Glossary
What Is a Prototype?
Updated · Team Fonda
Short answer
A prototype is an early, rough model of a product or service that you make to test how it should look and work. It can be a paper sketch, a set of clickable screens or a handmade sample. You show it to possible customers, watch what confuses them and fix that before you pay to build the real thing.
Key takeaways
- A prototype is for learning, not selling. It can be fake behind the scenes.
- Start with the roughest version that lets a person react, then add detail only where you need answers.
- Watching five or six target customers try it usually teaches more than asking many people for opinions.
- It sits between a proof of concept and an MVP.
What a prototype means in plain words
Describing an idea in words leaves a lot to the listener's imagination. A prototype gives people something to look at, click or hold, so their reaction is to the thing itself. It does not need to work properly. A booking app prototype might be six drawn screens linked together with no database behind them. A food product prototype might be a batch made in your own kitchen with a printed label.
The point is speed. A change that takes ten minutes in a sketch can take weeks once the product is built.
Types of prototype, from rough to realistic
Designers talk about fidelity, which means how close the prototype is to the finished product. Lower fidelity is faster and cheaper. Higher fidelity gets more realistic reactions.
| Type | What it is | Good for |
|---|---|---|
| Paper sketch | Screens or product drawn by hand | Checking the basic flow in the first days |
| Wireframe | Plain boxes and labels showing layout, no colors or branding | Agreeing what goes on each screen |
| Clickable mock-up | Realistic screens linked in a design tool, nothing works underneath | Testing whether people can complete a task |
| Concierge or manual version | You deliver the service by hand while it looks automated or finished | Testing a service before building systems |
| Physical sample | A handmade or 3D-printed version of a product | Testing size, feel, taste or packaging |
How to test a prototype
- Pick one task. For example, "book a dog walker for Thursday". A prototype that tries to show everything tests nothing well.
- Find people who match your target customer. Your ideal customer profile tells you who to ask.
- Give the task and stay quiet. Ask them to think out loud. Do not explain or rescue them.
- Write down where they stop, hesitate or guess wrong. Those are your fixes.
- Change the prototype and test again with new people.
Tip
Watch what people do, not what they say. "Looks great" from someone who could not find the button is a fail.
A worked example
This is an illustration with made-up numbers. Say you plan an app where pet owners book local dog walkers. You draw six screens in a free design tool and link them so they can be tapped on a phone. It takes one weekend.
- Round one: you ask 6 dog owners to book a walk for Thursday. Only 2 finish without help. The other 4 look for a calendar on the first screen and cannot find one, because you put the date choice after the walker choice.
- Fix: you move the date to the first screen. One hour of work.
- Round two: 6 new dog owners try it. 5 finish. Three of them ask, unprompted, whether they can see the walker's reviews first.
- What you learned: the booking order was wrong, and reviews matter more than the map view you had planned to build first.
Two rounds, twelve people and no code. Both findings change what goes into the first real version.
Why it matters for a first-time founder
You meet prototypes after you have evidence that the problem is real, which is called problem-solution fit, and before you build a minimum viable product. If part of your idea might not be possible at all, a proof of concept comes first.
A prototype is also useful in customer interviews. People give sharper answers when they can point at something. And if you are not technical, a tested prototype is the clearest brief you can hand to a developer or co-founder, which keeps the build smaller and cheaper. Fonda's Launch phase includes an MVP plan step, which comes after you have talked to real customers in the Validate phase. The guide on how to build an MVP explains that step in more detail.
Common mistakes
- Making it too polished too early. People hold back criticism of something that looks finished, and you become reluctant to change it.
- Testing with friends. They know you and are not the customer. Use strangers who have the problem.
- Explaining as they go. You will not be standing next to real users.
- Testing the whole product. One task per round gives clear answers.
- Reading it as proof of demand. A prototype shows whether people understand and can use the idea. Whether they will pay is a separate test.
Frequently asked questions
What is the difference between a prototype and an MVP?
A prototype is a model used to learn about design and usability, and it may not work at all behind the scenes. An MVP is a working product, however small, that real customers can use and pay for. You test a prototype with a handful of people and release an MVP to the market.
Do I need to know how to code to make a prototype?
No. Paper, slides and free design tools are enough for most early prototypes. For a service, you can run it by hand for a few customers. Code is only needed when the thing you want to learn depends on real data or real performance.
How many people should test a prototype?
A small group per round is usually enough to show the biggest problems, because the same issues come up again and again. Fix those, then test the new version with a fresh group. Several small rounds beat one large one.
Can I show a prototype to investors?
Yes, and it often helps them understand the idea faster than slides. Be clear that it is a prototype, and pair it with what you learned from testing it, since evidence from customers carries more weight than the screens themselves.