What Is an MVP?
An MVP (Minimum Viable Product) is the simplest version of your product that lets real people use it. You build only the core features — enough to test your idea with real users without spending months or hundreds of thousands of pounds.
No fancy onboarding. No polished animations. No feature that only 1% of users will ever click. Just the thing that solves the problem — shipped fast so you can learn fast.
What does MVP stand for?
Defined Term: MVP
MVP stands for Minimum Viable Product. Each word carries a specific meaning that shapes how you build.
M
Minimum
The smallest possible version. Not the best version. Not the complete version. The version with the fewest features that still works.
V
Viable
It must actually work. A viable product solves a real problem. It is not a mockup or a demo. Real users can sign up and get real value from it.
P
Product
It is a real product — not a spreadsheet, not a Notion doc, not a form. It is something a user can interact with independently, without you holding their hand.
Think of an MVP like an IKEA table — not the most beautiful table in the world, but it is a real table you can eat dinner on tonight. You can always upgrade to a better table once you know you like eating in that room.
Why do startups build MVPs?
Key Stat
90% of startups fail.
34% fail because there was no market need — meaning they built something nobody wanted. CB Insights, 2023. An MVP is the antidote to that. You find out if people want your product before you spend six months building it.
The core reason is simple: you do not know if your idea works until real people try it. No amount of planning, market research, or competitor analysis tells you what actually happens when a real user opens your product for the first time.
Validate fast
Find out in weeks if your idea solves a real problem — not after 12 months of development.
Spend less to learn more
A £10k MVP that fails teaches you more than a £150k product that never ships.
Attract investors
An MVP with 100 real users is infinitely more fundable than a pitch deck with no product.
Iterate (improve step by step) based on real data
User behaviour tells you what to build next. Guessing costs you runway (the time before you run out of money).
Reduce risk
The biggest startup risk is building the wrong thing. MVPs shrink that risk dramatically.
Reach market faster
Every week you are not live is a week a competitor could get there first.
What goes INTO an MVP?
The rule is brutally simple: include only what a user needs to get their first piece of value. Here is what that looks like in practice.
The one core user journey
There is one thing your product does. Map out every step a user takes to do that thing — sign up, enter data, see a result, take an action. Build exactly those steps. Nothing else.
Basic authentication (sign up / log in)
Users need to create an account and get back into it. Use an existing auth library. Do not build this from scratch.
The core feature or workflow
This is the thing your product actually does. It must work reliably. If it is a budget tracker, the user must be able to add a transaction and see their total. That is it.
A way to collect payments (if applicable)
If you are charging, use Stripe. Set up one pricing tier. You can add more later. The goal is to prove people will pay — even one paying user is a signal.
Basic data storage
Users need their data to persist (be saved and retrievable). A simple database is essential. It does not need to be optimised for scale at this point.
A feedback mechanism
A simple feedback form, an email link, or a chat widget. You need to hear from users. This is the most valuable feature in your MVP.
Whether you are building a SaaS MVP, a fintech MVP, an insurtech app, a healthtech product, or an edtech platform, the same principle applies: one user, one problem, one workflow. Everything else is a distraction at this stage.
What does NOT go in an MVP?
This is where most founders make the same expensive mistake. They build too much. Every feature you add doubles your build time and halves your focus. Cut these from your MVP without guilt.
The pancake rule: An MVP is like a pancake — the first one is never perfect, but you need to make it to learn what to fix. Stop trying to make the perfect pancake before you have ever turned on the stove.
MVP stages — from idea to launch
Building an MVP is not one big task. It is a series of small, focused stages. Here is how a typical MVP journey unfolds.
- 1Stage 1Week 1
Problem discovery
Talk to 10–20 people who have the problem you want to solve. Listen more than you talk. You are looking for patterns — the same pain, described in different words.
- 2Stage 2Week 1
Problem statement
Write one sentence: who has the problem, what it is, and why current solutions are not good enough. If you cannot write this sentence clearly, go back to stage one.
- 3Stage 3Week 1–2
Solution sketching
Draw the simplest possible solution on paper or a whiteboard. No code yet. Just enough to show a potential user and ask: 'Would this solve your problem?'
- 4Stage 4Week 2
Scope locking
Define exactly what is in the MVP and what is not. Write it down. Share it with your team or your agency. Do not start building until scope is agreed.
- 5Stage 5Week 2–5
Build
Build only what is in scope. Nothing else. When a shiny new feature idea pops up, write it in a backlog (a list for later) and keep building the MVP.
