The questions founders
actually ask, answered.
Short answer first, then the depth.
What to do this week, what works in this market, and the principles that have held since long before software.
A new playbook every Sunday.
Every Sunday, 09:00 UTC. Written ahead, dated, and released on its own. Nothing goes out that doesn't answer a real question.
How people launched before no-code
They sold first and built after. Without cheap tooling, the only affordable way to test an idea was to find a buyer, take a deposit, and deliver by hand — with a spreadsheet, a phone and their own labour. That constraint produced better businesses, because nothing got built without someone already paying for it. The method still works and is still faster than building.
Why some founders break and others endure
Not talent, not funding, and not resilience as a personality trait. The founders who endure have a low burn rate in their own life, a source of identity outside the company, evidence of progress they can see weekly, and at least one person they can tell the truth to. The ones who break usually had all their meaning, money and self-worth in a single line item.
What still wins when the tools change
Learn the things that were true before your tools existed and will be true after they're replaced: how to find someone with a problem, how to ask for money, how to keep a promise, how to decide with incomplete information, and how to keep going when it's dull. Tools are rented. These are owned.
- Building a business in 1990 vs 2026Is it genuinely easier to start a company now, or does it just feel that way?Sunday 2 August · released
- AI business vs traditional businessShould I build an AI company, or a normal company that uses AI?Sunday 9 August · released
- What still wins when the tools changeEvery year there's a new stack. What's actually worth learning deeply?Sunday 16 August · released
- Why some founders break and others endureWhat actually separates the people who last from the people who quit?Sunday 23 August · released
Pick the sentence that sounds like you.
Each path is an ordered route through the Library. Five playbooks, in the right order, for where you actually are.
A named buyer, a paid ask, and a first version you can ship in weeks.
- 01The Real First Steps of Starting a Business
- 02What Founders Get Wrong at the Beginning
- 03How to Start a Business in 2026
- 04Should You Build a Website or a Product First?
- 05How to Launch Without Wasting Time
A build path that doesn't require a co-founder or a large budget.
- 01What to Use If You Are Not Technical
- 02When to Use No-Code — and When Not To
- 03Should You Build a Website or a Product First?
- 04The Founder Tool Stack in 2026
- 05How to Launch Without Wasting Time
A business designed around thirty customers, not three hundred.
- 01The One-Person Business Playbook
- 02The Best Business Models to Start With Now
- 03Founder Finance Before You Have Any
- 04Marketing When You're One Person With No Budget
- 05Ownership: Keeping What You Build
An operating rhythm and clear decisions instead of constant reaction.
- 01How to Run a Company Without Keeping It in Your Head
- 02How to Make Decisions When Nobody Is Coming to Help
- 03Founder Finance Before You Have Any
- 04Why Some Businesses Last and Others Break
- 05Ownership: Keeping What You Build
The judgement layer: archetype, history, durability, turning points.
- 01Timeless Founder Lessons That Still Win
- 02The Four Founder Archetypes — and Which One You Are
- 03How Entrepreneurs Built Before Software
- 04Founder Turning Points: What the Big Journeys Teach
- 05Why Some Businesses Last and Others Break
The ones that change the most decisions.
The Real First Steps of Starting a Business
Write your idea as one sentence naming a specific person and a specific painful moment. Then have ten conversations with that person about the last time it happened. Then ask one of them for money — a deposit, a pilot fee, a pre-order. Everything else (name, logo, entity, website, product) comes after somebody has tried to pay you.
What Founders Get Wrong at the Beginning
Five failures cause most early deaths: building before talking, choosing a market too broad to describe, mistaking encouragement for demand, spending money to feel legitimate, and hiding behind work that cannot fail. All five are avoidance dressed as productivity.
Should You Build a Website or a Product First?
Build a website first if you sell something a human delivers — services, consulting, physical goods, local businesses. Build a product first if the product itself is the only way anyone can experience the value. When genuinely unsure: a one-page site plus a manual, human-delivered version of the service beats a half-built product every time.
How to Start a Business in 2026
The mechanics are cheaper and faster than ever — you can be selling within a fortnight — but the scarce thing has moved. Building is no longer the bottleneck; being believed is. In 2026 the winning sequence is: pick a specific buyer, sell before you build, assemble rather than construct, and spend your real effort on distribution and proof.
How to Start a Business in 2027
Three shifts are already visible: buyers will assume AI is in everything and stop paying for it as a feature; distribution will keep getting harder as generic content saturates; and trust, provenance and human accountability will become the premium. Build for a market where the product is assumed competent and the company is the differentiator.
What to Use If You Are Not Technical
Further than a technical founder could have got five years ago, and far enough to reach a real business. Your genuine constraints are debugging, security and scale — not building. The correct strategy is to build the first version yourself, sell it hard, and buy technical help only once revenue justifies it and you can specify exactly what you need.
How Entrepreneurs Built Before Software
They did exactly what works now, slower: found a group of people with an unmet need, took a personal risk to serve it, delivered by hand, and reinvested the proceeds. Every era's founders had the same four constraints — capital, trust, distribution, and delivery — and the only thing technology has changed is the price of the last one.
