Before Companies House: The Kitchen-Table Test for a UK Business Idea

It is tempting to start a UK business by registering it. A company name, a logo, a business bank account, a website — it all feels like progress. But registration is admin. The real question is whether the idea works when you put it on a kitchen table and look at it honestly.

This is a one-page test. No spreadsheet, no 40-slide pitch deck, no pretending you enjoy forecasting. Just five boxes, a pencil and a willingness to be slightly disappointed before you spend money.

Why one page is enough at the start

You are not trying to impress a bank. You are trying to find out whether the idea is viable enough to test in the real world. A single page forces you to be clear. If you cannot explain who pays and why on one side of A4, a longer plan will not fix that — it will just hide it.

Use a sheet of paper or a blank document. Draw five boxes. Fill them in with a pencil so you can change your mind.

Box 1: Who actually pays?

Not 'everyone'. Not 'small businesses'. Not 'people who like nice things'. Name the customer in one sentence.

For example: 'Busy parents in my town who need school uniforms altered within 48 hours.' Or: 'Small building firms that need help quoting jobs and chasing invoices.'

Ask yourself:

• Who has the problem badly enough to pay for it?
• Where do they currently spend money on this?
• Who signs off the purchase?
• Can you reach them without a huge marketing budget?

If you have more than one customer type, pick the primary one for now. You can add others later.

Box 2: What does it cost to deliver one sale?

List the direct costs of doing the work once. For a product, that might be stock, packaging, postage and payment fees. For a service, it might be travel, materials, software, subcontractors and your own time.

Add a small buffer — 10% is a reasonable starting point — for the things you have forgotten. You will have forgotten something.

Be honest about your time. If you plan to charge £20 an hour but the job takes three hours plus an hour of admin, the real rate is lower. That is fine for a test, but you need to see it.

Box 3: What should you charge?

Look at three competitors or substitutes. Not to copy them, but to understand what the market expects. Then set one clear price for your test offer.

Do not price at the bottom just to win the first sale. A cheap price attracts customers who complain about the price. Your price needs to cover the direct cost, contribute to your fixed costs and leave something for you.

If you are not VAT registered, you still need to know the VAT threshold because growth can change the maths. But for this page, keep it simple: what will one customer pay you?

Box 4: How many sales do you need to break even?

Break-even is the number of sales that covers your fixed costs. Fixed costs are the bills you pay even in a quiet month: insurance, software subscriptions, storage, website hosting, travel to a regular market, accountancy fees.

The sum is:

Break-even sales = fixed costs ÷ (price − direct cost per sale)

Say your fixed costs are £200 a month. You charge £45. The direct cost of each sale is £15. That leaves £30 per sale to cover fixed costs. £200 ÷ £30 = 6.7. So you need seven sales a month just to break even.

Now ask the useful question: is seven sales a month realistic in your first three months? If not, either the price is too low, the costs are too high or the customer is not ready to buy. Better to find that out now.

Box 5: What are the first three actions?

This is where most plans go wrong. The first actions should not be 'register the company', 'buy a logo' or 'build a website'. Those are admin. Your first actions should test demand.

Pick three actions you can complete in the next 14 days:

1. Talk to five potential customers and ask what they currently do about the problem.
2. Create a one-page offer with a clear price and send it to ten people.
3. Try a small, low-cost test: a market stall, a local advert, a pre-sale or a free taster session.

Write the date next to each action. If you cannot name a date, it is a wish, not an action.

A worked example

Imagine a mobile bike repair service. The customer is a commuter who does not want to drop a bike at a shop. The price for a basic service is £45. Direct costs — parts, travel, payment fees — are £15. Fixed costs — insurance, tools replacement, phone, software — are £200 a month.

Break-even is £200 ÷ £30 = 6.7, so seven services a month. First actions: speak to ten cyclists at the local station, post a simple offer in two community groups, and book a Saturday slot at a local market. That is a test, not a registration exercise.

Red flags on your page

• Your customer is 'everyone'.
• Your price is lower than your direct cost.
• Your break-even number is unrealistic for your area or schedule.
• Your first three actions are all admin.
• You have no evidence that anyone will pay.

None of these mean the idea is dead. They mean you need to change one thing and test again. Raise the price. Narrow the customer. Cut a cost. Run a smaller experiment.

After the page passes

Once the page holds together, then you can decide whether to trade as a sole trader or set up a limited company. You can look at registration, record-keeping and tax with a clearer head. But do the kitchen-table test first. It is cheaper than a logo, faster than a website and far more useful than another evening spent choosing a business name.