You validate a business idea by getting strangers to give up something of value for it before it exists. Money is best. A booked call, a signed letter of intent or a place on a waiting list comes next. Compliments, survey answers and encouragement from friends are not evidence, because they cost the person nothing.
You can do this without building the product. Customer conversations, a landing page with a real call to action, a small advertising test, pre-orders and a hand-run “concierge” version will tell you most of what you need in a few weeks.
- Test the riskiest assumption first, which is nearly always “will anyone pay for this?” and rarely “can it be built?”.
- Count commitments, not opinions: payments, deposits, bookings and sign-ups from people you don’t know.
- Set your pass mark before each test and stop, or change the idea, if you miss it twice.
- In the UK, check the name at Companies House and on the trade mark register, and know that taking pre-orders from consumers brings legal duties.
Why validate first: what the survival figures show
Starting a business in the UK is easy. Companies House recorded 193,789 company incorporations between July and September 2026 alone, though 226,091 companies were struck off in the same quarter. Staying in business is harder. According to the Office for National Statistics (ONS), of the UK businesses born in 2020, 92.9% survived their first year, 53.0% survived three years and 37.6% were still going after five, in 2025. The ONS counted around 326,000 business births and 271,000 business deaths in 2025 (the death count is provisional). The death rate of 9.4% was the lowest since 2006.
Published 19 November 2026. Source: ONS, Business demography, UK: 2025, Table 5.1a Source: Companies House, 29 October 2026View data
UK businesses born in 2020: share surviving each year 1 year 92.9% 2 years 71.3% 3 years 53% 4 years 44% 5 years 37.6%
The ONS figures don’t say why businesses close, and a closure is not always a failure. But the shape is clear: almost everyone gets through the first year, and the losses come in years two to five. Checking demand early is the cheapest protection you have.
Step 1 of 5
Interviews
Ask 10 to 15 people with the problem what they did last time, not what they would do.
Step 2 of 5
Landing page
State the offer and price, with one action that costs the visitor something.
Step 3 of 5
Ad test
Send a fixed budget of paid traffic to the page and count committed visitors.
Step 4 of 5
Pre-orders
Take payment or a refundable deposit, and follow consumer law when you do.
Step 5 of 5
Concierge
Deliver the service by hand to a few paying customers before building it.
1. Talk to potential customers, and ask about the past
Have 10 to 15 conversations with people who have the problem you want to solve. Don’t pitch. Ask what they did the last time the problem came up, what it cost them, what they’ve tried and what they pay for now. People are unreliable about what they would do and accurate about what they did.
Listen for evidence of effort: a spreadsheet they built, a tool they pay for and dislike, a member of staff who spends Fridays on it. If nobody has tried to solve the problem, it may not be painful enough to pay for. End each conversation by asking for something small, such as an introduction or a second call to see a mock-up. Whether they agree is a data point.
2. Put up a landing page with a real call to action
A single page that states the offer, the price and who it’s for, with one action to take. The action must cost the visitor something: a deposit, a booked slot in your calendar, or at least an email address given in exchange for a specific promise. A “notify me” button with no price on the page measures curiosity.
Be honest that the product is not available yet. Say when you expect it to be. If you collect email addresses, the marketing rules apply from the first message you send, so read our summary of the UK email and SMS marketing rules before you build the list.
3. Run a small ad test
Paid traffic gives you strangers quickly, which is the point. Friends and followers are too kind. Set a fixed budget you can afford to lose, send the clicks to your landing page and measure what share of visitors take the action. Search ads suit ideas people already look for. Social ads suit ideas people don’t know to search for. We compare the two in Meta ads or Google Ads: where to start, and how much Google Ads cost in the UK covers budgets.
Don’t read much into a handful of clicks. The useful outputs are the cost of getting one committed person and the words in the adverts that drew them, which tell you how customers describe the problem.
4. Take pre-orders, and follow consumer law when you do
Nothing validates like payment. A pre-order or refundable deposit tells you the price is acceptable as well as the idea. It also makes you a trader with obligations.