- 6Stage 6Week 5–6
Internal testing
Walk through the product yourself, then with one or two trusted people. Fix anything that is broken or confusing. This is not a full quality assurance process — just catch the obvious issues.
- 7Stage 7Week 6
Private beta launch
Give access to 10–50 real users who have the problem. Not friends and family — real potential customers. Watch them use it. Note every moment of confusion.
- 8Stage 8Week 7+
Learn and iterate (improve step by step)
Collect feedback. Prioritise the most common problems. Fix them. Ship an update. Talk to users again. Repeat until your core users are getting consistent value.
How to build a SaaS MVP — step by step
SaaS (Software as a Service — products you pay for monthly and access in a browser) MVPs follow a specific playbook. Here are the 10 steps that work.
- 1
Write your problem statement
In one sentence: who has the problem, what is the problem, and why does it hurt them? Do not start building until you can write this clearly.
- 2
Define your one core user
Pick one type of user to build for. Not five types. Not three. One. A solo freelancer or a team manager or a clinic admin — pick one and build for them exclusively.
- 3
List every feature you want
Write them all down with no filter. Get everything out of your head and onto the page.
- 4
Cut ruthlessly
Keep only the features that directly solve the core problem. Delete or defer everything else. If you are unsure, defer it. You can always add features — you cannot get time back.
- 5
Map the core user journey
Write out every single step a user takes from signing up to getting value. If any step is missing, the product does not work.
- 6
Choose your tech stack
Pick boring, proven technology — Next.js, React, Postgres, Stripe. Now is not the time to experiment with new frameworks. Predictability wins.
- 7
Build the core loop
Code only the steps in your user journey. Every feature not in that journey is out of scope.
- 8
Add basic auth and payments
Use a library like NextAuth or Clerk for authentication (login/signup). Use Stripe for payments. Do not build these from scratch — ever.
- 9
Deploy and share with 10 real users
Do not launch publicly yet. Find 10 people who have the exact problem and put the product in their hands. Watch what happens.
- 10
Interview your users and iterate
Call them on Zoom. Ask what confused them. Ask what they wished the product did. Fix the most common issues and repeat. This loop is the entire point of the MVP.
Need help following this process? A MVP development agency like Greta can run the entire playbook for you — from scoping to deployment — in 2–4 weeks.
How to measure MVP success
Your MVP is not a success because you shipped it. It is a success when it gives you clear signals about whether to keep going, pivot (change direction), or stop. Here are the metrics (measurements) that matter.
Signups
How many people signed up when you told them about the product? If nobody signs up, the problem may not be painful enough — or your explanation is unclear.
More than 10% of people you personally pitch should sign up.
Activation rate
Of the people who signed up, how many actually completed the core action? (e.g. created a project, sent an invoice, booked a call). This tells you if your product delivers on its promise.
Aim for at least 40–60% activation in early beta.
Retention (coming back)
Do users come back the next week? The week after that? Retention is the strongest signal that your product is genuinely useful. No retention means no product-market fit (a product people actually want).
Even 20–30% week-2 retention is a strong early signal.
Qualitative feedback
What are users telling you in their own words? Write down every piece of feedback. Look for patterns. The same complaint from three different users is a feature request.
At least 3–5 user interviews in your first month.
Revenue
Are any users willing to pay for this? Even one paying customer changes everything. It proves the problem is painful enough to spend money on.
One paying customer in your first 100 users is a valid MVP signal.
The 'disappointed' test
Ask users: 'How disappointed would you be if this product stopped existing tomorrow?' If 40% or more say 'very disappointed' — you have product-market fit.
The Sean Ellis benchmark: 40%+ 'very disappointed' = fit confirmed.
Track these in a simple spreadsheet. You do not need a custom dashboard yet. The goal is to answer one question: are people getting enough value to keep using this?
Building an MVP in 2026 — the vibe coding era
Something fundamental changed in 2024 and 2025. AI coding tools became good enough to write real production software. The result: building an MVP is faster and cheaper than at any point in history.
What used to take 6 months now takes 2–3 weeks with vibe coding.
Vibe coding is like describing your dream house to an architect who builds it while you watch. You describe what you want in plain English. The AI generates the code. An engineer reviews it for quality and security. You ship. Repeat.
Before (2022)
Now (2026)
This is what Greta is built for. We combine AI-assisted development with experienced engineers who review every line of code. You get the speed of vibe coding with the reliability of a professional software team.
