Daniel sent us this one — top seven surprisingly profitable lemonade-stand-scale software businesses. By lemonade-stand-scale he means tiny teams, maybe one or two people, building things that don't look impressive from the outside but generate revenue that would make a venture-backed startup blush. The kind of thing where you hear the numbers and think, wait, that's just a PDF generator for dentists. He wants the actual list, real examples, and the mechanics of why these work.
Oh, I love this. Because the entire venture capital ecosystem trains people to chase billion-dollar markets, and meanwhile there's someone in Tulsa running a scheduling tool for dog groomers who's clearing four hundred grand a year and hasn't taken a meeting since twenty-nineteen.
Nobody writes about them because there's no funding round to announce. No launch on Product Hunt. Quietly printing money.
Before we dive in — fun fact, DeepSeek V four Pro is writing our script today. So if anything sounds unusually coherent, that's why.
Or if it starts hallucinating a pet-sitting SaaS that also does nuclear non-proliferation, we'll know something went wrong. Alright, let's do this. Where do we even start?
I think we start by defining what makes these different. A lemonade-stand-scale software business typically has three characteristics. One, the team is tiny. Solo founder, maybe a husband-and-wife duo. Two, the product solves a deeply unsexy problem for a specific industry. Three, the revenue comes from a small number of customers paying real money, not from ad-supported freemium with millions of users. You've never heard of them unless you work in that industry.
The profitability part matters. A lot of these are doing seventy, eighty, ninety percent margins because the infrastructure costs are negligible and there's no sales team.
Let's get into the list. Number one — the canonical example everyone in the indie hacker world points to — is a tool called TinyPilot. Have you come across this?
I have not.
TinyPilot is a hardware-plus-software product that lets you remotely control a computer even when the operating system is crashed. Think BIOS-level access, pre-boot, the kind of thing that would normally require driving to a data center at three in the morning. The founder, Michael Lynch, built the first version for himself, then started selling kits.
This is a one-person operation?
At its peak it was basically Michael and some contractors. In twenty-twenty-one he did over a million dollars in revenue. And he documented everything publicly — revenue, costs, churn, the whole thing. The product solves a real problem for system administrators and homelab enthusiasts, a small but fanatical market that will pay real money because the alternative is physically being there.
The margins on the software side are basically pure profit. The hardware kits have some cost of goods sold, but the software-only licenses are just money appearing in a Stripe account.
Number two is pure software. It's called CardPointers.
I've seen this one. It tracks credit card rewards and tells you which card to use for which purchase to maximize points.
Built by Emmanuel Crouvisier. He launched it in twenty-nineteen and by twenty-twenty-three was doing around thirty thousand dollars a month in revenue. That's three hundred sixty thousand a year. For an app that tells you which credit card to pull out at the grocery store.
The beautiful thing is, the people who care about credit card optimization are exactly the kind of people who will pay for a tool that does it well. They're already doing spreadsheets. They're already on the forums. They just needed someone to build the thing they were manually doing in their heads.
The switching cost is high once you've set up all your cards and preferences. The churn on these niche tools is often shockingly low — some are seeing two percent monthly or lower, where five percent is considered decent for general SaaS.
Which means the lifetime value per customer is enormous relative to the acquisition cost. What's number three?
Number three is a category more than a single product: PDF generation APIs. There's a company called DocSpring that started as basically a one-person project. Their entire product is: you send them data via API, they give you back a filled-out PDF form. That's it.
How much are we talking?
DocSpring was acquired, but before that the founder was running it profitably with revenue in the low seven figures. The key insight is that filling out PDFs programmatically is genuinely hard. The PDF spec is over a thousand pages. Every programming language has a library that sort of works, but when you're generating tax forms or legal documents or insurance certificates, "sort of works" doesn't cut it.
You're selling to developers who have tried to do it themselves and realized it's a swamp. They're happy to pay.
It's usage-based pricing, so it scales with the customer's business. You're charging per document generated, and if the customer is an insurance company sending out thousands of policies a year, that adds up fast. The founder doesn't need a sales team because developers find you when they're desperate.
There's something almost beautiful about building a business around a file format so awful that people will pay to never think about it again.
That's honestly a viable business strategy. Find the thing everyone hates dealing with and wrap it in an API. Number four follows a similar pattern: it's called Leave Me Alone.
It's an email unsubscription tool. You connect your inbox, it scans for every newsletter and marketing email you've ever signed up for, and lets you bulk unsubscribe. Built by a two-person team, Danielle Johnson and James Ivings. In twenty-twenty they were doing about twenty thousand dollars in monthly recurring revenue.
A quarter million a year, give or take. For unsubscribing from emails.
