How much to charge for an MVP
How much to charge for an MVP comes down to one question: what decision is this build supposed to unlock, and what is that decision worth to the founder? Nobody wants an MVP. They want to know whether to raise, whether strangers will pay, or whether a process they run by hand deserves software. That decision has a price, and the founder usually knows it, because it is the cheque they are chasing or the cost they are carrying. Quote a fixed fee that is an obvious fraction of that number, built around one core loop, in three tiers. Do not quote by feature, and never by hour. Claude Code can get a working loop in front of users in days, and if you price by the day, the founder will pay you like a day.
This is the CCC Decision-Value Method, and it exists because the MVP is the job where new builders lose the most money in the most flattering way. The founder is excited, the idea is interesting, and the scope has no edges. So the quote comes out as a features list, the features multiply, and six weeks later you are a technical co-founder on a contractor's invoice. Priced against the decision, an MVP is a clean fixed-scope build with a high margin and the most natural retainer upsell in the business, because a founder who gets their answer needs the next thing built immediately.
Find out what the MVP has to prove before you say a number
You cannot price an MVP until you know what it is for, so the first call is a measurement. Five questions get you there, and I ask them in this order because each one makes the next one easier to answer honestly.
- What happens if this works? A pitch to investors, a launch to a waiting list, a pilot with one named customer. That is the decision, and it is the thing you are actually building.
- What is that worth to you? Let the founder say it. A raise has a target. A first customer has a contract value. A workflow has a monthly cost in someone's salary. Your quote is going to sit next to this number, so get it from them.
- Who is the first user, by name if possible? If the answer is "everyone," the MVP is not defined yet and the price has to include defining it. Paid discovery covers this; [paid discovery for AI projects](/blog/paid-discovery-phase-ai-projects) is how I sell that week.
- When is the deadline, and what set it? A demo day, an investor meeting, or a customer's renewal date is a real deadline and it is worth money. "As soon as possible" is not a deadline and does not change the quote.
- What have you spent so far? A founder who has already paid an agency for a deck and a designer for screens has a sunk cost and a mental price for the next step. You want to know both before you name yours.
Say the answers back before you quote. "So this exists to close your first three pilot customers, and each pilot is worth roughly this to you" is a sentence the founder has to agree with, and once they have, the price of the build is a fraction of a number they already accepted. The full question set I run on every deal is in [client discovery questions before you quote](/blog/client-discovery-questions-before-you-quote).
Define the MVP on paper as one loop, or you will build the whole product
The CCC One-Loop Rule: an MVP is one user, doing one thing, seeing one result, and that loop written down in three lines is the scope. A user signs up, uploads a contract, gets a summary. A customer books a slot, pays, receives a confirmation. If the founder cannot say the loop in a sentence, the product is not ready to be built and the price should cover the discovery that gets it there. If they can, everything outside that sentence is phase two, priced separately, and written on the proposal as phase two so nobody is surprised.
This rule is what makes the fixed fee safe. Claude Code will happily add an admin panel, a settings page, a second user role, and a referral system in the same afternoon, and the founder will happily ask for them, because from where they sit every one of those is "part of the MVP." The written loop is the line you point at. [How to scope a Claude Code client project](/blog/how-to-scope-a-claude-code-client-project) is the longer method, and [how to write a project brief for Claude Code](/blog/how-to-write-a-project-brief-for-claude-code) turns the loop into the brief you actually build from.
The CCC MVP Price Ladder: three tiers
Present three tiers, all built around the same loop, and let the founder pick how much of the decision they want proven. The bottom tier is for a founder who needs something to show. The middle is for a founder who needs something to sell. The top is for a founder who already knows the product is real and wants the first weeks of learning built in.
The CCC MVP Price Ladder
| Tier | What the founder gets | How to price it |
|---|---|---|
| Prove | The core loop working end to end with real data, deployed to a private URL, good enough to demo to an investor or a pilot customer. No payments, minimal auth, no polish outside the loop | A fixed fee that is a small, obvious fraction of the decision's value. Priced so a funded founder says yes on the call |
| Launch | Everything in Prove, plus real sign-up and login, payments wired to their account, a production deploy on their domain, and the basic analytics to see whether strangers use it | The anchor tier. Priced against the first customers or the raise it is meant to unlock. Most founders land here |
| Traction | Everything in Launch, plus a fixed post-launch window where you ship changes weekly against what real users do, and a handover of the codebase and accounts at the end | Priced for the founder who has already decided. This tier is what becomes the retainer |
The middle tier wins because of payments and the deploy. A loop on a private URL proves the idea. A loop that a stranger can pay for proves the business, and that is the decision most founders are actually trying to make. Put the difference in the table and they will choose it themselves. The three-tier shape is the same one in [how much to charge for a landing page](/blog/how-much-to-charge-for-a-landing-page) and [how much to charge for a chatbot](/blog/how-much-to-charge-for-a-chatbot); the tiers change, the logic does not.
What moves an MVP quote up or down
- Payments. A loop that takes money is a business, and the price reflects it. Subscriptions, one-off charges, and marketplaces with payouts are three different amounts of work, so ask which one before you quote.
- Who the users are. One kind of user is an MVP. Two kinds who see different things, a customer and an admin, is two products sharing a database. Price the second role as its own line.
