Stay Tuned for Special Announcements from WRC
According to Predictable Designs:
You won’t need any prior knowledge of electronics design to understand this article. Although, it will be easier to comprehend if you already have a technical background in computers, engineering or another field of science.
I won’t lie to you, though. It’s a long, difficult journey to launch a new physical product. Although hardware is known for being hard, it’s easier than ever for individuals and small teams to develop amazing new electronic devices.
However, if you are looking for an easy, quick way to make money then I suggest you stop reading right now because bringing a new product to market is far from easy or quick.
Although I present each step in a linear fashion, product development is never a smooth linear progression and at times you will find yourself taking two steps back for every step forward. Don’t get too frustrated though when this happens because it’s just part of the process.
Part 1 – Upfront Work
Before we get into the actual product development steps I feel that it’s important to quickly cover some of the upfront work that needs to be done before full product development begins.
Unfortunately, too many people skip past these critical early steps, since most entrepreneurs want to get right into developing their product. Skip these initial steps at your own peril.
Market research, patents and product validation
You need to first research the competition and confirm that a market exists for your product.
When entrepreneurs do competitive research a common misconception is to think that finding any competition is bad. In fact, most entrepreneurs panic and think it’s the end of their project when they find a competing solution.
In reality though, having some competition is actually good because it proves there is a market for your product idea.
Once you’ve determined that a market exists for your product, then the next step is to validate that people will actually buy your product.
For example, one way to validate your new product idea is to make a sales flyer to share with potential customers or retailers that carry similar products. Request their feedback on your tech product.
You can also setup an online sales page to test your idea. Crowdfunding is another great form of validation. The key is you have to share your product idea with lots of unbiased people who will give you honest feedback.
But, what about a patent? Don’t you need one before you can safely tell others about your product?
Being overly secretive is probably the most common mistake made by new entrepreneurs. In reality, I’ve never seen a hardware startup fail because their idea was stolen.
That just does not happen. Instead, entrepreneurs usually fail because they’ve been unsuccessful at marketing their product.
However, I do understand concerns over privacy and why most people want to start with a patent. But patents are really expensive and take a long time to get. They are simply not the best use of your limited time or money when you are first starting out.
Instead, I recommend what is called a Provisional Patent Application (PPA) in the United States. Other regions likely have something similar available.
For only a couple hundred dollars, a PPA will give your product protection for one year. And you won’t need to hire an expensive patent attorney at this stage.
Simplify your product
I spent years developing my own consumer product, and have helped other entrepreneurs develop their products for much longer. One of the most important lessons I’ve learned is that you need to simplify your device as much as possible.
You must embrace simplification in order to have a realistic shot at getting your product to market in a timely fashion and without going bankrupt.
Product complexity can be a death trap for new entrepreneurs and startups!
Most entrepreneurs, and even most engineers, don’t understand all the consequences of various product features. The addition of what seems like a minor feature can often drastically increase your development cost and the time it takes to get to market.
For example, something as simple as the position of a button could waste thousands of dollars, if it creates the need for more expensive injection molds.
Another example comes from an entrepreneur who recently asked me for feedback on his product concept.
He is very technically minded and had spent a lot of time specifying his product in great detail. But he didn’t understand all the future implications of these features.
I quickly explained to him why one feature of his product was going to drastically drive up all of his costs including: development, prototyping, electrical certifications, and mold costs.
I proposed several alternative solutions, and he was off to a much better start. These changes made his product about 3 times easier to bring to market. Overall, he easily saved at least $50,000 with this one change!
The catch is that unless you understand the entire development and manufacturing process it’s difficult to know the future implications of every product decision.
If you outsource product development be aware that asking your primary developer(s) to help simplify your product may create a conflict of interest. This is because the more complex your product the more money they get paid.
Instead, work with independent experts to help you simplify your product and better understand the long-term impacts of your early product decisions.
Proof-of-Concept (POC) prototypes
At this point you have proven that there is a market for your product and you’ve simplified the product concept as much as possible.
Now the question you must answer is whether your concept really solves the intended problem as expected.
This is the goal of a Proof-of-Concept (POC) prototype which is an early prototype created using off-the-shelf components.
A POC prototype doesn’t have any custom electronics design, and is typically based on development kits such as an Arduino or Raspberry Pi.
I’ve already stressed the importance of not starting right way with full development, but a POC prototype is different because it’s not usually very expensive to build.
Unfortunately, a POC prototype is rarely something that can be brought to market. The production cost will be too much, the physical size too large, and the appearance far from ideal.
