How Many Prototypes Are Needed Before We Get 'Perfection'

Some people expect a perfect product that’s ready for manufacturing after just one round of prototyping. Is this realistic, though? Let’s explore how many prototypes might be required, why different rounds of prototyping are undertaken during the new product introduction process, and how many samples of each iteration might be needed for testing to get to the bottom of prototyping here…


What is perfection when it comes to prototypes?

If you want the perfect prototype, you’re looking for a prototype that looks and works as expected with the expected Color Material and Finish. In some cases, the prototype is already what is known as the “golden sample(s)” which is the ideal representation of our mass-produced product made using the same materials, components, and production processes. But, when tooling still needs to be made (and some of the parts made with tooling will look a bit different), or in certain other situations, the “golden samples” can only be prepared later.

As our head of New Product Development, Andrew wrote in this post:

Golden samples are the culmination of a lot of product development work and several rounds of validation and product builds. For instance, we might make 5,000 units and have all kinds of problems, make modifications and fixes and build another 5,000 in a second run. In this run, the products are 90+% fine and meet the desired specification, quality, performance, and testing requirements; therefore can be considered golden samples.

So, your prototypes won’t all be ‘perfect!’ They don’t need to be.

Early prototypes are created to prove your concept…can your product idea even come together into a physical product, however rough? Later on, different iterations of the prototype are made, each fixing issues found and getting closer to a final version, or golden sample, that we can ‘lock in’ and mass produce.

According to Innovative Design Products, prototyping is a process, not an end result, and we’re inclined to agree with them:

Prototyping is how you test your engineering, improve and ultimately perfect your product design prior to manufacturing.

Simple products go through at least 2 rounds of prototyping, and complex products may need multiple rounds of prototyping and testing.

So, let’s see prototyping as a journey, rather than a destination.


Why do we need multiple prototypes?

The different rounds of prototypes all cost money, so it’s understandable that some SMEs can be surprised at the amount of cash new product development hoovers up (excuse the pun) to create different prototype iterations and test them before locking in the product design and preparing for mass production. A lot of testing is done on prototypes if a manufacturer is responsible and follows a suitable NPI process.

Typically many prototypes are created for tests:

  • To assure that the product works and looks like it needs to
  • That show that the product can be made with the correct quality
  • To assess that it will be reliable for at least the length of the warranty period, or longer depending on product type and cost
  • To assure that the product is safe and compliant with your market regulations
  • That check whether the product will be damaged during transit

The tests will vary, but oftentimes they come from a certification specific to your product type, although your supplier’s test engineers may well devise a custom test plan in many cases, too, depending on your requirements.

You can see the types and scale of testing done during new product introduction in this graphic:


Pay particular attention to the prototyping, EVT, DVT, and PVT sections as this is where the majority of the testing occurs.

Some more points from John Teele that are related to the NPI process you can see above:

Feasibility study / POC

“A POC prototype rarely functions exactly like the final product, and it will never look like the final product. It has only one goal – to prove the fundamental concept of the product at the lowest cost possible.”

“A proof-of-concept prototype is usually only used to determine the practicality of a new product idea. It will rarely be seen by customers.”

Prototyping / EVT

“A common strategy is to separate the appearance and feel of your product from the functionality. These are called looks-like prototypes and works-like prototypes.”

An engineering prototype (also sometimes called a works-like-looks-like prototype) is the first time that appearance and functionality come together in a single prototype.”

“This is a works-like-looks-like prototype that has been optimized for manufacturing. This is very close to the final product your customers will see. In most cases, it should also include the retail package if the product will be sold via retail outlets. Although the pre-production prototype may look and function very similar to the works-like-looks-like prototype, the key difference is manufacturability.”

Tooling / DVT

“The Design Validation Test (DVT) is one of the most complex stages. It’s goal is to ensure the product meets any necessary cosmetic and environmental specifications.

A significantly larger number of units will be needed than for the EVT stage, typically 50-200 units.”


Can we skip some of the testing in order to save money and get to market faster?

Yes, of course. But there’s too much risk involved to do this.

Our resident reliability testing guru would agree that skipping tests is a terrible idea, and unfortunately, reliability testing seems to be one that can be skipped or overlooked by importers quite often.

