MVP cost

How much does it cost to build an MVP

Most answers to that question are a single number with the reasoning left out, or a form that wants your phone number first. This page gives you a range in about thirty seconds and shows the arithmetic behind it. Then it explains what moves the number up and down. United States pricing, fixed price, no email required.

Use the calculator See how we build MVPs

Projects start at $4,000. Most fixed-price builds land between $15,000 and $60,000.

Cost drivers

What actually drives the cost of building an MVP

Software pricing feels arbitrary from the outside because the thing being priced is invisible. Two products can look identical in a demo and differ by a factor of five underneath. So it helps to know which parts of a build actually consume time, because time is the only real input.

Five things move an MVP quote more than anything else.

The number of distinct things a user can do

Count actions rather than screens. Posting a listing is one action. Editing that listing afterwards is close to free, because the hard part already exists. Letting somebody else respond to it, get notified, and pay for it is three more. Founders usually count screens and undercount actions, which is why a two-screen product sometimes costs more than a ten-screen one.

Whether more than one kind of person uses it

A product with one user type is a straight line. Add a second type who sees different data and every query and every screen now carries a branch. Add an admin who can act on a customer's behalf and there is a third set of rules to write and test. Second user types are the most underestimated cost in early product work. When a quote comes back higher than a founder expected, this is usually why.

Money

Taking a card payment once is a solved problem and costs about a week. Recurring billing is a set of features that arrive together. That means plans, upgrades, downgrades, failed cards, refunds, proration and the email that goes with each one. Founders who say they need billing usually do need it, and they should know they are buying a subsystem rather than a button.

Connections to software you already run

Every outside system you connect to is a small project of its own, with its own sign-in, its own rate limits and its own way of being unavailable at the worst moment. One connection is fine. Four connections is a sizeable share of the budget, and it is always worth asking whether the first version requires all four or whether two of them can be done by hand while you have twenty customers.

Design

If your product can use clean, standard patterns, design is nearly free, because good defaults already exist and users already know them. If you need original screens drawn before anybody writes code, that is real design work, and it sits ahead of the build in the schedule, so it stretches the timeline as well as the price.

Two things that matter less than founders expect

Which programming language gets used almost never changes the price. Any competent team is roughly as fast in the three or four stacks they know well, and the choice should be made on what is easiest to hire for later rather than on cost. The number of users you expect at launch does not change the price much either, because the difference between serving one hundred people and one thousand is a configuration setting rather than an architecture. Planning for a million users before you have ten is the most expensive mistake in this category.

One thing that matters more than founders expect

How clear your requirements are on the first day. A vague brief gets priced with a risk premium, because the team has to protect itself against the parts you have not decided yet. Founders who arrive able to state the core user action in one sentence consistently get lower quotes and shorter schedules than founders who arrive with a long list of maybes. This is free to fix, and it is worth doing before you talk to anybody.

Cost by product type

What an MVP costs by product type

The ranges below are our own fixed prices for delivery from the United States. Each one assumes a written specification signed off before the build starts, a deployed product you can put in front of real users at the end, full source code handed to you, and a 30-day warranty. These are not offshore hourly rates, and the number does not get revised upward once work is underway.

MVP cost and build time by product type
Product type	Typical fixed price	Build time	What is included
Single-purpose web app	$4,000 to $12,000	2 to 4 weeks	One core action, one user type, real database, deployed
Multi-screen web app	$12,000 to $28,000	4 to 7 weeks	Accounts, saved records, several connected screens, an admin view
SaaS product with billing	$20,000 to $42,000	5 to 8 weeks	Sign-up, plans, recurring charges, per-account data separation
Mobile app for iOS and Android	$18,000 to $40,000	5 to 9 weeks	One shared codebase, both app stores, accounts, notifications
Two-sided marketplace	$30,000 to $55,000	7 to 10 weeks	Two user types, listings, search, messaging, payments in and out

A single-purpose web app that tests one idea

A founder wants to know whether people will pay to have a document checked before they submit it. The first version requires an upload, a queue a reviewer works through by hand, and an email back to the customer. Accounts and billing come later, and the first invoices go out by hand. That is $5,000 to $8,000 and about three weeks. It is deliberately thin, and it answers the only question that matters at that stage, which is whether anybody pays.

A scheduling product for a service business

Customers pick a time, the business confirms, both sides get reminders, and the calendar has to stay in step with the one the owner already lives in. Two user types, one outside connection, a one-time card payment at the point of booking. Around $16,000 to $24,000 and five to six weeks. The calendar sync surprises people, because keeping two calendars honest in both directions is harder than the booking screens it supports.