If you sell to consumers online, the Consumer Contracts (Information, Cancellation and Additional Charges) Regulations 2013 apply. GOV.UK’s guidance for businesses says you must give customers certain information before they order, including your business name and address, a description of the goods or service, the total price with taxes, delivery costs and timescales, and how to cancel. Three points matter most for pre-orders:
- Delivery date. You must deliver goods within 30 days unless you agreed a different timescale with the customer. A pre-order needs a clearly stated date that the customer accepts.
- Right to cancel. Customers can cancel up to 14 days after their order is delivered, without giving a reason. If you fail to tell them about this right, they can cancel at any time in the next 12 months.
- Refunds. You must refund within 14 days of getting the goods back. Some items, such as personalised or custom-made goods, are exceptions.
Our guide to online returns and UK consumer law goes through this in detail. Keep pre-order money separate until you deliver, and refund promptly if you decide not to go ahead. This is general information, not legal advice.
5. Deliver it by hand first
A concierge test means doing manually what the product will later do automatically. Say you want to build an app that plans weekly menus for families with allergies. Before writing code, find five families and plan their menus yourself by email for a month, at the price you intend to charge. You learn what customers value, what they ignore and whether they renew, and you get paid while learning. It doesn’t scale, and it isn’t meant to.
What counts as evidence?
| Weak evidence | Strong evidence |
|---|---|
| “I’d definitely use that” | A deposit or pre-order from a stranger |
| Survey answers about willingness to pay | A signed letter of intent from a business buyer |
| Likes, followers and page views | Booked calls that people turn up to |
| Friends and family signing up | Concierge customers who pay a second time |
When to stop or change direction
Write down a pass mark before each test: for instance, five paid pre-orders from a fixed ad budget, or three of fifteen interviewees agreeing to a paid pilot. The numbers are yours to set, and the discipline is in setting them first. If you miss the mark, change one thing (the audience, the price or the offer) and run it again. If two honest attempts fail, the kindest reading is that the demand isn’t there in this form. Stopping at that point is a good result. You have spent weeks, not years.
If you pass, you have something more useful than an idea: named customers, a tested price and a known cost of reaching buyers. That is also what investors want to see, and our overview of SEIS, EIS and Innovate UK funding explains the routes once you get there.
UK checks: the company name and the trade mark register
Before you print anything, check the name. Companies House has an online company name availability checker, and it will not register a name that is the “same as” an existing one. A different spelling is not always enough: GOV.UK warns that you may have to change your name if someone complains and Companies House agrees it is “too like” an earlier one, or if it is too similar to another company’s trade mark.
So search the Intellectual Property Office’s trade mark register as well, by keyword and phrase. Registering your own mark costs at least £205, usually takes around three to four months if there are no objections and lasts 10 years. Check that a suitable web address is free at the same time. If the .uk you want is taken, our list of premium .uk domain names may help.
Common questions
How long should validation take?
We usually suggest four to six weeks of focused testing. Longer than that and you are probably avoiding the pre-order step, which is the one that gives a clear answer.
What if someone copies the idea while I’m testing it?
It’s a smaller risk than building something nobody wants. Ideas are rarely the scarce part, and you can’t learn what customers will pay for without telling them what you plan to sell.
If you’d like help designing the tests, that is part of our startup consulting work. If the tests point to software, our guides to scoping a SaaS MVP and what an app costs to build in the UK are the next step.
Sources
- ONS: Business demography, UK: 2025 (19 November 2026)
- ONS: Business demography, UK, reference tables (19 November 2026)
- Companies House: Incorporated companies in the UK, July to September 2026 (29 October 2026)
- GOV.UK: Online and distance selling for businesses (accessed 18 November 2026)
- GOV.UK: Accepting returns and giving refunds (accessed 18 November 2026)
- legislation.gov.uk: Consumer Contracts (Information, Cancellation and Additional Charges) Regulations 2013, regulation 30 (2013)
- GOV.UK: Set up a limited company, choose a name (accessed 18 November 2026)
- Intellectual Property Office: Search for a trade mark (accessed 18 November 2026)
- Intellectual Property Office: Apply to register a trade mark (accessed 18 November 2026)