Claude Code ships blind — no context gauge, no quota counter. I built a 7-segment statusline and baked in a 50% context handoff rule, backed by Lost-in-the-Middle research. Every segment, every color rule, and the full production script explained.
Intelligence isn't about what you know — it's about how you interact with the world. That one insight explains why AI is growing up in exactly the same order as a human child, and doing it a million times faster.
70 prompt tips for Claude Code are floating around. 7 actually moved the needle for me. Real before/after terminal sessions, a copy-paste cheat sheet, and the 3 patterns I dropped.
I had five business ideas in a notes app for three months and picked none of them. This is the framework that unstuck me — two circles, six categories, a seven-step market test, and a 30-day validation sprint. For non-technical solopreneurs.
TL;DR
For three months, I had five solopreneur business ideas sitting in a notes app and shipped none of them.
AI tools. An online course. A paid newsletter. A services agency. A curated directory.
Every morning I'd open the app, stare at them, re-read each one, maybe edit the wording, close the app. By afternoon I'd be back on Twitter watching other people ship things. This went on until a friend asked me, "Which of those five are you actually doing?" And I realized the honest answer was none.
I didn't have an idea problem. I had a picking problem.
This post is what I wish someone had handed me on month one of that stuck period. If the previous post convinced you that one person is enough, this one is about the next question: what is the one thing you're going to do?
If you're earlier than that ("I don't have any ideas at all"), start with Why One Person Is Enough for the case, and read the skill-transfer section below — it often unlocks ideas you didn't know you had.
The instinct most people reach for first: do what you love.
It's not wrong. It's incomplete.
A study of US college students
— but only about 3% of jobs exist in those three fields combined. Loving tennis doesn't make you the next Serena Williams. There aren't enough seats at that table.
Cal Newport's So Good They Can't Ignore You flips the usual narrative: passion is a byproduct of mastery, not a prerequisite. You don't love a thing first and then get good at it. You get good at a thing, and somewhere along the way, you start to love it.
This matters because "follow your passion" pushes you toward whatever you're most emotional about today. That's rarely where you have the skills, positioning, or market. The better question isn't what do I love? but what do I have skills in that people will also pay for?
That's two circles, overlapping. The overlap is your opportunity.
Draw two circles.
Left circle: your passions and your skills. Not just what you like — what you've gotten good at, whether or not it's a hobby, a side thing, or part of your day job.
Right circle: what the market is willing to pay for. What people are already spending money on, right now, in the world you actually live in.
The overlap is your territory.
Written as a formula:
Passion or skill × Usefulness = Opportunity
Scott Adams, creator of Dilbert, is the textbook example. He wasn't the best illustrator in the world. He wasn't the best writer. He wasn't the funniest person. He had basic business experience. But Dilbert is the intersection of all four of those skills, each at a good enough level, applied to a specific audience (office workers). Plenty of people were better than Adams at any one of those skills. Almost nobody had the exact combination he did.
Adams calls this the "talent stack," and it's one of the most useful frames I've found. You don't need to be the best at any one thing. You need a rare combination of skills that are each at least decent.
There's a version of the two-circle model most people get wrong on their first try. They list their job title's skills on the left circle ("I'm a teacher — I can teach"). They miss that every job contains a second skill set, the one underneath the job description.
A teacher doesn't only teach. She also:
Any of those, repackaged, can be a business. An online course creator uses the same skills. A B2B trainer uses the same skills. A corporate communications consultant uses the same skills.
Don't ask "what am I good at?" Ask "what am I good at that I've been giving away for free inside my current role?" That's usually where the real talent stack is hiding.
Sahil Lavingia's origin story for Gumroad is exactly this pattern. He was a designer at Pinterest. The skill he commercialized wasn't design — it was knowing what creative people need to sell their work, which he'd built up during years inside a creative-tools company. The job title was "designer." The real talent stack was "empathy for creators + product sense + writing."
Once you have a sense of your talent stack, the next question is: which type of business is this a fit for? There are roughly six categories I've seen work consistently for non-technical solopreneurs. Here's each, with its 2026 fit rating.
You package knowledge into something you can sell repeatedly — a course, an ebook, a paid newsletter, a template library, a research service. Justin Welsh built a $5M+ LinkedIn-driven solopreneur business primarily around information products. This is also what AWP is.
2026 fit: extremely high. AI accelerates content production 3-10x without replacing the underlying insight you bring. The cost structure is nearly zero.
Consulting, coaching, copywriting, specialized freelancing. The difference between a $75/hour freelancer and a $15,000/engagement consultant is framing and positioning, not talent.
2026 fit: high. AI lets one person deliver what used to take a small agency. The premium rate is justified by outcome, not hours.