The rules of the MVP have not changed — validate fast, learn from real users, iterate. But the cost and time of following those rules has dropped by 80–90%. If you have an idea and you are not testing it, you are running out of excuses. Get started today.
Real MVP examples
The best MVPs in history look embarrassingly simple in hindsight. That is the point. Here are three that changed the world.
Three air mattresses on a living room floor
The founders were broke and could not pay rent. They set up three air mattresses in their San Francisco apartment and built a basic website called “Air Bed and Breakfast.” They charged guests $80 a night to sleep on the floor.
The MVP had no search, no reviews, no maps, no payments system. Just photos, a description, and an email to book.
The lesson: They needed to know if strangers would pay to sleep in someone else's home. They found out for almost nothing. Today Airbnb is worth over $70 billion.
A three-minute explainer video
Before writing a single line of code, Dropbox founder Drew Houston made a video. He explained the product, showed a demo, and asked people to sign up for early access.
The product did not exist. The video was the MVP. Overnight, the waitlist grew from 5,000 to 75,000 people.
The lesson: You can validate demand before you build anything. A landing page with a waitlist is a valid MVP if your goal is to test whether people want the product.
One city. One button. No driver management.
The first version of Uber (called UberCab) only worked in San Francisco. It only connected to black car services (no regular drivers). The app had one button: “request a ride.”
No driver ratings. No price estimates. No multiple car types. No surge pricing. No in-app payments at first — they used credit cards on file.
The lesson: One geography. One use case. One button. They validated the core behaviour — “will people use a phone to summon a car?” — before building anything else.
Every one of these companies would have failed if they had tried to build the full product first. The MVP is what let them learn fast enough to survive and grow. The same principle applies to your MVP development today.
Frequently asked questions
What does MVP stand for?+
MVP stands for Minimum Viable Product. It is the simplest version of your product that real users can actually use and test. Minimum means as few features as possible. Viable means it genuinely works and delivers value. Product means it is a real, usable thing — not a mockup or a spreadsheet.
What is the difference between an MVP and a prototype?+
A prototype is a mock-up or demo — it looks like the product but does not work. An MVP is a real, working product that real users can sign up for and use independently. A prototype tests if your design is right. An MVP tests if your idea is right. Both are useful, but they serve different purposes.
How much does it cost to build an MVP in 2026?+
With AI-assisted development and vibe coding, a focused MVP can be built for £5,000–£25,000 in 2026. Traditional development used to cost £50,000–£150,000 for the same output. The key variable is scope — the tighter your scope, the lower your cost. A good MVP development agency will help you cut scope without cutting the core value.
How long does it take to build an MVP in 2026?+
With modern vibe coding tools and AI-assisted development, most MVPs take 2–6 weeks in 2026. A very simple MVP (landing page, waitlist, basic account creation) can be live in a week. A SaaS product with auth, a core workflow, and Stripe payments typically takes 3–5 weeks. Traditional development used to take 3–9 months for the same output.
What should NOT go into an MVP?+
An MVP should not include advanced analytics dashboards, multiple subscription tiers, full admin panels, complex third-party integrations, polished onboarding tours, email notification sequences, or any feature that only power users would ever need. The rule: if removing it does not break the core user journey, remove it. You can always add it later.
Can I build an MVP without coding?+
Yes, in many cases. With vibe coding tools like Cursor, Bolt, and Lovable, you can describe your product in plain English and an AI will generate working code. You can also use no-code platforms like Webflow or Bubble for simpler products. For more complex SaaS products, working with an experienced MVP development agency is often the fastest and most reliable path — especially if you want production-quality code you can scale later.
What is product-market fit and when do I have it?+
Product-market fit (PMF) means enough people want your product badly enough to keep using it — and ideally to pay for it. You know you have it when users come back on their own, retention stays high week after week, and people tell others about your product without being asked. The Sean Ellis test: if 40% or more of your users would be 'very disappointed' if your product disappeared tomorrow, you likely have product-market fit.
Should I patent my MVP idea before building it?+
In almost every case, no. Most software ideas are not patentable in the way founders imagine. More importantly, your biggest risk is not that someone copies your idea — it is that nobody wants it. Build the MVP, validate the idea, then worry about intellectual property if you have real traction. Speed to market is your best protection at the MVP stage.
Build your MVP in 2–4 weeks.
Stop planning and start validating. Greta is a specialist MVP development agency that ships production-quality products in weeks, not months. AI-assisted, human-audited, founder-focused.