Here's what makes it work — the problem is universal but the willingness to pay is concentrated. Everyone gets too much email. Most people just delete it. But a small subset — busy professionals, executives, anyone whose inbox is a source of genuine stress — will happily pay a few dollars a month to make the problem go away. The product costs almost nothing to run. The margins have to be north of ninety percent.
It's a one-time use case for a lot of people, which you'd think would kill the business. But there's a constant churn of new people hitting email overwhelm. The top of the funnel never dries up.
Plus they added a feature where it periodically scans for new subscriptions, so it becomes a maintenance tool. Number five — and this one is my personal favorite because of how absurdly specific it is. A product called Ranch Manager.
Wait, actual ranches?
Cattle, grazing rotations, breeding records, veterinary schedules. Built by a solo developer — I want to say in Montana. This person built it for their own family's ranch, realized other ranchers wanted it, and started selling it. Last I checked, it was doing around fifteen thousand dollars a month.
The total addressable market is what, a few thousand ranches in North America?
Which is exactly why no venture-backed startup would ever touch it. The market is too small for venture scale. But for one person? A few hundred customers paying fifty to a hundred dollars a month each is life-changing money. And the competition is basically Excel spreadsheets and maybe some ancient desktop software from the nineties that still requires Windows XP.
The moat isn't technology, it's domain knowledge. You have to actually understand how a ranch operates. You can't just hire a product manager from Google and have them figure it out in a quarter.
The customers don't churn because the software becomes the system of record for their entire operation. Every calf born, every vaccination given, every pasture rotation — it's all in there. Leaving means re-entering years of data somewhere else.
Alright, what's number six?
Number six is closer to something our listeners might actually stumble across: the WordPress plugin ecosystem. Specifically, a plugin called Advanced Custom Fields, or ACF.
Oh, I know ACF. That's practically infrastructure at this point.
Built by Elliot Condon, a solo developer from Australia. ACF lets WordPress developers add custom fields to posts, pages, and custom post types. It became so essential that it was eventually acquired by Delicious Brains, and later by WP Engine. But for years, Elliot ran it as a one-person operation with a pro version that generated something like two million dollars a year in revenue.
From a WordPress plugin.
From a WordPress plugin that solved a problem so fundamental it became a dependency for hundreds of thousands of sites. The free version drove adoption. The pro version added repeater fields, flexible content layouts, gallery fields — things professional developers needed and were happy to pay for.
The WordPress ecosystem is enormous. Something like forty percent of the web runs on WordPress. Even a tiny slice of that market, if you're selling a pro plugin, is massive.
The distribution is built in. The free version is listed in the WordPress plugin directory. Millions find it through search inside their WordPress dashboard. There's no customer acquisition cost beyond maintaining the listing and answering support tickets. It's about as close to passive income as software gets.
Until the acquisition, at which point it's not passive anymore, it's just a very large check.
Number seven, and this is the one that I think best captures the spirit of the lemonade stand. It's called Picnic — not the grocery delivery company. This is a tiny app for designing picnic tables.
You're making this up.
I am not. A woodworker built a tool to help customers design custom picnic tables. You pick the dimensions, wood type, finish, leg style, and it generates a 3D preview and a cut list. The woodworker started selling access to other woodworkers and DIY enthusiasts. It did something like eight thousand dollars a month at its peak.
Eight thousand dollars a month. From picnic table design software.
The lesson is that the internet connects you to every person on earth who might want your extremely specific thing. There might only be two thousand people in the world serious enough about building picnic tables to pay for design software. But those two thousand people are findable, and they're willing to pay because the alternative is doing trigonometry on graph paper.
This is the Chris Anderson Long Tail thing made real. Except it turns out the tail is made of picnic table software and dog grooming schedulers.
Let's pull back and talk about the patterns, because there are some non-obvious lessons. Every example we just listed shares a few characteristics. One, they solve a real, specific, painful problem for a definable group of people. Not a hypothetical problem. Usually the founder had the problem themselves.
The founder was the first user. Michael Lynch needed remote server access. The ranch software guy needed to manage his family's cattle. The woodworker needed to design tables. They weren't doing market research. They were scratching their own itch and then discovered other people had the same itch.
Two, the market is too small for venture capital but enormous for an individual. If the total addressable market is five thousand customers paying a hundred dollars a month, that's six million dollars a year in potential revenue. A VC would yawn. A solo developer would be ecstatic to capture ten percent of that.
Three, the technical complexity isn't the barrier. None of these are solving novel computer science problems. The barrier is the combination of domain knowledge and the willingness to spend years building something boring.
That's the part that gets undersold. Building a PDF generation API isn't technically glamorous. It's understanding a thousand-page spec document and implementing it correctly. That takes patience and a certain temperament that is not evenly distributed among software developers.
Most developers want to work on interesting problems. They do not want to spend six months reading the PDF specification and figuring out why form fields render differently in Adobe Reader versus Chrome's built-in viewer.
