The worst thing you can do is spend months, even quarters, building what feels like the perfect product without testing or validating with real users. I have watched teams pour time and budget into polish while the market shifts or never existed to begin with. Progress looks like being busy, but that doesn’t mean an idea is ready to launch.
Instead of building, I chase proof early. A small live test beats a beautiful build. One real click, conversation or paid trial. These are the signals that shape the roadmap and fund the next step.
I surveyed 312 founders across the Launching Next audience in November 2025:
- 69 percent are only somewhat confident or unsure their idea solves a real problem
- 44 percent mentally anchor on “Minimum” and 39 percent on “Viable”
- Top hesitations were “won’t look professional” (31 percent) and “not sure how to measure success” (28 percent)
- First MVP choices cluster with Landing page or video (33 percent) and Wizard of Oz (24 percent)
This guide gives you the shortest path from idea to evidence. You will pick a test, put it in front of real people, measure what matters, then decide your next move with confidence. We’ll walk through 16 real minimum viable product examples. Steal my MVP Picker and ship your test this week.
By the end of this guide, you’ll have a picked MVP type, created a 7-day test plan, and have a path to decide pivot vs. continue.
What is a minimum viable product?
A minimum viable product (MVP) is a working first version built to test your single riskiest assumption with real users and real feedback. Minimally means we cut everything that does not serve that experiment. Viable means the core value works well enough that feedback is trustworthy.
Success for an MVP is a validated learning you can act on, not code shipped.
TL;DR: Pick the MVP that matches your riskiest assumption
- Demand risk: Landing page, fake door or explainer video
- Behavior or trust risk: Wizard of Oz MVP
- Workflow or problem fit risk: Concierge MVP
- Feasibility or integration risk: Single feature coded as an MVP
- Speed or budget risk: Piecemeal using stitched third party tools
We’ll go into each of these in this article.
MVP, vs. Prototype, vs. Proof of Concept, vs. MLP, vs. MMP
You move faster when you choose the right level of fidelity for the question you are asking.
Quick roles
- MVP: A minimal version that delivers core value to real users and returns data you can act on.
- Prototype: Internal learning built to explore flows or UX. No real users and no market value.
- Proof of Concept (PoC): Feasibility only that proves a technical claim can work once. No distribution and no customers.
- Minimum Lovable Product (MLP): An MLP raises the bar from workable to worthy of word of mouth. You keep scope tight yet add the one touch that creates delight. Use it when competition exists and you need adoption energy and referrals from day one.
- Minimum Marketable Product (MMP): An MMP is the first version you can sell consistently. The value is proven, onboarding holds, support is serviceable and pricing is coherent. Use it when your MVP cohorts retain and you are ready to turn on a repeatable channel.
The 5 core MVP types, and when to use each
Landing page, Fake door or Explainer video
- Choose when: You must validate demand and messaging
- Looks like: A clear promise plus a single CTA or a short demo video
- Metric: Qualified click or sign up intent and message replay in replies
- Beginner move: Single landing page, plus email capture
- Operator move: Paid channel split tests across headlines and offers
Wizard of Oz MVP
- Choose when: Users must believe automation works and you can manually fulfill
- Looks like: A front end UI while humans simulate the backend
- Metric: Completion rate and willingness to repeat or pay
- Beginner move: Manual fulfillment in DMs
- Operator move: A light CRM with scripts and tagged outcomes
Concierge MVP
- Choose when: You need deep workflow learning from a handful of users
- Looks like: The founder doing the job one to one with optional fee
- Metric: Repeat usage plus referrals across a tight window
- Beginner move: Three to five customers and weekly check ins
- Operator move: Calendar scheduling with forms and templated SOPs
Piecemeal MVP using stitched tools (Google Sheets, Zapier, Make, n8n, JotForm, etc.)
- Choose when: Speed matters and third party tools can simulate the full flow
- Looks like: Airtable, Jotform (my favorite given cheaper pricing and a large feature set, I’m not affiliated with them) or Google Sheets with forms and no code automations
- Metric: End-to-end success rate across a real task
- Beginner move: A simple Zapier or n8n stack
- Operator move: Add payments and lightweight analytics with events
Single feature coded MVP
- Choose when: The risk is technical feasibility or the UX of one killer feature
- Looks like: One core feature built well with rough edges elsewhere
- Metric: Feature adoption and week over week return
- Beginner move: Narrow scope to a single use case
- Operator move: Track the feature with logs and session replays for rapid fixes
16 real MVP examples
Copy the MVP that maps to your riskiest assumption, not the vertical or aspirational brand.
