out to adding roughly 100 new pharmacies over four months. That's not a fantasy number. It's a number I can look at and know exactly what I have to do each week to hit it.

The part people won't like

The opportunities out there are endless. What's stopping people isn't the technology, and it isn't money. It's that they're blindsided by their own busy lives. The news, the day job, the noise. They try to keep up with the latest models on X but they don't test them. And if they do test them, they don't create anything with them. They read the announcement, they form an opinion, and that's where it ends.

You only see it by trying. That's the whole thing. You cannot reason your way into knowing what these tools can do for you — you have to hand one a real problem and watch what comes back.

And trying means failing, most of the time. I've failed at dozens of projects since Opus 4.5. Dozens. This is the one that struck. I've had partial successes elsewhere, but this was a brand new idea from scratch with minimal effort, and it's the one that turned into a paying customer overnight.

That ratio doesn't bother me, because the cost of each attempt is now so low. When a failed project costs you a weekend instead of six months and a payroll, failure stops being something you avoid and becomes something you cycle through. You're not making one careful bet. You're taking a lot of cheap shots at the net, and you only need one to go in.

Stay close to the edge

That's what I mean about staying close to the edge. Not to have an opinion about which model is best this week. To have your hands on the newest thing early enough that you can point it at a real problem before everyone else does.

The edge isn't a spectator sport. Fable 5.1 came out today and by tomorrow there will be a thousand takes on it and a handful of people who actually built something. The gap between those two groups is where all the opportunity lives right now.

My advice is simple: find a problem small enough that nobody thought it was worth solving, and solve it. Make it easy to find. Then go back and make it so good that nobody who has it would ever give it up.

I'm 3K in with zero marketing spend. Let's see where the year ends.