Test your business idea for two hundred pounds or less

Validation isn't research. It's a stranger giving you money, or a firm commitment to give you money, before the thing exists.

Everything else — surveys, focus groups, your brother-in-law saying it's a great idea — is noise dressed up as data. The test you're running has one question at its center: will someone who doesn't love me pay the price I need to charge?

Two hundred pounds is plenty to answer that. Most of it you won't spend.

Write the offer down in one sentence

Before you spend anything, write this: I help [specific person] with [specific problem] for [specific price].

If you can't fill in all three blanks, you don't have an idea yet, you have an area of interest. "I help people with marketing" isn't testable. "I help independent dental practices fill cancelled appointment slots, £400 a month" is testable by Friday.

The price matters more than people expect. Free interest is infinite. Paid interest is scarce, and the gap between them is where most businesses die. Put your real number in the sentence, the one that would make the work worth doing, not a discount you've invented to make the conversation easier.

Then write the version where you're wrong. What would have to be true for this to fail? Usually it's one of three things: they don't have the problem badly enough, they already solve it some other way, or they have it and can't pay. Name yours. You're about to go and check.

Talk to ten before you build anything

Not a survey. A conversation, on the phone or in person, fifteen minutes.

Find twenty people who fit your buyer description. LinkedIn, trade associations, local business directories, the Facebook group where they complain to each other. Message all twenty. Roughly half will ignore you. Talk to whoever replies.

Don't pitch. Ask about the last time they dealt with the problem. What did they do? What did it cost them in time or money? What did they try before that, and why did they stop? You're listening for a specific, recent, expensive irritation. Vague agreement that something "would be useful" is worth nothing.

Two questions do most of the work:

• "What are you doing about it now?" If the answer is "nothing", the problem probably isn't painful enough to open a wallet for.
• "Who would have to approve the spend?" If it isn't the person you're talking to, you've been having a pleasant chat with someone who can't buy.

Cost so far: nothing but your time. Most ideas die right here, and that's the cheapest possible place to lose.

Then ask for money

The conversation is a filter. The pre-sale is the test.

Go back to the three or four people who leaned in and offer to do the work. Now. Manually, badly, at a real price. A founding-customer rate is fine — twenty or thirty percent off is a reason to move today, not a signal that your pricing is broken. Anything deeper than that and you've tested a different business.

For a service, take a deposit. For something you'd deliver over a month, take the first month up front. If someone won't put a card down but says "definitely, once you're set up", write them down as a no. Kind words aren't revenue.

Three paying customers out of ten conversations is a strong signal. One is worth continuing with. Zero, with ten real conversations behind it, is your answer, and it arrived for free.

If you're taking deposits, keep it clean: written terms, a clear refund position, money you can actually give back if you don't deliver. Ask an accountant early about how to record it and whether anything you're doing needs registering.

What the money is actually for

You'll barely touch the two hundred. Sensible caps, not a shopping list:

• $15 or so for a domain. Your name works. Clever names slow you down.
• $0 for the website. A single page on a free builder, one screen tall: who it's for, what it does, the price, a button. Nothing else.
• $40 for a modest paid test, and only if you need to reach people you can't find by hand. Tradespeople, local services, anything with a search term behind it — a small spend tells you whether anyone's looking. B2B niches with forty buyers in the country? Don't bother. Email them.
• $30 for coffee, travel and a couple of lunches. Underrated. Sitting in front of someone gets you the truth faster than any advert.
• Hold the rest. You may need it to deliver the first job properly, and having the delivery money is more useful than having spent it on a logo.

What you don't buy: branding, business cards, incorporation, a CRM, an accountant's retainer, stock, a course about starting a business. Every one of those feels like progress and none of them answer the question.

The signals that lie to you

Landing page sign-ups. People will hand over an email address for anything. Treat a list of sign-ups as a list of people you now have to phone.

Compliments from anyone who knows you. Your family wants you to be happy. They're a terrible test panel.

"I'd definitely pay for that." Future tense, no card. Discount it entirely.

Traffic. You can buy attention for a fiver. Attention isn't demand.

And the worst one: your own enthusiasm at eleven o'clock at night. Ideas are seductive in the dark. Write the plan down and read it again at eight the next morning, sober and unexcited.

Be honest about what most of these tests find

Most ideas fail this. Not some — most. Sensible, well-thought-through ideas from capable people fail it routinely, and the failure usually isn't the idea, it's the price, or the buyer, or the fact that a decent-enough alternative already exists and nobody's angry enough to switch.

That's the whole point of doing it for two hundred pounds over three weeks rather than fifteen thousand over a year. You're not trying to prove yourself right. You're trying to find out cheaply, and a fast no is a genuinely good outcome. It hands you back your evenings and your savings.

Give yourself a stopping rule before you start, because you won't be objective later. Something like: twenty approaches, ten conversations, three weeks, and at least one paid deposit. Miss it and you stop, change one variable — the buyer, the problem, or the price — and run it again. Not all three. One. Otherwise you'll never know which change mattered.

If it passes, don't celebrate, deliver

A yes is the start of the harder test: can you do the work at that price without hating your life?

Deliver the first three jobs by hand. No systems, no software, no hiring. Do it yourself, badly and slowly, and time it. Service businesses live or die on whether the hours you actually spend match the price you actually charge, and you can't know that from a spreadsheet. Most people who quit a working business quit because the maths never worked and they only found out at customer twelve.

Get through three. Then raise your price on the fourth and see if the world ends.

It usually doesn't.