MVP testing is how you find out whether the market wants what you have built before you spend the rest of the runway. This walkthrough takes a UK founder from hypothesis to pivot-or-kill in five concrete steps, with the methods, sample sizes, costs, and regulatory checks for each one.
Masum Shamjad
Founder & CEO
May 22, 2026
MVP testing is the process of putting your minimum viable product in front of real users, capturing the right signals, and using those signals to decide whether to scale, pivot, or stop. The outcome of this walkthrough is a decision you can take to your board with evidence behind it.
This guide assumes you have an MVP already built or close to it. If you are still scoping, read the MVP development guide first.
You will need three things to start: a written hypothesis the test will confirm or reject, a small budget for user recruitment (£300 to £3,000 depending on method), and a way to capture both numeric and verbal feedback. Everything else is detail.
Five prerequisites belong in place before the test begins. Missing any of them turns the test into a vanity exercise.
A written hypothesis in the form "we believe X type of user will do Y when shown Z, and that proves W." If the hypothesis cannot be written as a falsifiable statement, the MVP test cannot have a clear outcome.
Pass and fail thresholds stated in advance. Decide before the test what conversion rate, completion rate, or qualitative signal will count as a pass.
A user recruitment plan with budget, panel, and screening criteria attached. Most UK MVP testing budgets run £300 to £3,000 for the recruitment line alone, depending on whether you use a panel platform or recruit through a community.
A data capture stack that records both behavioural events (clicks, completions, drop-offs) and qualitative feedback (interview notes, survey responses, session recordings).
Regulatory cover if the MVP touches personal data, money, or health information. UK GDPR applies the moment the first test user logs in, and a Data Protection Impact Assessment is required at any meaningful risk level (ICO).
With the prerequisites locked, the first practical step is to write the hypothesis the MVP test must answer.
Most MVP testing fails at this step. The team builds the test before naming the question.
A clear hypothesis names three things: the user segment, the behaviour you expect, and the threshold that counts as a pass. "Twenty percent of UK SME founders who visit the landing page will start a trial within seven days" is a hypothesis you can test. "Founders will like the product" is not.
Write down the hypothesis. Show it to a colleague. If they cannot tell you what data point would falsify it, rewrite it until they can.
Use the format: "We believe [target user] will [behaviour] when shown [our solution], and a [metric threshold] proves it." This forces specificity on every line.
Examples we have used in real UK projects: "We believe operations managers at 50 to 250-employee logistics firms will book a demo within five minutes of landing, and a 4% landing-to-demo conversion proves intent." Or: "We believe NHS dietitians will complete a full diet plan in under eight minutes using the MVP, and 70% completion across a panel of 12 proves usability."
Tie every test that follows back to this hypothesis. A test that answers a question the hypothesis did not ask is a distraction.
The hypothesis tells you what to measure. The next step is choosing the testing method that measures it cheapest.
There is no single best MVP testing method. There is only the right method for the hypothesis you wrote in step one.
The methods split into three fidelity tiers. Low-fidelity tests are fast and cheap, and answer demand questions.
Mid-fidelity tests answer service and workflow questions. High-fidelity tests answer behaviour and conversion questions at scale.
Landing page test: a single page describing the product with a sign-up or pre-order call to action, driven by paid traffic or organic posts. Measures demand before any product is built. UK cost: £500 to £2,000 in ad spend plus a half-day to design the page.
Smoke test: a landing page that captures sign-ups but reveals after submission that the product is not yet live. Measures intent at scale before committing to development.
Fake door test: a real product surface with a feature button that, when clicked, surfaces a "coming soon" message and a sign-up. Measures appetite for a single feature without building it.
Explainer video MVP: a short video showing the product working, used the way Dropbox famously did before writing the sync engine. Measures whether the value proposition lands.
Paper prototype: hand-drawn or low-fidelity sketches walked through with a target user. Measures whether the core flow makes sense at all.
Concierge MVP: you manually deliver the service to a handful of customers using whatever tools fit, before automating anything. Measures whether the service itself is wanted and what the manual workflow actually looks like.
Wizard of Oz MVP: the customer sees an automated product, but a human is doing the work behind the curtain. Measures whether automation would be valued before the automation is built.
Single-feature MVP: a real product with one feature implemented properly and nothing else. Measures whether the single feature is enough to convert paying users.
Piecemeal MVP: a product stitched together from existing third-party tools (Zapier, Airtable, Stripe, Webflow). Measures the full workflow before any custom code is written.
Beta release: the real MVP, on a real domain, opened to a closed cohort of paying or pre-registered users. Measures every signal at once at the cost of a real build.
Usability testing with a high-fidelity prototype: a clickable Figma flow tested with target users either in person or remotely. Measures task completion and confidence on the journeys that matter most.
A/B testing: two versions of the same MVP screen shown to randomised user groups. Measures which variant performs better, but only at traffic volumes high enough to be statistically meaningful.