Part 2 – Product Development Strategies
There are essentially five options for entrepreneurs and startups to develop a new hardware product. However, many times the best overall strategy is a combination of these five development strategies.
If you don’t have the technical knowledge to properly manage product development and to judge the quality of work being done, then having a technical advisor is a must regardless of which of these development strategies you follow.
This technical advisor can help you make key technical decisions and provide oversight of your primary developers. Even if you are an experienced engineer it’s highly beneficial to have an advisor who understands the entire process from idea to market.
1) Develop the Product Yourself
This is rarely a viable strategy completely by itself. Very few people have all of the skills needed to develop a market-ready electronic product completely on their own.
Even if you happen to be an engineer, are you an expert in electronics design, programming, 3D modeling, injection molding, and manufacturing? Probably not. Also most of these specialties are made up of numerous sub-specialties.
That being said, if you have the necessary skills, the farther you take the development of your product yourself the more money you will save.
For example, I brought my own hardware product to market about 6 years ago. The product was more complex mechanically than it was electrically. I’m an electronics engineer by training and not a mechanical engineer, so I initially hired a couple freelance mechanical engineers.
However, I quickly became frustrated with how slow things progressed. After all, I was thinking about my product almost every waking hour. I was obsessed with getting my product developed and on the market as fast as possible. Sound familiar?
But the engineers I hired were juggling it with lots of other projects and not giving my project the attention I felt it deserved.
So I decided to learn everything needed to do the mechanical design myself. No one was more motivated than myself to get my product developed and on the market. Ultimately, I was able to complete the mechanical design much faster (and for much less money).
The moral of the story is to do as much of the development as your skills allow, but also don’t take that too far. If your sub-expert skills cause you to develop a less than optimal product then that is a big mistake.
Also, any new skills you must learn will take time and that may ultimately lengthen the time to market. Always bring in experts to fill in any gaps in your expertise, but be sure to remain involved in all key decisions.
2) Bring on Technical Co-Founder(s)
If you are a non-technical founder then you would definitely be wise to bring on a technical co-founder. One of the founders on your startup team needs to at the very least understand enough about product development to manage the process.
If you plan to eventually seek outside funding from professional investors then you definitely need a team of founders. Professional startup investors know that a team of founders is much more likely to succeed than a solo founder.
The ideal co-founder team for most hardware startups is a hardware engineer, a programmer, and a marketer.
Bringing on co-founders may sound like the perfect solution to your problems, but there are some serious downsides as well. First of all, finding co-founders is difficult and will likely take a tremendous amount of time. That is valuable time that isn’t being spent developing your product.
Finding co-founders is not something you should rush and you need to take time to find the right match.
Not only do they need to compliment your skills, but you also really need to like them personally. You are essentially going to be married to them for at least a few years so be sure you get along well.
The major downside of bringing on co-founders is they reduce your equity in the company. All founders of a company should really have equal equity in the company. So if you are going solo right now, be prepared to give any co-founder half of your company.
3) Outsource to Freelance Engineers
One of the best ways to fill in any gaps in your teams technical ability is by outsourcing to freelance engineers.
Just keep in mind that most products will require multiple engineers of different specialties so you will need to manage the various engineers yourself. Ultimately, someone on the founding team will need to serve as the project manager.
Make sure you find an electrical engineer that has experience designing the type of electronics required by your product. Electrical engineering is a huge field of study and many engineers lack any experience with circuit design.
For the 3D designer make sure you find someone that has experience with injection molding technology, otherwise you’re likely to end up with a product that can be prototyped but not mass manufactured.
4) Outsource to a Development Firm
The best known product design firms such as Frog, IDEO, Fuse Project, etc. can generate fantastic product designs, but they’re insanely expensive.
Startups should avoid big design firms at all costs. Top design firms can charge $500k+ to fully develop your new product.
Even if you can afford to hire an expensive product development firm, don’t do it!
However, there are many smaller, less known design firms that I’ve worked with that can develop your product at a much more reasonable price. The cost for these smaller firms usually falls closer to the cost of hiring freelancers yet with more oversight and better quality procedures in place.
One final comment on working with product design firms. You must discard the false belief that you can just have someone else handle all of the development while you just wait on the sidelines.
Those that attempt that path almost always get burned and lose a lot of money and time. You absolutely need to be involved in development and understand at least the basics of the development process as well as the various design tradeoffs being made.
5) Partner with a Manufacturer
One avenue to pursue is partnering with an overseas manufacturer that already makes products that are similar to your product.