Duolingo closed beta
- Risk tested: Can free language learning with a tight daily loop create retention without paid content on day one
- What shipped: An invite only web app with a handful of courses and a single core loop that rewarded short daily practice
- Metric to copy: Day 1 and day 7 retention and the share of users with a 3 day streak
- Steal this: Run a private beta with a waitlist and one core loop then do a weekly cohort read to decide whether to open the gates
Zoom desktop pilot
- Risk tested: Will small teams switch for reliably clear video and simple join flows
- What shipped: A lightweight Mac and Windows client with join links and screen share for limited teams
- Metric to copy: Join success rate on first attempt and average call quality rating after each meeting
- Steal this: Recruit ten teams and instrument time to first video and post call thumbs up then fix friction daily
Crazy Egg explainer video
- Risk tested: Do teams want heatmaps enough to give emails or prepay before the tool exists
- What shipped: A short video demo that showed the experience end to end
- Metric to copy: Video viewers to email signups and pre purchase intent on the pricing click
- Steal this: Record a two minute click trail and pair it with a single field email capture and a pricing tease
Wealthfront concierge MVP
- Risk tested: Will consumers trust an algorithmic portfolio if it earns trust in a human delivered plan first
- What shipped: A manual planning service that mirrored the target allocation and rebalancing rules
- Metric to copy: Plan acceptance rate and funded account rate within seven days of the plan
- Steal this: Serve five clients by hand using the future rules and document every question that blocks funding
Turo peer to peer car sharing
- Risk tested: Will owners list and renters book if insurance and handoff are solved locally
- What shipped: A city-level marketplace with manual onboarding and key exchanges while the team handled insurance and support
- Metric to copy: Completed bookings per active car and first repeat booking rate within thirty days
- Steal this: Pick one neighborhood and one airport route then personally handle the first twenty handoffs and log every policy and trust blocker
Dropbox explainer video (Digital MVP)
- Risk tested: Do people want seamless multi-device file sync
- What they shipped: A short demo video that showed the experience before code existed
- Metric to copy: Waitlist growth from highly qualified audiences
- Steal this: Script a 90-second screen-record and post it where your early adopters already hang out
Zappos Wizard of Oz
- Risk tested: Will customers buy shoes online without trying them on
- What they shipped: Simple site while the founder bought and shipped pairs by hand
- Metric to copy: First-purchase conversion and repeat orders
- Steal this: Photograph inventory you can access locally and fulfill the first 10 to 20 orders manually
Airbnb concierge or piecemeal MVP
- Risk tested: Will travelers book spare space during peak events
- What they shipped: Basic site while the founders hosted early guests and learned trust levers
- Metric to copy: Bookings per listing and host or guest satisfaction notes
- Steal this: Curate 3 to 5 hosts you personally know and handhold the first stays
Groupon piecemeal MVP
- Risk tested: Will people redeem limited-time local deals in volume
- What they shipped: A simple blog with manually created PDF coupons and emails sent through basic tools
- Metric to copy: Email sign-ups to offer clicks to redemptions
- Steal this: Start with one merchant and one irresistible offer and send PDF codes manually
Buffer landing page MVP
- Risk tested: Will users pay for scheduled social posting
- What they shipped: A landing page with pricing tiers that collected emails and signaled intent
- Metric to copy: Percent choosing paid plans on the pricing clickthrough
- Steal this: Show pricing before product and tag interest by tier inside your email tool
DoorDash concierge MVP
- Risk tested: Will local delivery demand sustain manual ops
- What they shipped: A basic site with a phone number while founders delivered themselves in one neighborhood
- Metric to copy: Order frequency per customer and delivery time consistency
- Steal this: Pick one neighborhood and one lunch window and run the route yourself for a week
Product Hunt email digest
- Risk tested: Will a daily curation of new products become a habit
- What they shipped: A link-sharing group that became a daily email before the site took shape
- Metric to copy: Daily open and click rates plus community submissions
- Steal this: Run a curated newsletter for your niche for two weeks and invite replies
AngelList email list and directory
- Risk tested: Can founders and investors match efficiently online
- What they shipped: Simple profiles and email introductions before a full productized network
- Metric to copy: Warm intros made and reply rates
- Steal this: Build a lightweight introductions form and track outcomes in a sheet
Uber early UberCab SMS and basic app pilot
- Risk tested: Will riders summon and pay for on-demand black cars
- What they shipped: A narrow city pilot with a minimal way to connect rider and driver and text requests
- Metric to copy: Request completion rate and rider repeat use
- Steal this: Constrain geography and hours and nail reliability before you widen coverage
Spotify closed desktop beta
- Risk tested: Can streaming feel instant and reliable enough to replace downloading
- What they shipped: An invite-only desktop player with a limited catalog to stress test speed and playback
- Metric to copy: Session starts, skips, and return cohorts
- Steal this: Run a private beta around a single killer listening experience and add catalog later
Twitter twttr and SMS
- Risk tested: Will people post ultra short status updates from their phones often enough to form a new habit
- What shipped: An internal SMS broadcaster for the Odeo team that evolved into an invite only web app
- Metric to copy: Posts per user per day and percent of new users who post in week one
- Steal this: Spin up a simple SMS workflow with Twilio and a bare web feed then watch daily posting and reply loops for one small community
Every one of these MVPs trimmed the build to the smallest unit that produces trustworthy feedback. Steal the structure and the metric, then adapt both to your riskiest assumption.