Match method fidelity to hypothesis type. Demand and pricing questions belong in low-fidelity tests.
Workflow and trust questions belong in mid-fidelity tests. Conversion and retention questions belong in high-fidelity tests.
A common mistake is to skip straight to high-fidelity when a £500 smoke test would have answered the question. Method choice sets the whole budget.
The method tells you what you are building. The next step is who will use it.
Test users are the single most decisive input into MVP testing. Five well-recruited testers outperform fifty random ones.
The first decision is whether you need many users or the right users. For usability questions, the answer is the right users in small numbers. For demand questions, the answer is many users from the target segment.
The Nielsen Norman Group research on usability testing finds that five users surface roughly 85% of all usability problems, and additional testers return diminishing insight per session. Three rounds of five testers outperform one round of fifteen, because each round can incorporate the previous round's fixes.
Use this rule for any test focused on whether the product is usable, learnable, or trustworthy. Use a different rule for demand testing, where conversion rates require larger samples.
Prolific: UK-based research panel with a £6 per hour platform minimum and roughly £9 per hour recommended UK rate (Prolific). Suitable for consumer products and general population testing.
Respondent: specialist B2B panel where session rates run £60 to £150 depending on seniority. Suitable for testing products aimed at managers, directors and specific job roles.
UserTesting: subscription platform with a managed panel and built-in session recording. Suitable when speed and managed logistics matter more than panel customisation.
Your own waiting list: people who have already signed up to hear about the product. Suitable when the waiting list is large enough and the product addresses a clear pain.
LinkedIn outreach: direct messages to people in the target role, offering a £40 to £80 voucher for a 45-minute session. Suitable when the target user is hard to reach through panels.
A screener with three to five questions filters out unsuitable participants before they take a session slot. Screen on the specific behaviour your hypothesis names, not on demographics alone.
For a fintech MVP aimed at finance directors, screen on "have you reviewed an expense in the last 30 days" rather than job title. People over-claim role match; behaviour-based screening cuts the noise.
With the right testers booked, the test itself starts. The next step is what to capture while it runs.
Running an MVP test well comes down to capturing both numbers and words. Numbers tell you what happened. Words tell you why.
Every MVP test should capture a behaviour stack and a feedback stack in parallel.
Activation rate: the percentage of testers who complete the core "aha moment" task at least once. The most important MVP metric, since it predicts retention.
Time-to-value: the median time from first login to the first value-delivering action. Long times signal a usability or onboarding gap.
Drop-off points: the exact step in the workflow where the largest share of users abandon. A funnel chart of three to five key steps surfaces this quickly.
Conversion rate: the percentage of test users who take the paid action (subscribe, purchase, book a call). For paid MVPs only; pre-revenue tests use a proxy such as sign-up or demo booking.
Net Promoter Score or equivalent: a one-question signal of recommendation intent. Useful as a comparator across rounds, not as an absolute number.
Watch where the tester hesitates. Hesitation in front of a button signals confusion that the click event will not surface.
Note the exact words the tester uses to describe the product back to you. If their language differs from your marketing language, the marketing language is wrong.
Listen for unprompted comparisons to other products. They reveal the alternatives the tester is already using, and the gap your MVP needs to fill.
Record sessions with consent and pull three short clips that capture the most important moments to share with the team and investors. A 20-second clip of a real user struggling beats a slide of bar charts.
The data tells you what happened. The next step is deciding what to do about it.
The final step of any MVP test is a binary call dressed up as a three-way one. Either the test passes (persevere), the test fails on the hypothesis but signals an adjacent opportunity (pivot), or the test fails outright (kill).
Most UK founders we work with delay this call. They want one more round, or to retest a different segment.
CB Insights data on startup post-mortems finds that 35% of failed startups cite no market need as the primary cause (CB Insights, 2021). Another round of MVP testing rarely changes that signal once it is clear.
Persevere when the activation rate, conversion rate, or qualitative confidence beats your pre-stated threshold. Scale the same test with the same hypothesis.
Pivot when the test fails on the hypothesis but a recurring qualitative signal points to a different user, problem, or solution. Rewrite the hypothesis around the new signal, then test again.
Kill when neither the quantitative nor the qualitative signal supports the hypothesis or any adjacent one. Killing is not failure; it is the cheapest possible answer to the question "should we keep going?"
A clear decision at this stage is what the entire MVP testing process was designed to produce. The next two sections cover regulation and common mistakes that affect every stage of the test.
UK regulation does not go on holiday during an MVP test. Three regulatory bodies shape what you can and cannot do once real users are involved.
UK GDPR applies the moment personal data is processed (ICO). For most MVP tests, that moment is the first sign-up form.
The Information Commissioner's Office requires a Data Protection Impact Assessment when processing carries any meaningful risk to test users. The DPIA names the lawful basis (usually legitimate interest or consent), the data minimisation steps taken, the retention period, and the deletion process at the end of the test.
