BuildThe 10 Prototypes Rule. Shows in emojis how you first draft a prototype. Then it fails, you examine and then redo it.

What’s The 10 Prototypes Rule?

So, you want to build, market and sell your own products. Creating your first prototype is probably not new to you. But make no mistake. Just creating a single prototype is not going to cut it. In fact, you should create at least 10 prototypes before releasing your product.

The value of a prototype

Why are we even creating prototypes? To build a product, we need to be creative. Often, we also need to solve many engineering problems. In both cases, it’s hard to plan or predict how something will work out in the end. So, often you can only try to take the best approach to solve a problem and go for it. But, building a product in low volumes is very expensive. Tooling costs are high and ordering a small number of components is expensive.

To sum it up, a prototype proves if a solution is a good fit for a problem. Additionally, a prototype must be cheap, easy and fast to build.

Prototyping gone wrong 🔥

When I started the development of Moment-Rec, I wanted to create a prototype. I started with a development kit for the microcontroller I had chosen. Soon, I realized that it would be a pain to connect all sensors and peripherals to the kit. So, I started to develop a Moment-Rec PCB and write embedded software for it. It took me 7 months to complete this prototype.

Unfortunately, I was too busy solving engineering problems. So, I didn’t notice that I was not prototyping right.

The curse of an engineer’s mind 😈

When I started Moment-rec, I had just left a job as a Hardware Designer. So, naturally my mind was focused on solving engineering problems. It’s no surprise that I kept going with this mindset. Instead, I should have adopted the mindset of a product designer. As a product designer you start by making sure your product serves a purpose. Testing the environment and situation a customer would use it in. You also research what your target audience is worrying about.

Prototyping done right ✔️

In a nutshell, here is how to really prototype a product.

  1. Test if the product solves the problem
  2. Test if it blends with the environment/situation
  3. Work out engineering problems

The most important question during prototyping is: How can I solve the problem? Every product tries to solve a problem. If we’re not able to do that, then our product is garbage. Second, the product needs to blend in with the environment and be fitted to the user.

  • A sports gadget must be able to handle drops
  • A hearing aid should be hidden and fit the user
  • A smart lock must be reliable and robust

Once you solved the first two issues, only then you should worry about how to engineer it. In the end, engineering problems are just a question of how much money you are willing to pay.

The 10 Prototypes Rule 📜

After I realized my mistake, I came up with “The 10 Prototypes Rule”. It’s how to avoid the engineer’s mindset. The idea is to at least do 5-6 prototypes before starting to worry about engineering problems. Solving engineering problems requires much effort, time and is expensive. So it’s best to leave those tasks until the end, when we’re sure we got everything else right. First, we should focus on solution, function, features and environment. We have to iterate through various prototypes until the non-engineering criteria are met.

Exponential curve explaining the 10 Prototype rule.

It’s best to start with the simplest and quickest prototype build possible. Ideally, it shouldn’t take more than 5 minutes. Start with pen, paper and a bit of tape. From this early prototype we can gradually increase the complexity with every iteration. We should ignore engineering problem areas as often as possible and trick, cheat and hack. So, you should use finished hardware solutions such as the Raspberry Pi, Arduino and Kits. Use simple and high-level languages that mask the problem. Don’t optimize the prototype for size, performance, or efficiency. In fact, we’re trying make the prototype more effective at solving the problem.

The Engineering Prototypes ⚙️

Once we completed between 5-6 prototypes, we can get into engineering prototypes. Here we’ll do exactly the opposite. We try to minimize resource usage, price and size. We have to try and make it work with existing technologies. It’s a fight to comply to requirements we set during the non-engineering prototype phase.

Depending on how much experience, man power and time you have, you could also outsource this step.


The 10 Prototypes Rule guarantees that you gather enough knowledge to write good specification. It will also prevent many expensive engineering iterations. Finally, the 10 Prototypes Rule will help you with crowdfunding. If you plan to outsource engineering to a company, then you’ll need funding. All non-engineering prototypes you created are great to show off in your campaign.

