Can you make money building a niche version of old, boring software?

A buddy of mine built a Google review management tool for self-storage owners. He's making about $12,000 a month and is probably headed past $25,000 in a few months. Review management has been around for 10-plus years. He went super deep into one seemingly tiny niche that's still big enough for one-person software.

Why does an old category work?

The demand is already proven. Nobody needs convincing that reviews or forms matter. What's missing is the version shaped for their industry.

"I have taken a concept that's been around 10, 15 years, forms ... and built the niche version of it."

That's greetform: a survey, conditional lead cards and an offer wall that show up right after someone subscribes to a newsletter. I didn't realize it at the time, but it's basically a form system like Jotform, built specifically for newsletters. On paper it's just a form. For a newsletter operator it's so much better than any general form software, because it was built for them.

Old categories in niche industries are often stuck in the past too. The software local media companies use was largely built around 2004 and is expensive and terrible. Adding a rewrite button to legacy software is nothing. Real AI-first software is different: agents scrape every interesting thing happening in your community overnight, draft it in your voice, and you edit and publish in the morning.

Why is a boring niche better than a hyped one?

Nobody is sitting there going, "I want to make a legacy media AI-powered OS." That's the point. There's much less competition, so you can be 10 times better. It doesn't look big enough to most builders, but for a team under five people it is big enough.

My hypothetical: get to 1,200 users at $500 a month in a space nobody cares about. Nobody's inviting you onto the big business podcasts for that. You'd still make a lot of money.

If you build something too obvious, you get too many big competitors. If you build something small, the opportunity right now is huge.

Why do you need to live the problem?

Only the person living it would think to solve it. A Silicon Valley builder doesn't wake up asking what happens after someone subscribes. A newsletter operator does.

My software lasts because it solves specific problems I only know because of what we do every day. A random AI guy on Twitter who heard about local media won't know the actual problem. Get in the weeds and you start to find the wedge. Once you own a niche, you also spot small side projects that could make $5,000 or $10,000 a month and launch them easily.

Does the software have to look polished?

No, it has to serve the right people. A software founder I met sells a tool for finding newsletter advertisers at about $300 a month. It has that vibe-coded look. Mostly it scrapes newsletter data and LinkedIn profiles. For brands selling $20,000 to $50,000 a month in ads, paying $300 for that data is a no-brainer.

If you know what to scrape and how to present it, you can build a tool that's insanely valuable to one audience. Think about that in every industry, because these problems are invisible unless you're inside it.

Can I copy an expensive niche tool with AI?

Often a big chunk of it. A friend showed me a real estate marketing tool charging about $300 a month that he said makes $6 million a year (his number, not one I've verified). I asked Claude, and it said about 80% of it could be duplicated relatively easily. A lot of what it offers is content creation, which is what my team already does.

In the end, they're all wrappers. The winner is the best wrapper, the one that does all the things users actually need that nobody else does.

What does winning against weak incumbents look like?

Back in 2025, software I built for insurance agencies had only two competitors, and people disliked both. One was missing the retention system we had. The other cost about $2,400 a year against our $99 a month. Everybody wanted a reason to leave what they had.

• We closed about 60% to 70% of the people we talked to.
• We got zero complaints about price. Those agents spend so much money that $100 or $200 a month means nothing if you solve their problem.
• The bottleneck was getting in front of enough of them.

How do I find a problem worth solving?

Start from an audience. There's a famous 2008 essay by Kevin Kelly called "1,000 True Fans." My version of the math: find a thousand people who like and trust you and will pay you $10 a month, and you have $10,000 a month.

AI changes the build side of that math. If you have an audience of 1,000 people in a vertical who all have a problem they'd pay $20 to solve, and you can build it in a week, build it. There are probably ten more problems like it. Problems that never got solved because the build cost more than the benefit are worth solving now.

That's what happened to me. I'd started in this space about 18 months earlier knowing nothing, but I had the audience, which most software builders lack.

How do I sell it?

Go where the industry already gathers: its conferences and Facebook groups. Then let the product market itself:

• Every greetform page carries a light gray "powered by" link in the footer that opens in a new tab, so it doesn't interrupt anyone.
• When people buy a ticket to my local media event, they go through a greetform survey, and the offer wall at the end asks if they want greetform for their own local media. They've already used the product.
• A freemium tier, so I can market it to anyone with a newsletter.

When is a niche too small?

When the audience is small and incumbents have a 10-year head start. I shut down my earlier insurance-agency software for that reason, after a lot of money and effort. Niche software should either be easy to build or aimed at a bigger market. More in why good software products fail.