Reliability testing is done “to get the product to function in extreme conditions (temperature, humidity, vibrations…), in order to see if/when it fails. Ideally, the tests should mimic the normal use of the product.”

Even the largest companies can and have suffered catastrophic failures due to insufficient reliability testing.

These high-profile examples demonstrate how important product testing is and why it should not be skipped in the name of getting to market faster.


How many prototypes will we need to make?

It isn’t a case of deciding on the product design and making one prototype version that we then put through tests before settling on one that is ready for manufacturing.

As one iteration of the prototype product is tested, issues will be found and corrected, resulting in a second iteration, and so on, until we’re finally satisfied.

Numerous samples per prototype iteration will also be made. By necessity, many tests will damage or destroy the prototype as the product design’s limits are tested (very commonly when testing reliability, of course).

As I wrote here: “Very seldom are we able to get to an approved look-alike and work-alike prototype in 1 round. For simple products, there may be a need for 2 or 3 iterative rounds, and for more complex products it is sometimes more than 4 or 5 rounds.

The more innovative the product design is, and the higher the requirements for the finish, the more rounds will be needed. Large companies like Apple go through dozens of rounds before they launch the first version of a new product (and dozens more on its packaging).”

An electronic design expert, John Teele, wrote regarding electronic devices: “On average I would expect to need at least 3 prototype iterations. This is true for both the product’s enclosure and the electronics. More advanced products may require as many as a dozen prototype iterations in order to get everything just right.” 

Konstantin Dolger of LA NPDT echoes this. He suggests that: “On average, our clients need 2 to 4 prototypes.”

Examples of the number of prototype samples required for testing

As James Dyson, British inventor of the wildly popular Dyson vacuum cleaners and other consumer products, says in this interview:

Folklore depicts invention as a flash of brilliance. That eureka moment. But it rarely is, I’m afraid. It is more about failure than ultimate success.

After four years of building and testing 5,127 hand-made prototypes of my cyclonic vacuum cleaner, I finally cracked it.

Most clients who work with us here at Sofeast expect their new product to get from design to launch in under four years, and that’s absolutely reasonable! Dyson’s slow route to a working prototype that spawned many successful products and a business that posted US$7.6 billion in revenue in 2019 is unusual. But what his many failed prototypes tell us is that expectations about how many prototypes we need before we can be satisfied to carry on to mass production need to be tempered. One probably isn’t going to cut it at all.

We gave this example of how many sample prototypes might be created for electronic devices, for example, on Agilian Technology’s blog:

Here’s an example (based on a real case) of the sample prototypes needed for the different tests of an electronic device to be imported into the USA to give you an idea of what’s required:

  • 8 units for FCC/CEE (ionizing radiation/energy efficiency)
  • 2 units for ESD (electrostatic discharge)
  • 23 units for various reliability tests
  • 5 additional samples to be held in reserve in case they’re required for reliability testing and ESD.
  • 3 boxes full of finished products for ISTA-2a transportation testing (packaging reliability testing) which will drop, vibrate, and shock the whole box

So that’s a total of 38 sample prototypes plus however many are inside three boxes.



To expect perfection after creating one prototype iteration of your product design is to misunderstand the role of prototyping. Prototyping is a process we follow to get to a prototype that looks and works as expected with the expected CMF. We may go through several rounds of validation and product builds after the ‘first prototype’ is made, and each round will consist of many prototype samples which are used for testing. Over time issues are found and fixed per build until we’re satisfied and can lock the design, then turning our sights to mass production.

This is the NPI process and it is more complex than some importers first imagine.

Do you need help bringing your new product to market? We can help you.

About Renaud Anjoran

Our founder and CEO, Renaud Anjoran, is a recognised expert in quality, reliability, and supply chain issues. He is also an ASQ-Certified ‘Quality Engineer’, ‘Reliability Engineer’, and ‘Quality Manager’, and a certified ISO 9001, 13485, and 14001 Lead Auditor.

His key experiences are in electronics, textiles, plastic injection, die casting, eyewear, furniture, oil & gas, and paint.

This entry was posted in New Product Launch and tagged , , , , . Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *