Most solo founders don't fail because the code was bad. They fail in the first few months, before anyone's paid, making the same eight mistakes. Every one of them feels like progress at the time. Here's what each one looks like, how to tell if you're making it, and what to do instead.
Short on time? Take The Self-Check first and read only the ones you didn't tick. Got a discovery call this week? Read Ask This Instead before it, and paste your notes into The Prompt after.
Asking friends, shipping the feature someone asked for, shaving $20 off the hosting bill. It all feels productive. Most of it is Building Theater: it looks like progress and it makes $0.
I'm building a place where you learn to build products that work like a real company's on your own, and get them in front of people who pay, with a CTO who's built two seven-figure software companies in your corner.
Make Them Once. Not Eight Times.
Free to join the waitlist. Opens at $49/mo, and the price goes up as seats fill, so first in pays the least.
Before product-market fit, a solo founder has two jobs. Find out whether enough people have a problem bad enough to pay to fix. Then find a way to reach them that you can repeat. Code is how you test the answer. It isn't the answer.
Each of these mistakes swaps a scary thing for a comfortable one. Asking a stranger becomes asking your mom. A sales call becomes a demo. Finding customers becomes building features. You stay busy, and you never have to hear that the idea isn't working.
Talk to strangers who have the problem. Ask about their past, not your future. Pay for the stuff that makes your app feel solid. Show up in one place every week. None of it is clever. All of it works.
If a week of work left you with no new facts about a customer's problem, it was the left column.
The first four happen before there's much of a product. They're about the conversations: skipping them, having them with the wrong people, or asking questions that can only get you a nice answer.
Your mom, your college roommate and your old coworker all want you to win. So they say it's a great idea, and some of them even mean it. But none of them is your customer, and none of them will tell you it's boring, too expensive or already solved by a spreadsheet. Praise from people who love you is free. As data, it's worthless.
Talk to strangers who have the problem today. If friends want to help, give them a job that actually helps: introduce you to the person they know who complains about it most.
Quick favor: who's the one person you know who complains most about [the problem]? I'd love an intro. I want to ask them how they deal with it, not sell them anything.
Cold outreach feels gross, so you build instead and tell yourself the product will speak for itself. It won't. Until people talk about your product, the only way anyone hears about it is you telling them. Founders who skip this end up with a nice app and an analytics dashboard showing three visitors, all of them you.
Make it small and boring. Ten messages a day for two weeks, to people who have the problem. Ask for advice, not a sale, and most of the ick goes away. Pick up the phone too: almost nobody calls anymore, so a short, polite call stands out. Most people won't answer. That's normal, not a verdict. Ask, Don't Pitch has ten ways to start and twenty scripts.
"Would you pay for this?" asks about the future, for a thing that doesn't exist, of someone who'd feel bad saying no. So they say yes to be nice. It costs them nothing. Then you launch, send them the link, and they don't sign up. They weren't lying. People are just bad at predicting what they'll do, and kind when you're in front of them.
Ask about the past. What they did the last time it happened, what it cost them, and what they've already tried and paid for. Money they've already spent is the only honest answer to "would you pay." And when you want to know if they'll pay, ask them to: a pre-order, a deposit, or a paid pilot. A card number is a yes. Everything else is a maybe. More in Ask This Instead.
Ten minutes in, they mention a problem that sounds like your feature, and you can't help it: "Oh, we actually do that!" Now you're sharing your screen, they're being polite, and you've stopped learning. A discovery call is for finding out how their world works. The moment you start selling, they stop telling you the truth and start managing your feelings.
Keep the product closed and let them talk for most of the call. When they say something that lines up with what you're building, don't pitch. Say "tell me more about that," and write down their exact words. Then, in the last two minutes, if it fits:
This was really useful, thank you. I'm building something for exactly the problem you described. Could I show you an early version in two weeks and get your honest take?
These two show up once there's something to try. One is about handing your product decisions to the people you're interviewing. The other is about the first impression your app makes on a stranger.
Customers are experts on their problem. They're not product designers. When someone says "it needs a dashboard" or "add a Slack integration," that's their best guess at a fix, made in ten seconds, using the tools they already know. Build every request and you end up with a junk drawer that sort of solves twenty problems and nails none. Their job is to tell you what hurts. Yours is to work out the best way to make it stop.
Treat every request as a clue to a problem. Ask "what would that let you do?" and "what happens today when you can't?" until you hit the real pain. Then design the fix yourself and test it with them. If three people ask for three different features to fix the same pain, you've found the problem. None of them found the answer, and that's fine. That part's your job.
Got it. If you had that tomorrow, what would you do with it? And what do you do today, when you don't have it?
Free hobby tiers are great for building. They're terrible for the first strangers who try your product. The app goes to sleep when nobody's using it, so the first visit hangs for ages or times out. The database pauses after a quiet week. The free email plan sends your login link to spam. Your beta tester doesn't know any of that. They just think your product is broken, and they don't come back to check.
