Types of MVPs

An MVP is not a single form — it is a category of experiment. Ries identifies several distinct archetypes, each designed to test a specific kind of assumption with the minimum possible effort. Choosing the right MVP type is a strategic decision: the wrong type produces answers to questions you weren’t asking.

The four primary MVP archetypes:

1. Video MVP

A demonstration video that shows the product working as if it already exists, used to test whether customers actually want the product before a single line of production code is written.

• Classic example: Drew Houston’s Dropbox explainer video (2007). The video showed the product’s core value proposition — seamless file syncing across devices — to a waitlist. Signups jumped from 5,000 to 75,000 overnight.
• What it tests: Value-Hypothesis — does anyone want this? Is the problem-solution fit compelling enough to generate intent?
• Best for: Products where the user experience is the differentiator; when the concept requires demonstration to be understood

2. Concierge MVP

A human-powered service that manually delivers the product experience without any automation or technology. The team performs every step by hand to learn what customers actually need before building a system to automate it.

• Classic example: Food on the Table (2009). Manuel Rosso personally visited customers’ homes, walked to their local grocery stores, and manually prepared custom meal plans and grocery lists each week — before writing any software.
• What it tests: Whether the product experience produces the desired outcome; reveals exactly which steps add value and which are irrelevant before automating them
• Best for: Service-oriented products; cases where customer needs are unclear; when premature automation would lock in the wrong process

3. Wizard of Oz MVP

The product appears to be automated and functional to customers, but is actually operated manually behind the scenes. The customer experience mimics a real product; the “engine room” is human effort.

• Classic example: Zappos (1999). Nick Swinmurn photographed shoes from local stores, posted them on a website, then — when orders arrived — bought the shoes at retail price and mailed them himself. Proved demand for online shoe purchasing without inventory, logistics, or warehousing.
• Origins: The technique predates startups. HCI researcher Jeff Kelley documented the “Wizard of Oz” experimental methodology at IBM in 1984, describing how researchers simulated intelligent interfaces by having a hidden operator respond to user inputs — allowing testing of AI interactions before the AI was built.
• What it tests: Whether customers will take real action (purchase, sign up, engage) given the product experience — not just whether they say they will
• Best for: Testing complex technical systems; validating two-sided marketplace demand; testing behaviour rather than stated preference

4. Smoke Test / Landing Page MVP

A minimal web page describing the product that measures conversion intent — signups, email captures, or purchases — before the product exists. The “smoke test” name comes from electronics: if you turn something on and smoke appears, it fails; if you capture zero conversions, the concept fails.

• What it tests: Whether customers are interested enough to take a concrete action (provide an email, click “buy”) before the product is built
• Best for: Early concept validation; B2C products where the acquisition funnel is the primary variable; testing messaging and positioning

Choosing Between Types

The choice depends on what assumption you are testing:

MVP Type	Tests	Effort Level
Video	Problem-solution fit / desire	Low
Smoke Test	Conversion intent / messaging	Low
Concierge	What customers actually need	Medium
Wizard of Oz	Real behaviour under product conditions	Medium-High

The key distinction between Concierge and Wizard of Oz: in a Concierge MVP, the customer knows the service is manual; in a Wizard of Oz MVP, they believe it is automated. Both serve Validated-Learning, but they test different things.

Ash Maurya (Running Lean, 2012) further categorizes MVPs along a “solution interview” → “smoke test” → “concierge” → “single-feature product” continuum — prioritizing customer conversation before any artefact is built.

Related Concepts

• Minimum-Viable-Product
• Leap-of-Faith-Assumptions
• Early-Adopters
• The Lean Startup - Ries - 2011

Future Connections

Will connect to Validated-Learning, Value-Hypothesis when confirming links.

Sources

• Ries, Eric (2011). The Lean Startup: How Today’s Entrepreneurs Use Continuous Innovation to Create Radically Successful Businesses. Crown Publishing. ISBN: 978-0-307-88791-7.

  • Chapter 6 (Test) — primary source for the Dropbox, Food on the Table, and Zappos MVP examples and the typology of MVP approaches

• Kelley, John F. (1984). “An iterative design methodology for user-friendly natural language office information applications.” ACM Transactions on Information Systems, Vol. 2, No. 1, pp. 26–41. DOI: 10.1145/357417.357420.

  • Foundational paper documenting the “Wizard of Oz” simulation technique in HCI research; predates the startup application by more than two decades and establishes the methodological rigour behind the technique

• Maurya, Ash (2012). Running Lean: Iterate from Plan A to a Plan That Works (2nd ed.). O’Reilly Media. ISBN: 978-1-449-30517-8.

  • Extends and systematises MVP thinking beyond Ries; introduces a structured sequence from customer interviews to smoke tests to concierge MVPs; provides a practitioner-level taxonomy of experiment types

• Boehm, Barry et al. (1988). “A spiral model of software development and enhancement.” IEEE Computer, Vol. 21, No. 5, pp. 61–72. DOI: 10.1109/2.59.

  • Iterative prototyping as risk-reduction — the theoretical precursor to MVP methodology in software engineering; demonstrates that early, incomplete versions reduce development risk

• Klein, Laura (2013). UX for Lean Startups: Faster, Smarter User Research and Design. O’Reilly Media. ISBN: 978-1-449-33232-7.

  • Extends MVP typology from a UX research perspective; emphasises that non-functional prototypes (paper mockups, clickable wireframes) are valid MVPs for testing user behaviour; shows how each MVP type maps to specific user research questions

Note

This content was drafted with assistance from AI tools for research, organization, and initial content generation. All final content has been reviewed, fact-checked, and edited by the author to ensure accuracy and alignment with the author’s intentions and perspective.