How to get to $10k MRR as a solo founder

You can build a product in a weekend. AI writes the boring parts. Pieter Levels said it in June 2026: everyone can build apps with AI. Almost nobody has an audience, the cash for ads, or a free way to get attention. $10k MRR is not a prettier landing page. It is money that lands every month from people who still care.

A few days later he wrote about indie meetups full of people building spaceships: factories that spit out landing pages and dashboards, almost none with money or traffic. They say marketing starts when the system is finished. That order will not pay rent.

Indiecity is for one person or a team under twenty. You do not need a growth team. You need three things you can name: a price high enough that the math works, customers who stay, and one channel you work every week. If you do not have ten payers yet, stop and do how to get your first 10 paying customers without ads first.

What $10k MRR actually is

$10,000 a month is price times paying accounts, after people who cancel. That is all. One hundred customers at $100. Fifty at $200. Twenty at $500. Ten at $1,000 if the offer is sharp and the buyer has budget. Same line on a bank statement. Different lives for a solo founder.

Do not invent a month-by-month hockey stick to feel in control. Do not copy someone else’s public MRR screenshot as a schedule. Write the equation with your real average price and your real count of people who still pay. If the product is free to try and free forever, you are counting users, not MRR.

Levels has put real numbers on the internet for years. In 2018 he said Nomad List made $15,000 to $25,000 a month in membership fees after years of work, not after a weekend launch. In 2023 he published Photo AI at about $61,808 a month with 1,872 paying customers on a small PHP codebase. The point is not to copy his stack. The point is that $10k is ordinary math on a product people keep paying for.

Price so the math works alone

At $29 a month you need about 345 active paying accounts to hit $10k before refunds and failed cards. At $99 you need about 101. At $299 you need about 34. Support load follows customer count. A cheap sticker does not make support cheaper. Solo founders who underprice often invent a second product instead of fixing the number.

Patrick McKenzie wrote in 2006 that you can probably stand to charge more. It still holds in 2026. You see every wart. They see whether the problem gets smaller when they pay. Charge a number you can say on a call without wincing, large enough next to the mess you remove that $10k does not require a crowd you cannot serve.

If ten people already pay a joke price, raise for new buyers first. Keep early adopters on the old plan if that is how you sleep. The next person who has never paid only ever sees the number that makes the solo path possible. A pretty three-tier grid with invented columns is not research. One clear offer is enough until two different buyers show up.

Retention is the silent half of MRR

New sales feel like progress. Churn erases them while you sleep. If you add ten and lose ten, your dashboard can look busy and your MRR can sit still. Before you buy ads or open a second channel, know why people leave and whether the product delivered the outcome you sold.

Do not paste a “healthy SaaS churn” percentage from a fund deck into your plan. What a good early churn rate actually is has the public medians. Count exits this month. Write the reason in their words: never activated, wrong buyer, problem went away, support was slow, a cheaper tool appeared. Fix the top reason with product or positioning. Buying more traffic while people leave is a treadmill you will hate by month four.

• Who paid last month and still pays this month
• Who left, and the one sentence reason
• Whether they ever got the outcome you sold in the first two weeks
• Whether you sold the wrong job title because they were easy to reach

When someone renews, ask what almost made them cancel. That answer is worth more than a survey widget. Keep people who got a real result. Stop discounting people who were never going to stay. If the list is still small enough to call, reducing churn with twenty customers is the method.

One channel, not five hobbies

Rob Walling laid out the Stair Step Method as a path bootstrappers actually finish. Step one is not a multi-channel growth machine. He writes that the strategy with the best chance of success is a simple product with a simple marketing plan that only needs a single traffic channel. Get good at one way of finding customers. Do not try to master every channel at once.

In his own first step he got good at SEO instead of learning SEO, AdWords, and Facebook ads in parallel. The lesson for $10k is the same. Pick the motion that already produced a paid customer, or the room where buyers already complain in public. Work it every week until you know reply rate, booked calls, and paid. Kill the rest for a month.