- Integrations. Every outside system the loop touches, a calendar, a CRM, an email tool, a data source, is its own line because each one is its own thing that can break on launch day.
- Design. A founder with brand assets and screens from a designer is a build. A founder with a napkin is a build plus a design pass they will want to revisit, and that pass is priced.
- Web or native mobile. A web app that works on a phone is the default MVP. A native app with store review is a different timeline and a different quote, and most MVPs do not need it yet.
- Data that already exists. If the founder has a spreadsheet of customers or a legacy tool to migrate from, the migration is a line item, not a favour.
- The deadline. A demo day in three weeks is worth more than the same build with no date on it, and the founder already knows it.
- Account ownership. Build on the founder's accounts, under their billing. It goes on the quote, and it is why you are never the hosting company when the startup runs out of money.
The mistakes that turn a paid MVP into an unpaid co-founder job
- Taking equity instead of a fee. The founder will offer, because the equity costs them nothing today. Unless you would invest cash in this company, take cash for the build. If you genuinely would, take a fee and negotiate the equity as a separate conversation.
- Quoting hours or days. "About a week with Claude Code" becomes the price ceiling the moment you say it. Talk about the decision and the loop, nothing else. [Why I stopped charging hourly for AI builds](/blog/stopped-charging-hourly-for-ai-builds) is the longer argument.
- Starting without a deposit. Startups run out of money, and the ones that do not still have a board meeting to survive. [How much deposit to charge before an AI project](/blog/how-much-deposit-to-charge-before-an-ai-project) covers the number; for an MVP it is at the high end.
- Unlimited pivots. The founder will learn something on day four and want a different loop. That is healthy, and it is a change order. Write down what a pivot costs before the build starts. [How to handle scope creep on client projects](/blog/how-to-handle-scope-creep-on-client-projects) has the script.
- Building the roadmap. A founder's pitch deck has eighteen months of features on one slide. You are building the loop. Everything else is phase two with a price.
- No end date. An MVP without a handover date becomes a job. Even the Traction tier has a fixed window, and the retainer that follows is a new agreement, not a continuation.
How to present the MVP number so it closes
Lead with the decision you agreed on, then put the three tiers under it. "You said this exists to close your first pilot customers, and we agreed what a pilot is worth; here are three ways to prove it" is a sentence a founder can say yes to on the call, and it is a sentence they can repeat to a co-founder or an investor an hour later without you in the room. Send it as [a one-page proposal](/blog/one-page-ai-proposal-that-closes) with the loop written in three lines, the tier scope, the phase-two list, the deposit, and the handover date. [How to write a statement of work for an AI project](/blog/how-to-write-a-statement-of-work-for-an-ai-project) is the version I attach when the founder has a lawyer, and [what to put in an AI project contract](/blog/what-to-put-in-an-ai-project-contract) covers who owns the code, which for an MVP is always the founder once the invoice is paid.
The push-back you will hear is "a no-code tool could do this" or "a developer friend said it would take a weekend." Both are answered by the decision: a weekend build that the friend abandons in week three does not close a pilot customer or survive an investor's technical question. If the objection is really about budget, the honest answer is often the Prove tier, and [the "can you do it cheaper" objection](/blog/can-you-do-it-cheaper-objection) has the exact words.
Quote your next MVP against the decision, not the feature list
An MVP is the best client project in this business when it is priced right and the worst when it is not. Get the decision and its value from the founder, write the loop in three lines, quote the ladder, take the deposit, and put the handover date on the proposal. The founder who gets their answer is the easiest retainer you will ever sell; [how to turn a one-off client into a monthly retainer](/blog/how-to-turn-a-one-off-client-into-a-monthly-retainer) is what that conversation looks like.
Short, practical drops on skills, MCP, agents, prompts, and more. No spam, unsubscribe anytime.
Frequently asked questions
How much should I charge for an MVP?
Price it against the decision the MVP unlocks, not the features or the hours. Ask the founder what happens if it works, a raise, first paying customers, or proof that a workflow is worth software, and what that is worth to them. Then quote a fixed fee that is an obvious fraction of that value, in three tiers built around one core loop: Prove, Launch, and Traction.
Should I take equity for building an MVP?
Take a fee for the build. Equity costs the founder nothing today, and most MVPs do not become companies. If you would genuinely invest your own cash in the business, take the fee anyway and negotiate equity as a separate conversation with its own paperwork.
What should be included in an MVP?
One core loop: a user signs up, does the one thing the product exists for, and sees the result. Write it in three lines and that is the scope. Real sign-up, payments, a production deploy, and basic analytics belong in the Launch tier; a second user role, an admin panel, integrations, and native mobile are phase two with their own price.
How much deposit should I charge for an MVP?
The high end of whatever you normally charge, taken before any work starts. Startups run out of money, and the ones that do not still change direction. A deposit plus a written change-order price for pivots is what keeps an MVP a fixed-scope build instead of an open-ended job.
What if the founder says a no-code tool or a developer friend could do it cheaper?
Go back to the decision. A weekend build that gets abandoned in week three does not close a pilot customer or survive an investor's technical questions. If the objection is really about budget, offer the Prove tier: the core loop working end to end on a private URL, good enough to demo, with payments and launch as a priced next step.
Last reviewed by Duncan Rogoff on September 27, 2026