Timeless Founder Lessons That Still Win
Nine things survive every technology shift: proximity to the customer, focus, cash discipline, reputation, pricing courage, distribution ownership, patience with compounding, honesty about what is not working, and staying alive long enough to get lucky. Everything else in founder advice is a temporary tactic wearing a principle's clothes.
Your first 10 customers
They come from people you can name, contacted one at a time, with an offer specific enough to feel written for them. Not from a channel, not from a launch, not from an audience. At ten customers you are still doing manual work that will not scale, and that is correct — the point of the first ten is to learn what makes someone say yes, cheaply enough that you can change the answer.
How to run a weekly founder review
Sixty minutes, same slot, same six questions, written down. Money in, money out, what moved, what is stuck and who owns it, what I promised and did not do, and the one thing next week is for. The value is not the hour — it is that decisions stop depending on how you happen to feel that day.
Cash collection before it becomes a crisis
Collection is a system you set up before the invoice exists: terms agreed in writing, an invoice sent the day the work is accepted, a fixed reminder sequence, and one person who checks the ageing weekly. Companies rarely die from bad pricing. They die from good revenue that arrives too late.
Every playbook.
The Real First Steps of Starting a Business
Write your idea as one sentence naming a specific person and a specific painful moment. Then have ten conversations with that person about the last time it happened. Then ask one of them for money — a deposit, a pilot fee, a pre-order. Everything else (name, logo, entity, website, product) comes after somebody has tried to pay you.
What Founders Get Wrong at the Beginning
Five failures cause most early deaths: building before talking, choosing a market too broad to describe, mistaking encouragement for demand, spending money to feel legitimate, and hiding behind work that cannot fail. All five are avoidance dressed as productivity.
Should You Build a Website or a Product First?
Build a website first if you sell something a human delivers — services, consulting, physical goods, local businesses. Build a product first if the product itself is the only way anyone can experience the value. When genuinely unsure: a one-page site plus a manual, human-delivered version of the service beats a half-built product every time.
The Best Business Models to Start With Now
Pick a model by how fast it turns effort into cash and how well it fits your unfair advantage — not by which model is fashionable. For most founders starting today, the strongest openings are productised services, vertical software for an unglamorous industry, operator-led local businesses, and paid communities or education with a real outcome. Pure consumer apps and marketplaces are the hardest games on the board.
Founder Finance Before You Have Any
Four numbers run an early business: cash in the bank, monthly burn, cash collected this month, and the price you charge. Track them weekly in one sheet. Price higher than feels comfortable, collect faster than feels polite, and never confuse revenue signed with money received.
How to Launch Without Wasting Time
A launch is not a day, it is a sequence of small ones aimed at progressively colder audiences. Start with the twenty people who already know you, then one narrow community, then a public moment. If you do not have a list of thirty people to tell, you do not have a launch problem — you have an audience problem, and that is what to fix first.
Marketing When You're One Person With No Budget
Own one channel completely before touching a second. For most small companies the honest options are direct outreach, one narrow community, or publishing something genuinely useful — and only one of the three fits your temperament. Commit to it for ninety days with weekly volume you can actually sustain, and judge it on conversations started, not impressions.
How to Make Decisions When Nobody Is Coming to Help
Sort every decision into reversible or irreversible. Reversible decisions should be made in minutes with 60% of the information — the cost of deliberating exceeds the cost of being wrong. Irreversible ones (co-founders, equity, long leases, key hires, pricing architecture) deserve a written argument, a devil's advocate, and a night's sleep.
How to Run a Company Without Keeping It in Your Head
Replace memory with rhythm. One fifteen-minute weekly review covering cash, pipeline, customers and the one thing slipping; one monthly hour to review pricing, churn and priorities; one quarterly half-day to decide what to stop. Small repeated reviews outperform heroics and they are what makes a company survivable — and eventually sellable.
Ownership: Keeping What You Build
Ownership is a hygiene practice, not an event. Every asset in the entity, every contributor on a written IP assignment, every co-founder on a four-year vest with a one-year cliff, clean books from month one. Do it early and it costs hours; do it late and it costs equity or the sale itself.
How to Start a Business in 2026
The mechanics are cheaper and faster than ever — you can be selling within a fortnight — but the scarce thing has moved. Building is no longer the bottleneck; being believed is. In 2026 the winning sequence is: pick a specific buyer, sell before you build, assemble rather than construct, and spend your real effort on distribution and proof.
How to Start a Business in 2027
Three shifts are already visible: buyers will assume AI is in everything and stop paying for it as a feature; distribution will keep getting harder as generic content saturates; and trust, provenance and human accountability will become the premium. Build for a market where the product is assumed competent and the company is the differentiator.
AI Business vs Traditional Business
There is no longer a meaningful category called 'an AI business' — there are businesses where AI changes the unit economics and businesses where it is a tool in the back office. Ask one question: does AI let me deliver an outcome that was previously impossible or ten times too expensive? If yes, it is a strategy. If no, use it quietly and compete on everything else.