Find out more about how to build, market and sell your own products. Join our forum.

Notable Replies

  1. Yes, I have certainly fallen into the traps of prototyping the wrong way. Its so important to use tech that exists and stick it together for prototype 1. I have made the bad mistake of jumping ahead and wasting a lot of time on what was really the end product and not on the concept - ie the product designer as you say.

    Once you have a quick and dirty prototype that proves the concept (ie the point where you start to look at proper investment or product approval), then you can start in the engineering / coding detail and decisions. You will have learned a lot by then and have a much better idea of how to proceed.

    I can think of one project where I could have gotten a prototype up and running using arduino , off the shelf GPS modules etc in a few weeks. I ended up coding my own stuff from scratch and designing hardware that took way too long. (along with more mistakes that was necessary)

  2. It was exactly the same for me with Moment-rec. I basically jumped ahead and tried to produce the end product in one go :yum:
    Do you have much experience with Arduino?

    I think in the future I probably often use Raspberry Pi for early prototypes. The Raspberry Pi Zero W is very compact yet powerful. The advantage in my opinion is, that you can run python on it. It’s a very high-level language with tons of libraries available. I could even run the program on my computer first (maybe mocking some hardware) and then move the code to the raspberry Pi + connect actual hardware.

    Another interesting aspect for me is that python can run many AI algorithms. So, if AI becomes a huge thing in embedded, then I’d be pretty well prepared with a Raspberry Pi + python setup.

    Do you see any disadvantage in using Raspberry Pi instead of Arduino? Or would you just use the platform that is closest to a solution?

  3. I have to admit that I have only recently bought and played with Arduino. I had this stupid thought about how it was not ‘real’ (ie all done for you … must do it the hard way!)… but… wow - what a useful platform for getting things done quickly. Especially with prototyping and being able to test sensors and concepts etc before you plug them into other processors or projects.

    I have a few Raspberry Pi’s - On my to-do list is to get use them more. Python does seem to be a big player. Not used it much yet, but again on my to-do list. Its a very useful platform and as you say the Pi Zero starts to become a really good base for simpler projects.

    Its hard to compare Pi and Arduino. They are so different. I would use Arduino for quick prototyping but probably nothing esle (at this stage). Raspberry Pi for more complicated projects that need TCP/IP, Display etc. Also more useful for AI and data processing. For a hobbyist I think the Pi is better if you are a top level programmer and have little or no embedded knowledge.

    I still see Arduino as a play thing and a (very) useful prototyping tool. I’m not sure I would use it in any production product.

    Do people use Pi and Arduino in production? I guess so. I did have a client who was looking at a monitoring system and wanted to use the Pi - it makes sense from a production point of view as you just don’t have to worry about 95% of the electronics production and also its a known device and operation system. Easier to maintain.

    I wonder how much Raspberry Pi is disrupting the bespoke industrial computer controllers?

  4. I’ve actually never used Arduino so far. And mostly for the same reason you hesitated as well. Many people I talked to believe that Arduino is bad, because it hides so much from the user. So, people never learn the hard/real way of doing things. But, as you said, I think there is a lot of value in Arduino.

    I’ve done a couple projects with Raspberry Pi 3, but I’m also not an expert with python. I think you are right that Arduino is best for sensor stuff, when you need to access a lot of hardware. Raspberry Pi is (probably) really good for AI and TCP/IP stuff or very big complicated projects.

    I’ve recently read an article about people who tried to use Arduino in a real product (was for a kickstarter). I think the costs are a big problem. There’s probably quite much markup on Arduino’s so you won’t be very competitive with your product, if you try to sell in large quantities. Especially if it’s a consumer customer. But, I think it’s perfectly ok for industrial customers and also research.

    Raspberry Pi also has industrial versions, with almost all I/Os exposed via an industrial interface.

    I mean Raspberry Pi in general was a huge change for the whole market. It was the start of a new era. A computer that fits on the palm of your hand. It think it also played a huge part in shaping the maker community. It’s incredible what Makers create in their own free-time.

Continue the discussion at