You have a validated business idea. You have got a small budget. You need a working product in front of customers before that budget runs out. And every developer you talk to gives you a different number — 4 weeks, 3 months, 6 months, "let us discuss."
Here is the honest answer from someone who has shipped MVPs for himself and for others: a production-grade MVP using modern web tech should take 3 to 8 weeks for most business ideas. Not 6 months. This post walks you through the realistic week-by-week timeline, what has to be in scope, and what should be cut ruthlessly.
Before we get to the week-by-week, let me address the elephant.
Agencies quote long timelines for four reasons:
An independent engineer with modern AI-assisted tooling can ship the same MVP in 30-40% of the agency timeline. Not because the engineer is better than the agency's engineers. Because the process is not padded.
Not all MVPs are the same. Here are the three most common shapes I build:
You have a specific narrow tool. Users sign up, do the thing, pay a subscription.
Examples: a niche scheduling tool, a specialised calculator, a directory site with paid listings, a personal finance tracker for a niche audience.
Realistic scope:
Realistic price: ₹1,50,000 – ₹3,50,000
Users on both sides — buyers and sellers, patients and doctors, students and tutors.
Examples: a niche service marketplace, a booking platform, a mentor matching site.
Realistic scope:
Realistic price: ₹3,00,000 – ₹6,00,000
Any product where the AI is the core value proposition.
Examples: an AI writing assistant, an AI-powered analytics tool, an AI chat interface for a specific niche.
Realistic scope:
Realistic price: ₹2,50,000 – ₹5,50,000
These are floors. Bigger scope = bigger price. If a vendor is quoting you 3-4x these numbers for what sounds like an MVP, ask why.
Let me walk through the realistic timeline for the most common shape. This is what I actually do.
Nothing user-facing looks impressive yet. This is scaffolding.
End of week 2: a user can sign up, log in, use the core feature. It looks rough. It works.
Total realistic time: 4-5 weeks for a simple SaaS.
Anyone quoting 12 weeks for this exact scope is padded. Anyone quoting 2 weeks is skipping something (usually payments, testing, or reliability).
Every week during the build, you will think of a new feature that "would be so easy to add." Every one of those additions extends your timeline by more days than you think. Ship without them. Add them after launch based on real user feedback.
Best MVPs are shown to 5-10 real potential users during week 2-3, not the day of launch. Every week you delay showing it, you build more things nobody wants.
"We can add payments later." No. Payments are 3-4 days of work and need to be tested end-to-end. If your MVP charges money and does not have payments on day 1 of launch, you shipped a demo, not a product.
"We will add monitoring once we have users." No. Monitoring on day 1 catches the bug that would have made your first 20 users churn silently. Add it before launch, not after.
Do not ship your MVP on the shiniest new database or framework. Use boring, well-supported tech that works. Save the exotic choices for version 2.
For a shipped MVP running on modern serverless infrastructure:
Total monthly running cost for a launched MVP: ₹0 – ₹3,000/month for the first 500 users. Do not over-plan infrastructure.
If you want a fixed quote for your specific MVP, message me on WhatsApp with a one-page description. I will tell you honestly whether it fits the 3-8 week window and what a fair price is.
Why do traditional agencies quote 6 months for what an independent engineer builds in 5 weeks?
Padding, waterfall process, team switching costs, and overspecification. An independent engineer with modern tools does the same work with 30-40% of the elapsed time.
What features should be omitted from an initial MVP to hit a 4-week launch deadline?
Multi-language, multi-currency, team accounts, advanced admin, integrations beyond the essential one, native mobile apps, onboarding tours, reporting dashboards. Ship the core value plus payments plus monitoring. That is it.
Is no-code (Bubble/WordPress) faster than a custom Next.js build for an MVP?
For the first 1,000 users, sometimes yes. For scale, cost, and flexibility beyond that, no-code becomes a liability. If your MVP is a business idea you plan to grow, custom code from day 1 is usually cheaper over 2 years.
What are the typical monthly cloud running costs for a newly launched web app?
₹0-₹3,000/month for the first 500 users on modern serverless stack. Costs scale with usage, roughly linearly. A profitable SaaS at 5,000 paying users typically runs at ₹15,000-₹40,000/month total infrastructure.
How do we transition from an MVP to a scaling production system?
Usually not by rewriting. Add tests, add monitoring, refactor the specific parts that hurt as you grow, hire your first engineer. Full rewrites are almost always a mistake.
Or WhatsApp me with your MVP idea for a straight quote.