How to ship an MVP in 7 days
Give yourself one week and hold yourself to that target. I run this sequence with solo founders and tiny teams because it forces a single bet, real users and a decision you can stand behind.
- Day 1: Decide on the riskiest assumption, the MVP type, one metric and a decision rule. Write the pass threshold (a go or no go line) and the next step you will take for each outcome.
- Day 2: Build the smallest artifact you can ship today. It could be a page, video, form or a sales script. Turn on basic analytics and create a simple notes doc.
- Day 3: Recruit 10 to 30 target users. Use DMs, communities, warm intros or a small paid test. Log who they are and where they came from.
- Day 4 to 5: Run the experiment. Manually fulfill anything you can. Capture friction/hesitations, questions and unexpected behavior in your notes doc.
- Day 6: Hold eight to ten short interviews at fifteen minutes each. Ask what they tried to do, where they hesitated, what they expected next, and what they would pay.
- Day 7: Scorecard review. Decide to continue and grow, iterate or stop. Save the decision in writing for your future self.
During this one week of sustained effort, you will gain an unprecedented amount of insights, context and detail that you would never find growing Reddit, X or asking ChatGPT.
Common MVP failures and how to avoid them
Anchor the test to one thing. Most pain comes from fuzzy goals and bloated landing pages. Use these guardrails.
- Testing too many things at once: Lock one hypothesis per test and state the pass threshold first
- Over-polishing the non-core: Pour quality into the core value only, and keep everything else serviceable
- Skipping a decision rule: Write the decision rule before you launch, and share it with anyone helping you
- Too wide a scope: Constrain by segment and geography so you can recruit and fulfill quickly
- No audience plan: Line up one channel the same day you build and schedule outreach on your calendar
FAQs on minimum viable products
What are the main MVP types
Five workhorse patterns handle most idea risks.
- Landing page or explainer for demand and message testing
- Fake door for intent measurement without a full build
- Wizard of Oz for behavior and trust when you can fulfill manually
- Concierge for workflow learning with a handful of users
- Piecemeal or single feature coded for feasibility and end to end flow with stitched tools
Which MVP is best for B2B SaaS
Start where the risk sits.
- Demand or value prop unclear: Use a landing page with a Calendly ask for discovery
- Workflow and integration risk: Run a concierge with three design partners and a shared doc or Airtable
- Security and data risk: Ship a single feature coded MVP in a sandbox account with logs and guardrails
Is a fake door test ethical
Set clear expectations and give value.
- State that the feature is in development when they click
- Offer a waitlist or a fast interview slot
- Share what you learned with the list and invite them first when live
- Avoid collecting payment before you can deliver
MVP vs prototype vs PoC quick recap
Prototype explores user experience internally. Proof of concept (PoC) proves a technical claim once. MVP delivers core value to real users and returns decision ready data.
How many users do I need for an MVP
- Qualitative learnings: Ten to fifteen target users will surface the big blockers
- Simple demand test: 100-300 qualified visits can validate a basic conversion
- Behavior loops: Aim for 50 to 100 active users in a tight segment to read retention
What metrics prove MVP traction
Pick one primary metric before you launch.
- Demand tests: Qualified signup rate and percent choosing a paid tier on a pricing click
- Behavior tests: Week one activation and day 7 or week 4 retention
- Concierge or Wizard of Oz MVPs: Completion rate and willingness to repeat or pay
- Piecemeal or single feature: End-to-end task success and time to value
Should I charge for my MVP
Charge when money is part of the risk.
- Add a real price tier even if you plan to comp early users
- Use a credit card capture for pilots when trust and willingness to pay matter
- If the risk is purely behavior or feasibility, defer payment and charge on the MMP
How do I recruit users quickly without an audience?
Go where your ideal customer profile already gathers, use targeted communities, direct messages, and a small paid test, keep your ask clear and time bound.
When should I move from a Wizard of Oz MVP to code?
Switch once users repeat the behavior and the manual steps become the bottleneck, document the steps, then automate the highest friction link first.
How do I avoid common MVP problems?
Test one hypothesis per run, pour quality into the core value only, constrain scope to a tight segment and geography, and line up your channel the same day you build.