What AI Changes — and What It Does Not
AI reliably compresses production: drafting, coding, research, design, analysis, admin. It does not compress judgement, trust, distribution or taste. Use it to remove weeks from the making, and expect it to remove nothing from the selling — treat any output as a first draft by a fast, confident, occasionally wrong junior.
The Founder Tool Stack in 2026
Eight categories cover almost every early-stage business: a site, a way to take money, a place customers reach you, a customer record, accounting, a document store, one build tool, and one analytics view. Pick the cheapest credible option in each, resist the ninth category, and re-evaluate once a year — not once a month.
When to Use No-Code — and When Not To
Use no-code whenever the constraint is your speed to first customer, which is nearly always at the start. Move off it when a paying customer's complaint can only be fixed with real code, when per-user costs stop making sense at your scale, or when the data model has outgrown what the platform can express. Not before.
What to Use If You Are Not Technical
Further than a technical founder could have got five years ago, and far enough to reach a real business. Your genuine constraints are debugging, security and scale — not building. The correct strategy is to build the first version yourself, sell it hard, and buy technical help only once revenue justifies it and you can specify exactly what you need.
The One-Person Business Playbook
Yes, if you design for it from the start rather than backing into it. A one-person business needs high prices, narrow scope, no bespoke work, a low-support product, and one distribution channel you can sustain alone. The failure mode is not lack of ambition — it is accepting work that only you can do, at prices that require you to do a lot of it.
How Entrepreneurs Built Before Software
They did exactly what works now, slower: found a group of people with an unmet need, took a personal risk to serve it, delivered by hand, and reinvested the proceeds. Every era's founders had the same four constraints — capital, trust, distribution, and delivery — and the only thing technology has changed is the price of the last one.
The Four Founder Archetypes — and Which One You Are
Founders cluster into four durable archetypes: the Merchant who trades and connects, the Builder who makes the thing, the Operator who runs the machine, and the Creator whose taste is the product. Each wins in different businesses and fails in predictable ways. Knowing yours tells you what model to choose and what to hire against first.
Timeless Founder Lessons That Still Win
Nine things survive every technology shift: proximity to the customer, focus, cash discipline, reputation, pricing courage, distribution ownership, patience with compounding, honesty about what is not working, and staying alive long enough to get lucky. Everything else in founder advice is a temporary tactic wearing a principle's clothes.
Founder Turning Points: What the Big Journeys Teach
Almost every major founder story turns on the same handful of moments: a near-death cash crisis, a painful narrowing of focus, a decision to own distribution, a betrayal or exit that clarified ownership, and a long unglamorous period nobody talks about afterwards. The lesson is not the genius. It is that the turning points were survivable, and survival was mostly a choice about cost and focus.
Why Some Businesses Last and Others Break
Durability comes from five structural things, none of which are growth: pricing power, low fixed costs, a customer base nobody else owns, an owner who stays close to the work, and no single point of failure. Companies break in the opposite pattern — commodity pricing, heavy fixed costs, rented customers, an absent owner, and one concentration risk they never addressed.
Your first 10 customers
They come from people you can name, contacted one at a time, with an offer specific enough to feel written for them. Not from a channel, not from a launch, not from an audience. At ten customers you are still doing manual work that will not scale, and that is correct — the point of the first ten is to learn what makes someone say yes, cheaply enough that you can change the answer.
What to do when growth stalls
A stall is always located somewhere specific: fewer people arriving, the same people converting worse, buyers spending less, or customers leaving faster. Find which of those four moved before you change anything. Most stalled companies fix the loudest thing rather than the moved thing, and lose a quarter doing it.
When speed matters more than efficiency
Speed wins when the cost of being wrong is small and the cost of being late is large — a contested market, a closing window, an unproven assumption. Efficiency wins when the decision is expensive to reverse, or when you are compounding something that already works. The failure is not choosing one; it is applying the same setting to every decision.
How to run a weekly founder review
Sixty minutes, same slot, same six questions, written down. Money in, money out, what moved, what is stuck and who owns it, what I promised and did not do, and the one thing next week is for. The value is not the hour — it is that decisions stop depending on how you happen to feel that day.
How to decide what not to build
Rank by the evidence behind the request, not by how loudly it was made. Anything with no named buyer, no observed behaviour and no revenue attached goes to a not-now list with the reason written next to it. A backlog is a record of unmade decisions; the useful artefact is the list you refuse.
Cash collection before it becomes a crisis
Collection is a system you set up before the invoice exists: terms agreed in writing, an invoice sent the day the work is accepted, a fixed reminder sequence, and one person who checks the ageing weekly. Companies rarely die from bad pricing. They die from good revenue that arrives too late.
How to validate before scaling
Validation is not enthusiasm, it is repeatability: the same kind of buyer, buying for the same reason, at a price that leaves margin, through a route you could run again next month. Until those four repeat, spending on scale multiplies a guess. The test is whether you can predict the next sale, not whether the last one happened.
Looking for the long-form volumes and interactive workbooks instead?
Go to the Library