What is an MVP in app development?
MVP does not mean "small app". It means the minimum to test a hypothesis with real users. Here is the framework that separates real MVPs from undercooked v1.
"Let\'s build an MVP" is the most misused phrase in startup jargon. Most people use it to mean "a small app." That\'s wrong. MVP (Minimum Viable Product) is a hypothesis-testing tool — the smallest possible version that lets you learn whether your assumptions about users are correct. Here\'s the proper framework.
Quick answer — what an MVP actually is
- The smallest possible version of an app
- That delivers one core value to one user persona
- Tested with real users in a real market
- Designed to validate or invalidate a hypothesis
- Built in 6-10 weeks max for SMB-scale apps
The 3 dimensions every MVP must have
1. Minimum — really minimum
If you need to write down 6 features for v1, you defined the wrong MVP. Truly minimum apps have 1-3 features.
Example: Dial Your Meal\'s MVP wasn\'t the full marketplace. It was: customers can browse 10 partner restaurants in one Sydney suburb and place an order. That\'s it. No reviews. No promo codes. No driver tracking. Just place-an-order in one neighbourhood.
2. Viable — actually works
Common error: shipping a prototype and calling it MVP. Prototypes have fake data, broken edge cases, no real backend. MVPs work end-to-end:
- Real auth (not "skip login")
- Real database (not local mock)
- Real payment if commerce (not "coming soon")
- Real push notifications (not stubbed)
- Real error handling (not crash on edge cases)
3. Product — solves a real need
An MVP must solve a problem someone is willing to pay for or actively use. If users can\'t articulate what pain it solves, you don\'t have an MVP — you have a feature looking for a problem.
The MVP scoping framework
Step 1 — Write the hypothesis
One sentence: "I believe [user type] will [do specific action] because [their pain] makes them want [value]."
Example: "I believe restaurant owners in Sydney will pay 5% commission to use our app because the existing aggregators charge 30%+ and they want to keep more margin."
Step 2 — Identify ONE user persona
Not "people who want food delivery." Specifically: "restaurant owners in Sydney CBD with 20-50 daily orders."
Step 3 — Identify ONE core action
The action that, if it works, validates the hypothesis. Everything else is secondary.
For DYM: "owner can list menu, accept orders, and get paid weekly." That\'s the MVP.
Step 4 — List ONLY the features needed for that action
Self-test for each feature: "If we don\'t have this, can the user still complete the core action?" If yes — defer it.
Step 5 — Build with measurable signals
What metrics will tell you if it\'s working?
- Activation: % users completing core action within 24 hours
- Retention: % users returning in week 2
- Economics: revenue per user / cost per user
- Qualitative: 5-10 user interviews after 1 week
Real Indian MVP examples (clients we\'ve shipped)
| Idea | Bad MVP | Good MVP |
|---|---|---|
| Field rep tracking | "Track everything" + dashboard + reports + payroll | GPS check-in + photo + manager view. That\'s it. |
| D2C food brand | Full e-commerce + subscriptions + referrals + reviews | 3 SKUs, COD only, WhatsApp re-order |
| Coaching marketplace | Two-sided + payments + reviews + classes + recordings | 10 coaches, manual booking, IPay later |
| Restaurant ordering | Full menu + orders + delivery tracking + ratings | 10 menu items, table-side QR, UPI payment |
The MVP timeline (6-10 weeks honest version)
| Week | Focus |
|---|---|
| Week 1 | Hypothesis + persona + core action defined. Scope locked. |
| Week 2 | Wireframes + click-through prototype. Tested with 5 users. |
| Week 3-4 | Backend + auth + database. Critical SDKs integrated. |
| Week 5-6 | Frontend screens. Internal testing. |
| Week 7 | QA + beta testing with 10-20 users. |
| Week 8 | Launch in limited market. Watch metrics. |
What MVP is NOT
- NOT a feature-rich v1 — that\'s a full product
- NOT a prototype — that\'s a clickable mockup
- NOT a "lite version" — that\'s a deliberate downgrade
- NOT a 6-month build — that\'s a v1, not MVP
- NOT for everyone — that\'s a mistake; pick one persona
- NOT free if your business is paid — pricing is part of MVP
How to read MVP signals
After 4-8 weeks of MVP being live with 50-200 users:
| Signal | Meaning |
|---|---|
| 50%+ users return week 2 | Strong product-market fit signal — invest in v2 |
| 20-40% return week 2 | Some signal — fix biggest friction, retest |
| Under 20% return | Wrong product, wrong audience, or core hypothesis failed |
| Users use it differently than you expected | Pivot opportunity — listen carefully |
| Same 5 feature requests from everyone | That\'s your v2 roadmap |
Common MVP mistakes
- Confusing MVP with prototype
- Adding features "just in case"
- Targeting multiple personas in v1
- Free / no pricing — fails to validate willingness to pay
- No instrumentation — flying blind
- Stretching to 4-6 months instead of 6-10 weeks
- Polished design when fast iteration matters more
- Building for engineers, not for users
Pro tips
- Charge from day 1 — even ₹49 / month tells you 100× more than free
- Limited geography — one neighbourhood, one city, one industry
- Manual backstage — automate the user-facing flow, do operations manually
- Talk to 5 users every week for first 2 months — pattern matching
- Set a kill criterion upfront: "if X doesn\'t happen by week 12, we kill it"
Conclusion
An MVP is a hypothesis-testing tool, not a small app. Pick one persona, one core action, ship in 6-10 weeks, charge from day one, measure activation and retention. If you want a discovery sprint that scopes a real MVP rather than an undercooked v1, see Mobile app development or book a strategy call.
FAQs
Smallest version that lets ONE type of user complete ONE core action. If you need 5 features for the user to complete it, that is your MVP. If you need 12, you defined the wrong MVP.
Prototype: clickable mockup, no real backend, no real users. MVP: working app with real backend, deployed to a real test market with paying or active users. MVP teaches you market truths, prototype teaches design truths.
A real MVP for an SMB app: 6-10 weeks. Less than that is usually a prototype calling itself MVP. More is usually scope creep.
₹1.5L to ₹4L for single-platform with focused scope. ₹4-8L for cross-platform with payments. Below ₹1L is a templated prototype, not a real MVP.
Test pricing in MVP — even ₹100 paid users teach you 10x more than 1,000 free users. Free MVP = wrong product validation.
When you have signal — repeat usage, positive economics, clear feedback themes. Usually 2-4 months after launch with 100+ active users.