A SaaS reporting product for one industry

Customers connect their own data source, the product pulls from it nightly, and each account sees charts and exports for its own numbers and nobody else's. Sign-up, three plans, recurring billing, strict separation between accounts. Around $24,000 to $34,000 and seven to nine weeks. Recurring billing and per-account separation account for most of the gap between this and the scheduling product above, and neither is optional once you are charging monthly.

A mobile app with accounts and payments

One codebase, both app stores, sign-in, a profile, in-app purchase and push notifications. Around $22,000 to $32,000 and six to eight weeks, plus a review window at each store that nobody controls. If the idea can be tested in a browser first, do that instead. A web app that works properly on a phone answers the same question for around half the money, and it does not need anybody's approval to ship.

A two-sided marketplace

Two kinds of user, listings, search, messaging between the sides, payment collected from one and paid out to the other, and a set of admin screens to settle disputes. Around $32,000 to $50,000 and eight to ten weeks. Marketplaces are the most expensive first version on this list because nearly every screen gets built twice, and because moving money to third parties brings identity checks and payout rules with it.

See what we have shipped

Reading a quote

Why the same brief gets quotes from $800 to $100,000

Send an identical brief to twenty firms and the spread you get back is not noise. Each number is a different business model quoting a different scope, and once you can tell which is which, the spread stops being confusing and starts being useful.

How to read development quotes by price band
Quote	Who sends it	What you are actually buying
$800 to $3,000	Solo overseas freelancer or a template shop	A theme with your logo on it, or a first milestone priced to win the job before the change orders start
$4,000 to $12,000	A competent small team on a scope that really is small	A working but narrow first version, if the scope on paper is honestly that narrow
$15,000 to $60,000	A small onshore studio quoting against a written specification	A working product, source code, documentation, and one team that answers for all of it
$80,000 to $250,000	A mid-size or large agency	The same product plus account managers, a project manager, a designer and a sales margin

The bottom of that table is where founders lose the most money, and it rarely looks like a scam at the time. A $2,000 quote against a brief a serious team priced at $25,000 buys a different scope wearing the same words. What tends to happen is that the first milestone looks fine, mockups arrive quickly, and then the price starts moving as each unstated requirement surfaces. By the time the project stops, the total has often passed what the honest quote would have been, and there is no working product to show for it.

The top of the table is not a rip-off either. Large agencies do carry the overhead they charge for, and if you need three designers and a compliance review, that is where you go. Most first versions do not need it.

Three questions separate a serious quote from a hopeful one. Ask what happens to the price if a requirement changes, and listen for whether there is a written change process or a shrug. Ask who owns the code, and get the answer in writing. Then ask to see the specification the number was built from, because a firm that priced from a written scope can show you the scope, and a firm that guessed cannot.

One more test that costs nothing. Ask what they would cut to get to two thirds of the price. A team that knows the domain will answer immediately and specifically. A team that is guessing will offer to reduce the quality instead of the scope.

Timeline

How long it takes to build an MVP

Most first versions we build take 4 to 10 weeks from kickoff to a product real users can sign into. Scope size sets that number, and so does how fast decisions get made on your side. Very little else does. Adding people to the team does not compress it, and any firm offering to halve the timeline for more money is selling you a schedule it cannot hold.

Where the weeks go in an MVP build
Phase	Typical length	What happens
Scoping and written specification	3 to 7 days	Every screen, every rule and every user type written down and signed off
Design	0 to 2 weeks	None at all if standard patterns work, up to two weeks for original screens
Build	2 to 7 weeks	The bulk of the schedule, with something you can click on at the end of every week
Testing and fixes	3 to 5 days	Runs alongside the last stretch of the build rather than after it
Launch and handover	2 to 4 days	Deployment, source code, documentation and a walkthrough with whoever takes it over
Warranty	30 days	Anything that breaks gets fixed at no cost

Timelines stretch for reasons that have little to do with the feature list. Slow feedback is the biggest one. If a review that requires two days takes two weeks, the calendar absorbs every one of those extra days and no amount of engineering wins them back. Outside systems you do not control cost time too, including app store review and any partner whose API you need access to. Then there is scope that keeps moving. Every change is small on its own, and every change resets something already finished.

If a hard date matters more than the feature list, say so at the start. Cutting scope to protect a date is straightforward when it happens during scoping and expensive when it happens in week six.

Budget

What to cut to hit a budget

Almost every founder we talk to has a budget smaller than the product in their head. That is normal, and it is fixable, because a first version exists to answer one question rather than serve every customer you will ever have. Cut scope instead of quality. Underpriced work costs more later, and slow decisions cost more than either.