Large manufacturers will have their own engineering and development departments to work on their own products. If you can find a manufacturer already making something similar to your own product, they may be able to do everything for you – development, engineering, prototyping, mold production and manufacturing.
This strategy can lower your upfront development costs. Manufacturers will, however, amortize these costs, which means adding an additional cost per product for the first production runs. This essentially works like an interest free loan, allowing you to slowly pay back your development costs to the manufacturer.
Sounds great and easy, so what’s the catch? The main risk to consider with this strategy is you are putting everything related to your product into a single company.
They will surely want an exclusive manufacturing agreement, at least until their costs have been recovered. This means you can’t migrate to a cheaper manufacturing option when your production volume increases.
Also be warned that many manufacturers may want part, or all, of the intellectual rights to your product.
Part 3 – Develop the Electronics
Development of the electronics for your product can be broken down into seven steps: preliminary production design, schematic diagram, PCB layout, final BOM, prototype, test and program, and finally certification.
For a very basic introduction to electronics see this article which focuses mostly on the various types of electronic components. Then for an introduction to basic electronic circuits see here.
Below are the steps for developing a new electronic product:
Step 1 – Create a Preliminary Production Design
When developing a new electronic hardware product you should first start with a preliminary production design. This is not to be confused with a Proof-of-Concept (POC) prototype.
A POC prototype is usually built using a development kit like an Arduino. They can sometimes be useful to prove that your product concept solves the desired problem. But a POC prototype is far from being a production design. Rarely can you go to market with an Arduino embedded in your product.
A preliminary production design focuses on your product’s production components, cost, profit margin, performance, features, development feasibility and manufacturability.
You can use a preliminary production design to produce estimates for every cost your product will need. It is important to accurately know the costs to develop, prototype, program, certify, scale, and manufacture the product.
A preliminary production design will answer the following pertinent questions. Is my product feasible to develop? Can I afford to develop this product? How long will it take me to develop my product? Can I mass manufacture the product? Can I sell it at a profit?
Many entrepreneurs make the mistake of skipping the preliminary production design step, and instead jump right into designing the schematic circuit diagram.
By doing so, you may eventually discover you’ve spent all this effort and hard-earned money on a product that can’t be affordably developed, manufactured, or most importantly, sold at a profit.
Step 1A – System Block Diagram
When creating the preliminary production design you should start by defining the system-level block diagram. This diagram specifies each electronic function and how all of the functional components interconnect.
Figure 2 – A block diagram specifies each function and the system-level connectivity.
Most products require a microcontroller or a microprocessor with various components (displays, sensors, memory, etc.) interfacing with the microcontroller via various serial ports.
By creating a system block diagram you can easily identify the type and number of serial ports required. This is an essential first step for selecting the correct microcontroller for your product.
Step 1B – Selection of Production Components
Next, you must select the various production components: microchips, sensors, displays, and connectors based upon the desired functions and target retail price of your product. This will allow you to then create a preliminary Bill of Materials (BOM).
Figure 3: Selection of the production components is a critical first development step.
In the U.S., Newark, Digikey, Arrow, Mouser, and Future are the most popular suppliers of electronic components. You can purchase most electronic components in ones (for prototyping and initial testing) or up to thousands (for low-volume manufacturing).
Once you reach higher production volumes you will save money by purchasing some components directly from the manufacturer.
Step 1C – Estimate the Production Cost
You should now estimate the production cost (or Cost of Goods Sold – COGS) for your product. It’s critical to know as soon as possible how much it will cost to manufacture your product.
You need to know your product’s manufacturing unit cost in order to determine the best sales price, the cost of inventory, and most importantly your profit.
The production components that you’ve selected will of course have a big impact on the manufacturing cost.
But to get an accurate manufacturing cost estimate you also must include the cost of the PCB assembly, final product assembly, product testing, retail packaging, scrap rate, returns, logistics, duties, and warehousing.
Step 2 – Design the Schematic Circuit Diagram
Now it’s time to design the schematic circuit diagram based upon the system block diagram you created in step 1.
Figure 5: Example of a schematic circuit diagram.
The schematic diagram shows how every component, from microchips to resistors, connects together. Whereas a system block diagram is mostly focused on the higher level product functionality, a schematic diagram is all about the little details.
Something as simple as a mis-numbered pin on a component in a schematic can cause a complete lack of functionality.
In most cases you’ll need a separate sub-circuit for each block of your system block diagram. These various sub-circuits will then be connected together to form the full schematic circuit diagram.
