Introduction
The terms MVP, POC, and prototype get used interchangeably in pitches, in scoping calls, and in agency quotes. They are not the same. Using the wrong one for the job wastes money and produces artifacts that do not answer the question you needed to answer. This guide distinguishes them crisply.
Prototype — for design and interaction validation
A prototype is a high-fidelity clickable experience that does not have a real backend. Figma clickthroughs, Framer sites, and InVision mockups all qualify. Its job is to test whether the interaction model and visual design are comprehensible and desirable before you write production code.
Audience: users in research sessions, investors who want to see the feel of the product, internal stakeholders aligning on UX.
Budget: $5k–$25k for a senior designer with 1–3 weeks of work. Should not include engineering.
- High-fidelity clickable, no real backend
- Figma/Framer/InVision-level artifacts
- Purpose: test interaction model and design desirability
- Budget: $5k–$25k, 1–3 weeks, no engineering
POC — for technical feasibility
A Proof-of-Concept is a throwaway code spike that answers one technical question: can we build this? Can the LLM achieve acceptable quality on our data? Can we integrate with this API within the latency budget? Can this architecture handle the throughput we need?
A POC is not a product. It has no auth, no multi-tenant support, no billing, no polish. It is the smallest amount of code that retires the biggest technical risk.
Audience: internal engineering team, technical investors, enterprise buyers who require technical validation before a contract.
Budget: $15k–$50k for 2–4 weeks of senior engineer time. Output is a demo + a written technical assessment.
- Throwaway code spike answering one technical question
- No auth, no billing, no multi-tenant, no polish
- Audience: engineering team, technical investors, enterprise buyers
- Budget: $15k–$50k, 2–4 weeks
MVP — for market validation
A Minimum Viable Product is the narrowest possible slice of product that real users can complete a real job inside, with enough instrumentation to measure whether it is working. It has auth, core workflow, basic analytics, and is App Store or production-deployed.
An MVP is not a polished product. It has clear rough edges, limited admin tooling, minimal personalization, and a short feature list. Its job is to test one thing: will users adopt this for the value it provides?
Audience: real users paying real money (or providing real retention signal), investors who want evidence of product-market fit, early partners.
Budget: $55k–$180k for 6–10 weeks with a senior team.
- Narrow slice of product real users can complete a real job in
- Has auth, core workflow, basic analytics, production deploy
- Purpose: test whether users will adopt for the value
- Budget: $55k–$180k, 6–10 weeks
Deciding which one you need
If the question is 'will users want this shape of experience?' — build a prototype.
If the question is 'can we technically build this?' — build a POC.
If the question is 'will users pay for and return to this?' — build an MVP.
Most founders need an MVP and over-invest in prototypes or POCs for comfort. Most technically ambitious teams need a POC first to retire risk before committing to an MVP budget. Most design-heavy or new-interaction-model products need a prototype before an MVP.
- User wants the shape? → prototype
- Is this technically feasible? → POC
- Will users pay and return? → MVP
- Most founders need an MVP but fund a prototype for comfort
Combining them — the right order
For genuinely risky products, the right sequence is prototype → POC → MVP. Each stage retires different risks before the next investment. Total: $75k–$250k and 2–4 months, versus $55k–$180k for an MVP alone.
The extra cost is justified when the product has both novel UX (requiring design validation) and novel technology (requiring feasibility validation). For most products — recognizable SaaS, standard mobile apps, incremental innovations — skip straight to MVP.
- Order: prototype → POC → MVP for genuinely risky products
- Total: $75k–$250k, 2–4 months
- Skip straight to MVP for recognizable product categories
Conclusion
Prototype, POC, and MVP each exist to answer a different question. Matching the artifact to the question saves money and produces evidence you can act on. When in doubt, most founders should skip straight to a disciplined six-week MVP — it answers the market validation question that prototypes and POCs cannot.