Session recording tools (UXCam, FullStory, Hotjar) need particular care because they capture behaviour and screen content. Mask any free-text input fields and anonymise the recordings, or you carry the same risk as a full customer database.
A fintech MVP that touches regulated activities (payments, lending, savings, insurance distribution) needs FCA authorisation or a sanctioned route. The FCA Regulatory Sandbox is the sanctioned route for live MVP testing.
Sandbox cohorts open periodically and accept firms with a clear consumer benefit, a real innovation, and a need for live testing. Acceptance gives you restricted authorisation to test with real customers under FCA supervision for a defined window.
If your MVP processes payments, do not run a live test outside the Sandbox without confirming your authorisation status with a compliance specialist first. The fines for unauthorised regulated activity start at £50,000 and scale rapidly with consumer harm.
A healthtech MVP intended for NHS use must pass the Digital Technology Assessment Criteria before any clinical pilot. The DTAC covers clinical safety, data protection, technical security, interoperability, and usability.
Plan the DTAC sign-off into the MVP testing timeline from the start. Healthtech founders who skip it at MVP stage rebuild large parts of the product to meet it later.
Regulation shapes what an MVP test can do. The next section covers the practical mistakes that derail tests even when regulation is handled correctly.
Most MVP testing failures we see fall into the same six patterns. Each one is preventable.
Testing the product, not the hypothesis. The team launches a beta and watches metrics drift without a written question to answer. The result is data without decisions.
Recruiting friendly users. Friends, investors, and existing customers tell you what they want you to hear. Test with strangers who match the target user behaviour profile.
Running an A/B test below statistical significance. Most UK MVPs do not have the traffic to call an A/B test in a sensible window. If the expected weekly visitors are below 1,000, use sequential testing or skip the A/B entirely.
Ignoring qualitative signal because it is hard to measure. A dashboard of clicks will not tell you the product is confusing. A 15-minute interview will.
Pivoting on one round of data. A single test result is a hypothesis, not a verdict. Confirm the signal in a second round before committing to a pivot.
Killing the test before it has finished running. Cutting a test at the first negative day burns the recruitment budget and produces no answer. Stick to the pre-stated stop conditions.
These mistakes appear in every sector we have worked across. The next section is what to do once the test has produced a clear result.
A passed MVP test is the start of the next decision, not the end. Three things should happen in the week after the result.
Document the test in full. Write down the hypothesis, the method, the testers, the data, and the decision. The next test reuses the structure; the next funding round references the rigour.
Schedule the next test. A persevere decision needs a scale test. A pivot decision needs a fresh hypothesis test. A kill decision needs a debrief and a new project.
Review the regulatory file. If the MVP test triggered a DPIA, update it before any production data is collected. If it ran in the FCA Sandbox, prepare the exit report before the cohort ends.
MVP testing is not a project you do once. It is the discipline that lets you make every product decision with evidence behind it.
The five steps in this walkthrough have taken you from a written hypothesis to a clear decision, with the right testers, the right data, and the right regulatory cover. If the steps feel heavy for an MVP, that is the point. Cheap tests done well save the cost of expensive products built on guesses.
If you want a second pair of eyes on the MVP test you are planning, our software development team runs no-cost scoping sessions to map the hypothesis, method, and sample size for any UK MVP. We have run MVP tests across UK fintech, healthtech, retail and logistics for 14 years.
Tell us the question, and we will tell you how to test it.
MVP testing is the process of putting a minimum viable product in front of real users to find out whether the product solves a problem worth paying for. It works by stating a hypothesis, choosing a method that fits the hypothesis, recruiting the right testers, capturing both numeric and qualitative data, and using the result to persevere, pivot, or kill.
For usability questions, five well-recruited testers surface roughly 85% of issues (Nielsen Norman Group). For demand and conversion questions, you need a sample large enough to be statistically meaningful, which usually means several hundred to several thousand impressions through a landing page or paid ad campaign.
A focused usability test runs across one to two weeks with three rounds of five testers. A landing page or smoke test runs for two to four weeks to gather enough traffic. A live beta release typically runs for four to eight weeks with a closed cohort.
UK MVP testing budgets typically run £1,500 to £8,000 for a single round of structured testing. Costs include user recruitment (£300 to £3,000), session moderation, analytics tooling, and any paid traffic for landing-page tests. Beta releases of fully built MVPs cost more because the build itself is included.
MVP testing answers whether the product idea is worth pursuing. Beta testing answers whether the built product is ready for general release. MVP testing happens before or during the early build; beta testing happens after the build is feature-complete but before launch.
Contact Us
Get in touch with our team anytime today.
Our team is always here to listen, support, and guide you.
Stay informed with instant updates delivered straight to your inbox.
Providing top-notch technology solutions that you can trust to transform your growing business
Services
Get In Touch
Ahmedabad
714, Shivalik Shilp, Iscon Cross Road, SG Highway, Satellite, Ahmedabad – 380015, Gujarat, India.
+91 63532 74556
