Summarize with AI
A small fixed price for MVP design buys one thing done properly: the core flow your product is judged on. Expect the main user journey mapped, ten to twenty key screens designed, a clickable prototype, and the handful of reusable components a developer can build from. It does not buy a full design system, a marketing site, user research, or unlimited rounds. The way to stay inside a small budget is to cut the number of screens, not the quality of the ones you keep.
Most pre-seed founders come to this the same way. You have enough money to pay a studio once, not every month, and you need something a user can actually use. A fixed price answers the scary question first, which is what this will cost. What it does not tell you is what falls inside the line. Below is what usually does, what does not, and how to shrink the scope without shipping something broken.
What a fixed price usually covers
A fixed price is a studio's bet that it can finish a defined piece of work in a defined time. So the scope is written in units the studio can count. Screens, flows, and rounds of feedback.
On a small MVP budget that normally means one primary flow, from the moment someone signs up to the moment they get the result your product promises. Around ten to twenty screens covering that flow and its obvious failure states. A clickable prototype so you can hand a link to a user or an investor instead of a slide. And a small set of components, like buttons, inputs, cards, and a table, so the build does not invent a new style on every screen.
Two or three rounds of feedback are normal. Ask for the number before you sign, because an unstated round count is where fixed prices quietly become hourly ones.
What gets left out, and why that is usually fine
Four things almost always sit outside a small fixed price, and none of them should worry you yet.
A full design system. At your stage you need enough shared parts to stop the build drifting, not a documented library with tokens and rules. That work pays off when three designers are working at once. You have none.
A marketing website. The product and the site are different jobs with different readers. Bundling them into one small budget means both get half the attention. If you need both, do the product first and put up a single clear page while you wait.
Research. Interviews, surveys, and usability testing are real work and cost real money. On a first version, the cheapest research is shipping and watching. Spend the budget on something users can touch.
Every edge case. Settings, billing, admin views, empty states for features you have not built. Sketch them later, once the main flow has survived contact with people.
Design only, or design and build
This is the fork that decides what your money buys, and founders often do not realise they are choosing.
Design only means you get files and a prototype. Someone else writes the code. This is the right call when you have a technical cofounder or a developer already on the team. It is also the cheaper half, so more of the budget goes into thinking.
Design and build means the studio also ships a working front end. It costs more for the same scope, because building is where the hours go. It is the right call when nobody on your side can turn a prototype into a product, and when waiting to hire an engineer would cost you more than the price difference.
The trap is buying design only when you have no one to build it. A beautiful prototype that sits in a folder for four months is the most expensive thing on this page.
How to scope down without shipping something unusable
There is a good way to cut and a bad way. The bad way is to keep every screen and give each one less care. The result looks unfinished everywhere, and users read that as a product that does not work.
The good way is to remove whole features and finish what remains. Pick the one thing your product does that nothing else does, and design that path properly. A product that does one thing well reads as early. A product that does six things roughly reads as broken.
Three cuts that reliably save money and cost almost nothing. Use one account type instead of roles and permissions. Use a single list view instead of a dashboard with charts. Replace anything configurable with a sensible default you can change later in code.
And keep the failure states. Empty screens, loading, and error messages are cheap to design and are what makes a small product feel finished. Skipping them is the most common false economy we see.
What makes a fixed price go wrong
Fixed prices fail for boring reasons, and all of them are visible before you start.
The scope is written in adjectives. If the agreement says "a modern, intuitive MVP" instead of a screen count and a flow, nobody can tell when it is finished. Ask for a list.
The product is still moving. If you are likely to change what you are building halfway through, a fixed price will either stop you or cost extra. A short retainer suits a moving target better.
There is no named decision maker. Fixed prices assume fast answers. If feedback has to travel through four people, the timeline slips and the studio starts charging for the wait.
If you cannot compare two quotes because neither publishes a price, our guide to comparing design agencies without public prices covers the questions that force a real number out.
Where to go next
If you are picking who to work with, our shortlist of MVP design agencies for pre-seed AI startups compares studios on what they publish, including whether they name a starting price. If you also need a site, how fixed price website design works covers the same question for marketing pages, which are scoped differently.
A word on who wrote this. Maydit designs web and product work for AI founders across the US, the UK, and Europe. The building happens in Framer, Webflow, or custom code depending on what the thing has to do, and the same team stays on into product design once a first version is live. Dualite is the result we point to most often: design work supporting a repositioned ICP helped the product reach 100,000+ users in seven months. There are two ways to buy, and for a first version it is nearly always the first. A fixed scope takes three to four weeks and suits a team with a date to hit. A monthly retainer, with no long lock-in, suits a team already shipping every week. If you have a number in your head and want an honest answer about what it covers, book a 30 minute call and bring it with you.
