How to start a company from zero

Should you start a company?

Most people who could start a company never do. Not because the idea was bad. Because doing all of it looked impossible: building it, getting found, selling, support, the paperwork. That list used to need a team and a budget. It does not anymore.

So the question is whether you have something worth starting. You probably do if any of these are true:

• You have watched the same problem happen over and over at work or in your life, and the fixes people use are clumsy.
• You know a group of people better than most: a trade, a hobby, a neighbourhood, a condition, a craft.
• You keep coming back to the same idea, even after you have tried to drop it.
• Something just became possible with AI that was not possible two years ago, and you can see who would want it.

If one of those fits, you are already a founder. The rest is doing, and doing is the part doany takes off your hands.

Find an idea worth building

Good ideas rarely arrive as inventions. They arrive as complaints. Someone is paying for something clumsy, doing something by hand, or putting up with a bad option because there is no better one. That is the signal.

Three places to look

• Problems you already understand. The job you have, the job you had, the thing you do on weekends. You know the words people use, what they have tried, and what they would pay to make it go away.
• Things you personally want. If you would buy it tomorrow, you are the first customer and the best judge of whether the version you get is any good.
• Groups you belong to. Every community has a recurring “does anyone know a tool for…” thread. Read those threads. The answer “nothing, sorry” is your idea.

Skip anything that starts with the technology. “An AI for X” is not an idea until you can finish the sentence “so that a specific person can…”.

Test the idea before you build it

Building first and asking later is how most ideas die quietly. Test first. It is cheap, it is fast, and it changes what you build.

Four questions settle most ideas:

• Who has this problem, exactly? Not “small businesses.” A person with a job, a habit, or a situation.
• What do they do about it today? Spreadsheets, a freelancer, a worse tool, nothing. Workarounds are proof that people care.
• Would they pay? Ask ten of them. A “maybe” is a no. Two real yeses beat fifty likes.
• Why now? What changed so that this is possible or urgent today?

Pick your first customer, not your whole market

Start narrow. It feels backwards when you are dreaming big, but the fastest way to a big company is a small group of people who need you badly.

A good first customer is easy to describe, easy to find, feels the problem often, and has money to fix it. When you can say who they are in one sentence, you know where to show up, which words to use, and what to leave out of version one.

You will widen later. doany can point the whole company at a new audience when you are ready. For now, one kind of person.

Name it and claim the domain

A name has one job: do not get in the way. Simple, easy to spell after hearing it once, and roomy enough to survive a change of direction. You do not need a perfect name. You need one you will not have to explain.

• Say it out loud to someone and ask them to spell it.
• Avoid names that only make sense after a story.
• Check that nobody with a similar name does a similar thing.
• Check that the domain and the social handles are free. A .com is nice, not required. .co, .ai, .app, and .io are all fine.

Write the one sentence

Everything in doany starts with a sentence. Not a business plan, not a deck. One line that says who it is for, what it does, and what changes for them.

A formula that works: [what it is] for [who, specifically] that [the result they want].

• A weekly meal-plan service for first-time marathon runners that turns what is in their fridge into a training diet.
• An invoice-chasing tool for freelance illustrators that gets late clients to pay without awkward emails.
• A booking site for mobile dog groomers that fills the gaps in their week.

What comes next

You have a problem, a person, a name, and a sentence. Next comes the part that used to need a developer: turning the sentence into a live product and website. That is Chapter II.