Good solo channels look boring:

• Warm intros from people who already paid
• Outbound to one job title with a short, human note
• One niche forum, Slack, or community where the buyer already talks
• Search pages written only in the words buyers already used on sales calls
• A partnership with one adjacent tool or newsletter that serves the same buyer

Levels’s 2026 point fits here. Building is cheap. Attention is not. If you have no audience and no ad budget, your free path is still a room, a list of names, or a channel you can grind without a team. Spreading across five half-done experiments is how you stay stuck around $1k MRR while the factory looks impressive.

Stair step, not a spaceship

Walling’s stair step is a climb: first product with one channel, enough revenue and skill to own your time, then heavier recurring products when you can carry them. You do not need to reenact his exact companies. You need the order. Prove cash with something you can sell and support alone. Grow the channel that works. Add complexity only when the simple thing is boringly reliable.

That is the opposite of the AI factory Levels mocked: perfect systems, zero money, marketing scheduled for after the tooling is done. Ship the offer. Charge. Keep people. Repeat the one channel. The product can stay small. Photo AI’s public story was thousands of paying customers on a codebase Levels described as roughly 14,000 lines of PHP, not a platform rewrite. Small and paid beats large and empty.

A week you can run until $10k

1. Monday. Write average price, active paying accounts, and last month’s exits with reasons. No chart from the internet. Your numbers only.
2. Monday. If price times realistic headcount cannot hit $10k without a crowd you cannot support, set a new price for new buyers. Say it out loud once.
3. Tuesday to Thursday. Work only one channel. Fixed volume: messages sent, posts answered, intros asked, pages shipped. Track sent, replied, booked, paid.
4. Tuesday to Thursday. Talk to one person who almost churned or who left. Fix one activation or positioning issue, not five features.
5. Friday. Ask every happy payer for one intro or one public word. Write the exact phrase they use for the problem. Put that phrase on the page.
6. Friday. Kill any second channel you opened out of panic. If the primary channel produced zero paid after real volume, change the offer or the buyer, not the tool stack.

If this takes [pain in their words] off your week, it is [price] a month. I can start you today. Want to do that?

Glad [result] landed. What almost made you quit in the first two weeks? I am fixing that for the next person. Also: who else on your team still deals with [pain]?

Hey [name]. You wrote about [specific mess]. I charge [price] to [outcome] for [job title]. Fifteen minutes this week if that is still open. If not, who should I talk to instead?

How you know it is working

MRR up and to the right for boring reasons: higher price on new deals, fewer exits with the same complaint, and a steady drip of paid from one channel. People ask the price before you do. Customers introduce peers without a points program. Support questions shift from “how do I log in” to “can it also do this.” That second question is a new thing you can charge for, not a bug.

If price is a joke, fix price. If people leave before they get value, fix onboarding and who you sell to. If nobody new pays, the channel or the sentence is wrong. Do not open three new channels to hide one weak offer. Do not build v4 so you can avoid the calendar.

What to stop doing

• Building the AI factory before anyone pays enough to matter.
• Copying a stranger’s month-by-month MRR chart as your plan.
• Running five marketing experiments at 20% effort each.
• Charging $9 because that is what you have seen on other landing pages.
• Buying ads to avoid asking for intros or writing ten human notes.
• Counting free users and waitlist signups as progress toward $10k.
• Adding features to stop churn caused by the wrong buyer.

Then the number is real

$10k MRR as a solo founder is a price that makes the math possible, customers who stay because they got the outcome, and one channel you refused to abandon for shiny work. Walling’s stair step is still the honest shape. Levels’s 2026 warning still stands: shipping is easy, getting paid is the job.

When you cross it with a product people can name, write the story. When the product is real and the name is yours, put it on the map. You can join Indiecity while you climb. That gives you people to talk to. It does not replace the price, the retention work, or the channel on your calendar.

Until then, say the number, keep the people who got value, and work the one channel again next week.