
The Ten Questions Every Business Idea Should Survive
Evaluating IdeasSeptember 16, 2026
The Charlie Page Letter
26 years of experience and common sense delivered to your inbox.
A business idea should survive ten questions in one sitting. Not a market study, not a pitch deck, ten questions asked in order, with written answers, and a willingness to stop when one of them can't be answered honestly.
These aren't original questions, which is precisely why they work. They're the questions an investor asks in the last ten minutes of a meeting, the ones a buyer asks in the second phone call, the ones you ask a friend's idea because you have no stake in it. Founders are the only people who skip them.
I run this list on my own work before I let myself enjoy it, and it has cost me several ideas I liked. That's cheaper than the alternative, which is finding out late that the customer was a rumour and the price was a hope.
Who exactly is this for?
One sentence, one person, no plural nouns. "A freelance video editor with four repeat clients and no invoicing system" passes. "Creators" fails, because you can't find creators, but you can find that person, and the difference decides everything downstream.
The test is whether a stranger could locate twenty of them this week. If not, the customer is too vague or too rare, and no amount of product work will fix a person you can't point at.
What's the problem, in their words?
Not your phrasing. Theirs, copied from a forum or a conversation, warts and all. If you can't produce their wording, you haven't listened yet, and you're about to build for an imaginary person with great sincerity.
The problem also has to be one they'd recognise as a problem. Many real difficulties are simply lived with, and a solution to something the buyer considers normal weather is not a product, it's a lecture.
What do they do about it today?
A workaround, a hire, a spreadsheet, a monthly argument, nothing. The answer is your real competitor, and "nothing" is the most dangerous answer on the list, because it means you're not improving a habit, you're installing one.
Past behaviour is the only honest data available before a sale. Opinions about the future are nearly worthless, and the fastest way to lose a week is to conduct interviews where everyone predicts they'd use it.
What do they already pay, and to whom?
A number, and a name. Payment proves the problem is worth something, and the name tells you who to learn from and who to be better than. If nobody pays anything, you have a preference rather than a budget.
Resentful payment is the best kind. Someone paying for a tool they hate, or a service they grumble about, is a buyer who is already convinced and only needs a credible reason to move.
What's my price, said out loud?
A specific number you'd say to a stranger without qualifying it. Ranges are avoidance. If you can't name a price, you don't have an offer yet, and the idea should be treated as unfinished rather than defended.
Then check the price against the alternative the buyer already accepts. Not against your effort, against what they currently spend in money or hours. That comparison is what makes a price feel fair, and it isn't your opinion that decides it.
How do I reach the first hundred?
One channel you can actually work this week: a community you're already in, a list you own, a partner who serves the same customer, or outreach you can do yourself by hand. "Content" and "SEO" are not answers until you've named the words and the room.
If you can't name a channel, you don't have a business, you have a product looking for an audience. That's the single most common critical flaw in weak ideas, and it's invisible from the inside because the product is the fun part.
Does each sale make money?
Take one sale and subtract everything it costs: delivery time, tools, refunds, the hours of finding the next one. If the remainder is thin, growth makes it worse, because volume multiplies a loss rather than a profit.
This is the question new founders find least interesting and should find most urgent. A business that loses money politely on every transaction still loses money, and no amount of scale turns a negative into a positive.
Can I actually deliver it?
Not the demo, the tenth delivery, when the customer is awkward and the edge case shows up. Be honest about the skills and hours required, because a promise you can't keep repeatedly is a refund with extra steps.
Delivery problems are the ones borrowed mechanisms can't solve for you. Whatever the structure looks like on paper, someone has to do the work every week, and that someone is you at the start.
Why would someone switch?
There is always a competitor, including the current workaround. Your reason has to be one the buyer cares about: faster, safer, less embarrassing, better for the specific corner they're in. Cheaper is the weakest reason and the easiest to lose.
If you can't articulate the switch, the honest conclusion is that the market is satisfied. That's information, not failure, and it points you at a different customer rather than a bigger budget.
What evidence would rule this idea out?
Name it before you build. Five strangers who won't pay, a price ceiling you can't clear, a channel that produces nothing. If you can't name anything that would count against the idea, you're not testing it, you're believing it.
This is also the question an unsentimental second opinion handles well. A free pressure test asks these ten in a slightly different order, scores the answers, and tells you which one is critical rather than which one is uncomfortable. Keep your revisions in My Tests so you can watch the weak answer get fixed.
Ten questions, in writing, with the willingness to stop. That's the whole discipline, and it's available to anyone who'd rather be wrong cheaply than right expensively.
Run your idea through the free pressure test to get the same scrutiny in minutes, then read how to tell if anyone is already paying for the question most people can't bring themselves to ask.
Frequently asked questions
The Charlie Page Letter
26 years of experience and common sense delivered to your inbox.