Which is exactly why the people who do that work can charge a premium. The unpleasantness of the work is the moat. Everyone else looks at it and says, I could do that, but I really don't want to. And then they pay you instead.
Let's talk about the economics more concretely. What does the cost structure actually look like?
Take a typical solo SaaS doing fifteen thousand a month in revenue. Hosting costs — maybe a few hundred dollars. Payment processing — Stripe takes about three percent, so call it four hundred fifty dollars. Domain, email, basic tools — another hundred. Third-party APIs, maybe a few hundred more. Total operating costs: maybe a thousand to fifteen hundred dollars a month.
On fifteen thousand in revenue, they're keeping thirteen and a half thousand. That's a hundred sixty thousand a year in profit. For one person.
As a gross margin, we're talking ninety percent. The average public SaaS company is thrilled with eighty percent. These tiny operations are often doing better.
They don't have offices, a sales team, or a marketing budget beyond maybe some targeted ads. The customer support burden is manageable because the customer base is small and the product is well-defined.
The support thing is worth dwelling on. One of the hidden advantages is that the customers are usually sophisticated about the domain. A rancher doesn't need you to explain what a grazing rotation is. The support conversations are about bugs and feature requests, not about teaching people the basics of their own industry.
Which means the founder can handle support themselves without burning out. If you have five hundred customers and each contacts you twice a year, that's a thousand support tickets. Three a day. Compare that to a consumer app with a hundred thousand free users — the support volume is vastly higher and the revenue per user is zero.
Completely different business. Alright, so we've established these businesses exist, they're profitable, and the economics are fantastic. What's the catch? Because there's always a catch.
The catch is that they're hard to find, hard to start, and weirdly hard to sustain psychologically. Hard to find — you can't just brainstorm these ideas. You have to have lived the problem. If you're a twenty-two-year-old computer science graduate who's never worked anywhere except tech companies, you don't know what the ranching industry needs.
You might build a ranching app, but it'll solve the problem you imagine ranchers have, not the problem they actually have.
Hard to start — even once you have the idea, you need to convince the first few customers to trust you. These are industries where reputation matters enormously. A rancher isn't going to switch their entire operational workflow to software built by someone they've never heard of unless someone they know vouches for it.
The sales cycle in these niche industries is often relationship-based in a way that tech people find frustrating. You can't just run Facebook ads and expect signups. You have to go to the trade shows. You have to be in the forums. You have to build credibility over years.
Then there's the psychological piece. Running a solo software business is lonely. You're the developer, the marketer, the support person, the accountant. There's no one to celebrate wins with or commiserate with when something breaks. A lot of these founders burn out not because the business fails, but because the isolation gets to them.
I've seen this with indie hackers who post their revenue numbers publicly. The graph goes up and to the right, and then six months later they post that they're shutting down because they're miserable. Not because the money stopped. Because they couldn't do it alone anymore.
The ones who thrive tend to build community around the product — a forum or a Slack group where customers help each other. It creates social contact and reduces the support burden. Or they eventually hire one or two people, which changes the economics but saves their sanity.
Let's talk about the acquisition side. A lot of these businesses eventually get bought. What does that look like?
The acquisition market for tiny SaaS businesses has matured a lot in the last few years. There are funds now — SureSwift Capital, Tiny, and others — that specifically acquire small, profitable software businesses. They're looking for exactly this profile: stable revenue, low churn, minimal team, niche market.
The multiples are what?
Typically three to five times annual profit at this scale. So if you're doing a hundred fifty thousand a year in profit, you might sell for somewhere between four hundred fifty thousand and seven hundred fifty thousand dollars. It's not retire-to-a-private-island money, but it's a very nice outcome for a business you built on nights and weekends.
For the acquirer, they're betting they can professionalize the operation, maybe raise prices, add some features, and increase the profit over time. They're not looking for ten-x growth. They're looking for steady cash flow.
Sites like MicroAcquire, which rebranded to Acquire.com, have made it much easier to find buyers and sellers. A few years ago, if you wanted to sell your tiny SaaS, you had to know someone. Now there's a marketplace.
Which makes the whole thing more legible. Someone can build a business knowing there's an exit path that doesn't require venture funding or an IPO.
It means the lemonade stand isn't a dead end. You can build it, run it for a few years, sell it, and then go build something else. Or some founders just keep running them indefinitely because the income is great and the workload is manageable.
The infinite game. Alright, let's get practical. If someone listening wants to build one of these, where do they start? What's the playbook?
Step one is inventory your own weird domain knowledge. What industry have you worked in? What boring problem did you encounter that nobody has built good software for? If you've spent five years working in commercial laundry services, you know things about commercial laundry that most software developers don't. That's your advantage.
Step two is talk to people still in that industry. Not to pitch them software, but to understand whether the problem you saw is actually widespread. Maybe it was just your weird employer. Maybe it's a real pain point across the industry. You don't know until you have a dozen conversations.