Before you send the link to anyone you don't know, pay for hosting that stays on, a database that doesn't pause, and email that arrives. For most early apps that's somewhere around $20 to $50 a month. One lost beta tester costs you more than a year of it. Then test it like a stranger: open it in a private window on your phone after it's been idle for an hour. The SaaS Tech Stack has what to pay for, and the Speed Audit checks the rest.
The last two are the expensive ones. They're what happens when the first six go unfixed: a product nobody can find, and a founder pouring more work into it anyway.
Most solo founders spend months on the product and about an afternoon on how anyone will hear about it. But getting it in front of people is usually harder than building it, and no feature fixes that. "I'll figure out marketing after launch" tends to mean a launch post, a spike of friends, and then quiet.
Pick one channel and show up every week, starting before the product is ready. Social media, an email list, LinkedIn, cold outreach or paid ads all work for someone. Don't try them all. Pick the one where your customers already are and you can stand doing for a year. By launch day you want an audience waiting, not a post into the void. Pick A Channel compares them.
Product-market fit is usually obvious when you have it. People sign up without you chasing them, they get annoyed when it's down, and they tell their friends. When you don't have it, that's obvious too, but it's easy to look away. So founders add features, redesign the site, hire a contractor or turn on ads. All of that makes a product nobody's pulling for a bit bigger.
Admit where you are. Before fit, the job is to find ten people who'd be really upset to lose it, not to grow. Talk to the users who stayed and the ones who left. Narrow down to the kind of customer who uses it most, and change the product or the customer until somebody pulls. Then grow. A quick test: ask your users how they'd feel if they could no longer use it. If fewer than 40% say "very disappointed," you're not there yet.
Every channel works for somebody. The wrong one is the one your customers aren't on, or the one you'll quit in a month. Here's when each one fits before product-market fit.
Ask the last five people you talked to where they go to learn about their work. If three say the same place, that's your channel.
A channel you can stand for a year beats a better one you'll quit in a month. Most channels take months before they pay off.
Talk about the problem in public while you build. The people who reply are your first interviews, and later, your first customers.
Good discovery questions are about their life and their past, not your idea and their future. Start with the pain, get the facts, then close the loop. You won't need all twelve on one call.
Aim for them talking most of the call. If you're explaining, you're not learning.
Write down exactly how they describe the problem. That's the copy for your landing page.
If they're wrong about how something works, let it go. You're there to learn how they see it.
Count to five before you fill a silence. People often say the important thing right after a pause.
Tick what's true today, not what you plan to do. Each one you can't tick points to a mistake above. Start with the lowest number you missed.
Start With The ConversationsFix 01 to 04 first. Everything else depends on talking to the right people.
Most of these mistakes are invisible while you're making them. Paste your notes or the transcript from a discovery call into Claude, ChatGPT or whatever you use, under this prompt. It tells you where you pitched, what you actually learned, and what to ask next time.
You're reviewing a discovery call for a solo founder who hasn't found product-market fit yet. I'm building [WHAT YOU'RE BUILDING, in a sentence] for [WHO IT'S FOR]. The person I talked to is [THEIR ROLE, and how I found them]. Be honest, not encouraging. I'd rather hear it from you than from my bank account. Review the call below for these mistakes, and quote the exact lines: 1. Hypothetical questions: anything like "would you", "if I built" or "would you pay". 2. Pitching: where I described, demoed or defended my product before the last few minutes. 3. Leading questions: where I put the answer in the question. 4. Talk time: roughly what share of the call was me talking. 5. Feature requests I took at face value instead of asking what problem they'd solve. Then give me: - Facts I learned: things that happened, what they do today, what they've paid, what it costs them. Keep these separate from opinions and compliments. - The problem in their words: their exact phrases, ready to reuse on a landing page. - Signs of real pain, or signs it's a "nice to have". Say which way it leans and why. - The three questions I should have asked, and when in the call to ask them. - A short, friendly follow-up message to send them, with no pitch in it. Here's the call: [PASTE YOUR NOTES OR TRANSCRIPT]
Swap in your own product, customer and call. A transcript works better than notes, because notes leave out what you said.
After five calls, paste all five in at once and ask what problem came up most, in whose words. That's the start of your positioning.
Ask at the start: "Mind if I record so I can listen instead of taking notes?" Most people say yes. Follow the rules where you and they live.
Anybody can build software now. Not everybody can build a business.
It's hard to spot these from the inside. The Lab is for Founders who'd rather hear it early: weekly office hours with a CTO who's built two seven-figure software companies, and a room of other Founders who care whether anybody buys what you're making.
Less Theater. More Customers.
Want my 30-Day Launch Plan? It's for Lab members only: validate, plan, build and launch, one week at a time.
See The Solo Founder Lab ▶