Write down what must be true
Before you test anything, list the beliefs your idea depends on. Then mark the ones that would sink the business if they were wrong and that you have the least evidence for. Those are your riskiest assumptions, and they are where testing starts.
Turn each one into a test with a pass mark. For example: “At least 8 of the 15 independent dog groomers we interview describe a missed appointment that cost them money in the last month.” Setting the number in advance stops you reading whatever you hoped for into the results.
Talk to potential users first
Interviews are the cheapest test and often the most useful. The GOV.UK Service Manual, written for government digital teams, recommends learning about user needs by interviewing and observing actual or likely users, and understanding all kinds of users, not just “typical” ones.
Ask about the past, not the future. “Tell me about the last time this happened” gets you facts. “Would you use an app that…?” gets you kindness. Questions that work well:
- When did this last happen? What did you do?
- What did it cost you, in time, money or stress?
- What have you tried already, and what did you pay for it?
- Who else is involved in deciding?
- Can I come back to you when there is something to try?
Test demand with a landing page
A landing page describes the product as if it exists and asks for a real action: joining a waiting list, booking a call or placing a pre-order. Send people to it from the places your users already spend time, and count how many take the action.
Eric Ries, who developed the Lean Startup method, describes spending two weeks building a feature that nobody wanted, and says a simple advertising smoke test would have revealed the problem much sooner.
If you collect email addresses, UK rules apply. You need people’s specific consent before you send them marketing emails, so say clearly what you will send and include an easy way to unsubscribe. The ICO notes that the “soft opt-in” for existing customers does not apply to prospective customers or new contacts.
Show people a prototype
A prototype lets people try the product before it exists. It can be paper sketches or a set of clickable screens. GOV.UK’s guidance notes that sketches are useful for exploring and discussing ideas, while more realistic prototypes are better for user research.
Give each person a task, such as “book a groom for next Tuesday”, and watch without helping. Where they hesitate or get lost is where the design needs work. Ask what they expected to happen, not whether they like it.
Deliver it by hand, then ask for money
A concierge test means delivering the service yourself to a handful of customers, with no product behind it. It feels slow, but you learn exactly what people value, and it costs almost nothing in development.
The strongest evidence is money. A deposit, a pre-order or a paid pilot tells you far more than a sign-up. If you take payment from consumers before the product exists, be clear about what they are buying, when it will arrive and how refunds work. Online sales to consumers come with cancellation rights, so read the GOV.UK rules on distance selling first.
Decide what the results mean
Compare each result with the pass mark you set. If a test passes, move on to the next risky assumption. If it fails, find out why before you change the idea: was it the problem, the audience, the message or the price?
Record each experiment in the validation step of the MVP builder, with what it tests and its pass mark, so your brief shows the evidence behind it.
Validation never makes an idea certain. It makes the next decision less of a guess, and it tells you what the MVP itself needs to prove.
Before you finish
- Your riskiest assumptions are written down.
- Each test has a pass mark set before it runs.
- You have interviewed potential users about real past events.
- Your landing page asks for a real action and counts it.
- Email sign-ups are collected with clear consent and an easy way to unsubscribe.
- You have watched people use a prototype without helping them.
- You have asked for money, or a firm commitment, before building.
- You have recorded what each result means for the idea.