Step three — and this is the one most people skip — is to find ten people who will pay you before you build anything. Not ten people who say they would pay. Ten people who actually give you money. If you can't find ten paying customers for something that doesn't exist yet, you probably can't find a hundred after you build it.
The pre-sale is the ultimate validation. It also solves the motivation problem. If you have ten people who've already paid you, you have an obligation to deliver. That's very different from building something in the dark and hoping someone shows up.
Step four is to build the smallest possible thing that solves the problem. Not the platform. Not the extensible architecture. The minimum thing that makes those ten customers happy. Ship it, get feedback, iterate. The nice thing about niche markets is that the customers will tell you exactly what they need.
Because they've been suffering with whatever janky solution they've been using. They know exactly what they want. You just have to listen.
Step five is to raise prices. I see this over and over. Indie hackers launch at ten dollars a month because they're afraid nobody will pay more. Then they gradually raise to twenty, thirty, fifty, and discover that churn doesn't move. The people who need the software will pay what it's worth. You're not competing on price in these niches. You're competing on being the only thing that works.
That's the counterintuitive part. In a small market, you can actually charge more because the alternatives are so bad. A rancher isn't comparing your fifty-dollar-a-month software to a competitor's forty-dollar-a-month software. They're comparing it to spending three hours a week in Excel and still getting it wrong. And three hours of a rancher's time is worth a lot more than fifty dollars.
Step six: don't quit your day job until the software business is actually replacing your income. I know this sounds conservative, but I've seen too many people go all-in too early, run out of runway, and then have to take a job while the software business withers. These lemonade-stand businesses can be built on nights and weekends. They don't require the all-consuming effort that a venture-backed startup demands.
The sloth approach. Slow and steady.
I was not going to make that joke, but yes.
I just did. Alright, we've covered the list, the economics, the patterns, the playbook. Any cautionary tales?
One is platform risk. A lot of these businesses are built on top of other platforms. The WordPress plugin ecosystem is great, but if WordPress changes its architecture or terms of service, your business could be in trouble. Same goes for businesses built on Shopify or Salesforce.
The ACF story is instructive there. It survived because it was so deeply embedded that WordPress couldn't easily break compatibility without breaking a huge chunk of the web. But that's the exception. Most platform-dependent businesses are more fragile than they look.
Another risk is the key-person problem. If the business depends entirely on one person, what happens if that person gets sick, or wants a vacation, or just gets tired? There's no redundancy. The business is one bad flu away from a support backlog that drives customers away.
The solution is usually to eventually hire someone, even part-time, who can handle support and basic operations. But that's a big psychological step for a solo founder who's used to doing everything themselves.
It changes the economics. That hundred sixty thousand a year in profit becomes a hundred twenty after you hire someone. Still great, but different.
Worth it for the ability to sleep through the night without worrying about server alerts.
We should also mention that not every niche is equally good. Some are too small even for a solo founder. If there are only fifty potential customers in the world, and they're only willing to pay twenty dollars a month, the math doesn't work.
The sweet spot seems to be somewhere between two hundred and two thousand customers, paying between thirty and three hundred dollars a month. That gives you a revenue range of six thousand to six hundred thousand a month, which covers everything from a nice side income to a very comfortable full-time business.
The size of the addressable market matters less than the intensity of the need. If only five hundred people in the world need your software, but they need it desperately and will pay two hundred dollars a month, you've got a hundred thousand a month in potential revenue. That's a great business for one person.
The picnic table guy figured this out. There aren't that many serious picnic table builders in the world. But the ones who exist really care about picnic tables. They're enthusiasts and professionals. That's the ideal customer profile.
Now: Hilbert's daily fun fact.
Hilbert: The average cumulus cloud weighs approximately one point one million pounds, roughly the same as a hundred elephants, and stays aloft because the weight is distributed across millions of tiny water droplets spread over a vast volume of air.
When someone says they have their head in the clouds, they're carrying a lot of weight.
...right. Thanks, Hilbert.
So here's the forward-looking thought I want to leave listeners with. We're in this strange moment where AI is supposedly going to automate all software development, and yet the businesses we just described are more viable than ever. Because the hard part was never the code. The hard part was knowing what to build and who to build it for.
AI can help you write the PDF parser faster. It cannot tell you that the commercial laundry industry desperately needs better route optimization software and that the current market leader hasn't updated their product since two thousand twelve. That knowledge only comes from being in the industry.
The lemonade stand is safe, at least for now. The moat is knowing something the AI doesn't know and caring about a problem the AI doesn't care about.
Thanks to our producer Hilbert Flumingtop. This has been My Weird Prompts. You can find every episode at myweirdprompts dot com. We'll be back next week.
Until then, keep your businesses small and your margins high.