You design the product, outsource everything else. Printful, Printify, Shopify, Amazon FBA. You never touch inventory. Etsy is full of six- and seven-figure operations run by one person.
2026 fit: medium-high. Market is more saturated than it was in 2020, but AI helps with product research, ad copy, and design iteration.
A tool serving one specific niche — scheduling for dog walkers, CRM for dental practices, invoice templates for musicians. You don't have to code it yourself; outsourcing development for a v1 is cheaper than ever. Pieter Levels' Nomad List and RemoteOK fit this pattern.
2026 fit: medium. Requires more upfront investment and some technical collaboration, but AI is lowering the barrier quickly.
Real estate, stock photography, music licensing, course royalties on platforms. You build or acquire an asset once; it earns on its own schedule.
2026 fit: medium. Slow to build. Very stable once it's working.
Directories, job boards, niche newsletters with ads, curated community platforms. Arvid Kahl's Zero to Sold is built on this pattern and has become a small industry of its own.
2026 fit: high for curators who already have an audience. Hard from scratch.
Notice what's common across all six: none require you to be an engineer, none require you to do every task yourself, and none require outside capital to start. Being a solopreneur means being the only decision-maker, not the only worker. All six let you buy execution (design, development, fulfillment, customer support) while keeping ownership and direction.
So you have a direction. The instinct now is to build the product. Resist that instinct. Build validation first, product second.
Here's a compressed version of a market-test method that works for solo operators:
You have to be willing to work on this for 2+ years through plateaus. If you're not, it won't last. But "passion" alone isn't enough — see above.
Cheap test: search your target term on Google. Are there ads on the results page? Ads mean someone's paying to get in front of this audience, which means there's money there. No ads at all in a mature market is usually a bad sign.
Selling to someone who already knows they have a problem and is actively looking for a solution is 10x easier than selling to someone you need to first convince they have a problem.
Almost all purchases solve a deep pain or a deep desire. People don't buy luxury watches for timekeeping — they buy status, belonging, self-image. Pain-relief usually sells better than desire-fulfillment. Pain is urgent. Desire is optional.
Competing on price is a race to the bottom. One-person businesses can't win it; they don't have the volume. Differentiate on quality, positioning, audience, or delivery — never on price alone.
"What do you think of this idea?" asked to your brother is worthless. Build one clear customer profile — the specific person most likely to pay for this thing — and talk to five of them in depth. What do they currently do about this problem? What would they pay to avoid it? What would make them switch?
Build a minimum version. Offer it to 5-10 target customers for free or discounted in exchange for honest feedback and a case study. If no one wants the minimum version, no one will want the full one.
The underlying logic of all seven steps is sell first, build second. Someone on Twitter wrote up this exact pattern recently: a consultant posted an ad for a high-end service he hadn't actually built yet. Two buyers said yes. He told them the "premium version" would ship in 30 days, then spent those 30 days actually building what they'd paid for. That's pre-selling. That's the sequence.
You don't need a 40-page document. You need three things on one page.
| Element | The question it answers |
|---|---|
| One product or service | What are you selling? |
| One target audience | Who is paying? |
| One delivery / payment method | How does money flow in? |
Plus a one-sentence mission:
I help [customer] achieve [core benefit / outcome].
For a dog-walking service: "I help busy owners feel calm when they can't be there for their dog." Notice that's not "I walk dogs" — that's the function. "Feel calm" is the benefit. Customers don't buy your actions; they buy the outcome your actions produce.
For a knit hat pattern: "I help people experience the joy of making something with their hands." Not "a PDF with knitting instructions."
Functions vs. benefits is the single most common business-writing error. Every landing page, every social post, every email should lead with the benefit. Train yourself to ask, "yes, but what does this do for them?"
The standard advice is to build an MVP — a minimum viable product. For solopreneurs, there's a harder-edged version: minimum viable profit.
A product can lose money. Profit can't. If you define the smallest version of your business that breaks even, your timeline is much shorter than the VC-style "we'll make it up in volume."
Two classic examples:
The pattern: don't wait for the product to be polished. Get to the first paying customer as fast as possible.
Solopreneurs have one unfair advantage over big companies: you can serve markets too small for them to care about.
In a niche market, you don't compete with everyone. You only need to be the best in a very small pond. Niche isn't a limitation — it's a moat. Trust builds faster inside a narrow audience. Expertise reads as authority more clearly. Word of mouth travels faster in a tight community than in a broad one.
The phrase "the riches are in the niches" is a cliché for a reason. Justin Welsh is specifically a LinkedIn solopreneur educator. Arvid Kahl is specifically a bootstrapped SaaS coach. Pieter Levels is specifically for digital nomads. None of them are trying to be for everyone. All of them make more than the generalists.
Here's the compressed sprint I ran on AWP itself, which you can copy.