What to cut first and roughly what it saves
Cut this	Why it is usually safe	Rough saving
Role-based permissions	One shared account type answers the same question at launch	$3,000 to $6,500
Recurring billing	Invoice your first twenty customers by hand and learn what they will pay	$4,000 to $6,500
A native mobile app	A web app that works properly on a phone tests the same idea	$8,000 to $14,000
The third and fourth outside connections	Connect one, handle the rest by hand while volume is low	$3,000 to $6,500
Original design from scratch	Standard patterns are the ones your users already know	$4,000 to $6,500
Self-service sign-up	Create the first accounts yourself and talk to every one of them	$1,500 to $2,500

Four things stay in no matter how tight the number gets. The first is the one core action your product exists for, since without it there is nothing to test. Real data storage matters too, rather than a spreadsheet behind the scenes, because migrating fake data later costs more than doing it once. Keep whatever proves people will pay, even if that is a manual invoice you send yourself. And deploy on real infrastructure, because a demo running on somebody's laptop cannot go in front of a customer.

If the smallest honest version of your product still comes back above your budget, the answer is usually to split the build rather than shrink every part of it. Ship the half that proves the idea, get real usage, then price the second half against what you learned. We do this often, and it is why our projects start at $4,000 rather than at a number that assumes you have already raised.

Price bands

What you get for the money at each band

A range only helps if you know what sits inside it. Here is what fits at each level, assuming a fixed price and the work done in the United States.

What is realistic at each MVP price band
Price band	Build time	What is realistic at this level
$4,000 to $9,000	2 to 4 weeks	One core action, one user type, real data, deployed, no billing
$10,000 to $18,000	4 to 6 weeks	Accounts, several connected screens, an admin view, one outside connection
$19,000 to $35,000	5 to 8 weeks	Recurring billing, or a mobile app, or several roles. Pick two of the three
$36,000 to $60,000	7 to 10 weeks	A marketplace, or a subscription product with roles and several connections

What does not change between bands is worth naming, because that is where a low quote actually differs. Every band above includes a written specification you approve before work starts, weekly progress you can click on, full source code and infrastructure access at handover, documentation good enough for the next engineer, and a 30-day warranty. If a quote you are comparing leaves any of those out, it is not the same product at a lower price.

FAQ

Frequently asked questions

Is $4,000 really enough to build an MVP?

It is enough for a narrow one, which is often the right first move. At that level you get one core action, one kind of user, real data storage and a live deployment, with anything that can be done by hand left as a manual step. It will not have billing, several roles or a mobile app. It will tell you whether anybody wants the thing.

Why is an offshore quote a third of the price?

Sometimes the cost base really is lower, and there are good teams working that way. More often the low number is a different scope wearing the same words, priced to win the job before the change orders arrive. The way to tell is to ask both firms for the written specification behind the number. One of them will have one.

What does the fixed price include?

A written specification you approve before anybody writes code, the build itself, testing, deployment on real infrastructure, full source code, documentation, a handover walkthrough and a 30-day warranty. Every hour sits inside the fixed price, and the only invoices you ever see are ones you already agreed to. If a new requirement comes up mid-build, we price it separately and you decide whether to add it.

What if the estimate is more than my budget?

Then we cut scope rather than quality. In practice that means shipping the half that proves the idea, putting it in front of real users, and pricing the second half against what you actually learn. Most of the time the second half looks different from what was planned, which is the whole argument for splitting it.

How accurate is the calculator on this page?

It is built from the same arithmetic we use to price real projects, so it lands close for products that fit one of the shapes it offers. It cannot know about the one unusual requirement that changes everything, and those exist. Treat it as a range to plan against and the starting point for a proper conversation.

How long does it take to build an MVP?

Four to ten weeks for most first versions, from kickoff to something users can sign into. A narrow single-purpose product can be live in two to four weeks. A marketplace or a subscription product with several roles runs to eight or ten. Slow feedback on your side stretches every one of those numbers.

Do you take equity instead of payment?

No. We charge a fixed price, and you keep all of your company and all of your code. Equity pulls a build partner toward your next fundraise instead of toward the product working, which is the wrong incentive on a first version. A fixed price keeps the relationship simple and the ownership entirely yours.

Get a firm number for what you are building

Thirty minutes on a call and we will tell you straight whether we can help, what we would cut, and what the fixed price looks like. Nobody pitches you, and there is no obligation either way.

Book a free 30-minute call

Prefer email? Reach us at [email protected]

Projects start at $4,000. 30-day post-launch warranty included.