Special electronics design software is used to create the schematic diagram and to help ensure it is free of mistakes. I recommend using a package called DipTrace which is affordable, powerful, and easy to use.
Other popular PCB software packages include Altium Designer and Eagle. But be warned these are quite expensive and are best for those designing multiple products.
For a free, open-source PCB design tool also be sure to check out KiCad.
Step 3 – Design the Printed Circuit Board (PCB)
Once the schematic is done you will now design the Printed Circuit Board (PCB). The PCB is the physical board that holds and connects all of the electronic components.
Development of the system block diagram and schematic circuit have been mostly conceptual in nature. A PCB design though is very real world.
Figure 6: The schematic circuit diagram must now be turned into a PCB layout design.
The PCB is designed in the same software that created the schematic diagram. The software will have various verification tools to ensure the PCB layout meets the design rules for the PCB process used, and that the PCB matches the schematic.
In general, the smaller the product, and the tighter the components are packed together, the longer it will take to create the PCB layout. If your product routes large amounts of power, or offers wireless connectivity, then PCB layout is even more critical and time consuming.
Figure 7 – Example of a Printed Circuit Board (PCB) layout design
For most PCB designs the most critical parts are the power routing, high-speed signals (crystal clocks, address/data lines, etc.) and any wireless circuits.
To learn more about PCB design be sure to watch this free tutorial and check out this in-depth paid course which teaches you every step of PCB design.
Step 4 – Generate the Final Bill of Materials (BOM)
Although you should have already created a preliminary BOM as part of your preliminary production design, it’s now time for the full production BOM.
The main difference between the two is the numerous low-cost components like resistors and capacitors. These components usually only cost a penny or two, so I don’t list them out separately in the preliminary BOM.
But to actually manufacture the PCB you need a complete BOM with every component listed. This BOM is usually created automatically by the schematic design software. The BOM lists the part numbers, quantities, and all component specifications.
Step 5 – Order the PCB Prototypes
Creating electronic prototypes is a two-step process. The first step produces the bare, printed circuit boards. Your circuit design software will allow you to output the PCB layout in a format called Gerber with one file for each PCB layer.
These Gerber files can be sent to a prototype shop for small volume runs. The same files can also be provided to a larger manufacturer for high volume production.
The second step is having all of the electronic components soldered onto the board. From your design software you’ll be able to output a file that shows the exact coordinates of every component placed on the board. This allows the assembly shop to fully automate the soldering of every component on your PCB.
Your cheapest option will be to produce your PCB prototypes in China. Although it’s usually best if you can do your prototyping closer to home so as to reduce shipping delays, for many entrepreneurs it’s more important to minimize costs.
For producing your prototype boards in China I highly recommend PCBWay, Bittele Electronics, Seeed Studio, or Gold Phoenix PCB. They all offer fantastic pricing on quantities from 5 up to 8,000 boards.
In the U.S. I recommend Sunstone Circuits, Screaming Circuits, and San Francisco Circuits which I’ve used extensively to prototype my own designs.
It usually takes 1-2 weeks to get assembled boards, unless you pay for rush service which I rarely recommend.
Step 6 – Evaluate, Program, Debug, and Repeat
Now it’s time to evaluate the prototype of the electronics. Keep in mind that your first prototype will rarely work perfectly.
You will most likely go through several iterations before you finalize the design. This is when you will identify, debug and fix any issues with your prototype.
This can be a difficult stage to forecast in both terms of cost and time. Any bugs you find are of course unexpected, so it takes time to figure out the source of the bug and how best to fix it.
Evaluation and testing are usually done in parallel with programming the microcontroller. Before you begin programming though you’ll want to at least do some basic testing to ensure the board doesn’t have major issues.
Nearly all modern electronic products include a microchip called a Microcontroller Unit (MCU) that acts as the “brains” for the product. A microcontroller is very similar to a microprocessor found in a computer or smartphone.
A microprocessor excels at moving large amounts of data quickly, while a microcontroller excels at interfacing and controlling devices like switches, sensors, displays, motors, etc. A microcontroller is pretty much a simplified microprocessor.
The microcontroller needs to be programmed to perform the desired functionality. Microcontrollers are almost always programmed in the commonly used computer language called ‘C’. The program, called firmware, is stored in permanent but reprogrammable memory usually internal to the microcontroller chip.
Part 4 – Develop the Enclosure
Now we’ll cover the development and prototyping of any custom plastic pieces. For most products this includes at least the enclosure that holds everything together.
Development of custom shaped plastic or metal pieces will require a 3D modeling expert, or better yet an industrial designer.