The goal isn't success in 30 days. The goal is a clean go / iterate / kill decision in 30 days. Most ideas fail this test. That's fine — failing in 30 days is a feature, not a bug. Failing in 30 days is how you get to the idea that does work within a reasonable timeframe.
For context, this is how I picked AWP out of my original five.
I applied the two-circle model. My left circle was: writing, teaching complex tools to non-technical people, and running a one-person operation. My right circle was: what people currently pay for. The overlap was narrow but distinct — practical AI workflow content for non-engineers who want to build a business using AI tools.
Four of my five ideas failed one of the two circles. The "AI tools" idea didn't have a clear customer. The "course" idea I wasn't uniquely positioned for. The "services agency" idea required skills I'd have to hire for, which defeated the solo point. The "curated directory" idea was too passive for what I wanted to do every day.
AWP was the one that sat cleanly in the overlap. It also fit my talent stack (writing + teaching + working with AI tools at the application layer, not the research layer). It fit category #1 (information products) and #6 (curation/community). It survived all seven of the pre-sale validation questions.
Six months later, I'm still on it. That's the test. You want an idea that six months from now you're still excited to open.
Start with the "hidden skill set" exercise. Write down your current or most recent job title. Under it, list everything you actually do day-to-day — not the job description, the real tasks. Half of those are transferable skills that could be packaged into a business. Communicate complex things? Course creator, technical writer, consultant. Handle client relationships? Account manager as a service, boutique agency. The ideas often hide inside the role you already do.
Cheap test: search for your niche term on Google and scroll through the ads. Ads mean advertisers are willing to pay to reach this audience, which means there's money. No ads = either too small or not commercial. A more expensive test: check keyword search volume using a tool like DataForSEO or Ahrefs. Anything with 500+ monthly searches on a commercial keyword has real potential.
No. Serial, not parallel. Testing two ideas at once means you're doing a bad job of both. Pick the one you'd most regret not testing, give it a clean 30-day sprint, and kill it or commit to another 90 days. Parallel testing is what keeps you stuck in your notes app for three months.
No — you need an original angle. Most successful solopreneurs solve an existing problem for an underserved niche, not an unsolved problem for everyone. Paul Jarvis didn't invent consulting. Justin Welsh didn't invent LinkedIn content. Pieter Levels didn't invent job boards. They all took something established and narrowed it to a specific audience with their own voice.
Three options: (1) outsource the v1 on Upwork or a no-code platform ($2K-$10K typically gets you a working MVP), (2) find a technical co-founder — though be careful with equity splits, (3) start with a non-code version of the business (a manual service, a newsletter, a consulting practice) and fund the code build from revenue later. Crew/Unsplash did #3. So did CDBaby.
A fourth option has quietly become real in 2026: build the first version with an AI coding assistant yourself. Non-programmers using Claude Code ship small-but-real MVPs in 2-4 weekends — not production-grade, but enough to validate paying demand before you spend $10K outsourcing. If you're curious whether this path is realistic for you, my 6-month Claude Code review has the honest numbers on cost, time, and what it actually takes; start with 10 mistakes I made in week one before you open a terminal.
The common rule is six months of your real monthly expenses. I'd add: plus at least one paying customer on the new thing. Six months of savings without revenue buys you a long job-hunt, not a business. Six months of savings and a first customer buys you a real ramp.
You have the framework. The remaining work is doing the sprint — which is not something a blog post can do for you. Next in this series: Starting a Solopreneur Business in 2026 — the first 90 days of an actual launch, covering pricing, your first customers, and the mistakes I'd avoid if I started over.
Pair this post with the macro argument → the structural case for why one person is enough in 2026. And for the research that grounds the 2026 solo math — AI cost collapse, agentic commerce, the $1.7T non-employer economy — read eight numbers from Stanford HAI's 2025 AI Index and MIT Technology Review's 2026 AI predictions.
— Leo
Intelligence isn't about what you know — it's about how you interact with the world. That one insight explains why AI is growing up in exactly the same order as a human child, and doing it a million times faster.
The CLAUDE.md I actually use — 67 lines, rewritten three times. Full copy, what I deleted between versions, and the six rules that made the final version stick.
29.8 million US solopreneurs, $1.7T in revenue, 81.9% of small businesses with zero employees. Four assumptions keep most people from starting. Here's what breaks each one — with Pieter Levels, Pinboard, Midjourney, and the four pillars every durable one-person business runs on.
Stanford HAI's AI Index 2025 is the most comprehensive snapshot of AI we have. 457 pages, 8 chapters. The core story isn't any single number — it's multiple trend lines hitting tipping points at once. Here are the 8 numbers that capture it, and what each one means for a non-technical solopreneur.
Hands-on AI coding tutorials and workflow deep-dives, straight to your inbox every week.