If you're a hardware company, you are in the business of selling things. In order to make money, you need to sell your things for more than they cost you to make. This is basic economics, but the application of this principle is anything but simple.
In this day and age of hardware startups launching on Kickstarter, or offering presales on their website, it is even more important. If you price your product too low, you run the risk of cutting your margins down so low that you don't make enough money to survive. This death-by-success is worse than the alternative which is to over-price your product and not generate enough sales or interest to run a viable business.
When pricing your product you are walking on a knifes edge: too high and you can't launch, too low and you risk imploding. The only surefire way to avoid this is to actually produce and pay for your product before selling a single one. For most startups, this is impossible: so you must estimate.
Estimation by Quotation
The best way to estimate is to get a quotation from your manufacturing partner. Typically this is somewhat binding, and you can count on it being close to the true cost, especially since the factory is in the exact same boat as you: if they price it too low, they won't make money. If they price it too high, you won't want to buy it from them. The challenge here is that you need to have a final design that is unlikely to change or you run the risk of cutting into your own margins.
Another challenge with factory quotations is that they could just be making numbers up. You need to have your own idea of what your product actually costs, otherwise you could be getting taken advantage of. Of course, there are strategies for dealing with this as well. The simplest one is getting multiple quotes from different vendors. You'll quickly see who the outliers are, and what an average quotation looks like.
If you're worried about IP concerns, make sure you only send out a limited package for quotation, or break up your project into submodules and have them quoted individually. It is not unreasonable to ask for an NDA as well as this stage. If you really aren't comfortable with sending your design out for multiple quotations, you may need to get some Xanax and think about doing your own estimation. :)
DIY Cost Estimation
Estimating the cost yourself is extremely critical, and something I start doing before I ever touch the computer to start designing. If you can't produce your idea for the right price, it will never get off the ground. I also believe that it is very important for designers and engineers to understand the cost ramifications of every design decision. Each choice you make has an impact on the final cost, and that should not be overlooked.
The first thing to do for cost estimation is to start listing all of the things that will need to go into your product, starting with the most expensive. The more thorough you are, the more accurate your estimation will be. Think of everything: electronics, custom parts, cables, molded parts, metal parts, gears, pulleys, belts, lcds, nuts, bolts, etc.
If you have a prototype you can simply disassemble it and add each item to your spreadsheet as you go. Of course if your prototype is based on some sort of CAD design, you can likely just export a BOM and call it a day. Whichever way you choose the end result will be some sort of spreadsheet with parts, quantities, and prices.
Now, the tricky part: cost estimation!
If you have existing suppliers or quotations for any parts, it's simple: add in the estimated cost (as well as a note saying where you got this cost information from.) If you don't have a quote, well its time to put your beret on and go into artiste mode. There is no science here, and a ton of this will be guesswork from here on out.
First, start with the most expensive items. These will make up the brunt of the cost of your product, so getting them nailed down will allow you to get a ballpark estimate of the cost. With that in hand, you can start getting a feel for how this is going to turn out.
Assuming you plan on manufacturing in China, the best place to get rough price estimate data is Taobao.com. This Chinese-only website is a huge online marketplace for all sorts of goods. You can get everything from handbags to capacitors. All you have to do is know how to search for it.
The two most important tools for searching Taobao are Google Translate and Google Chrome. Google Translate allows you to take your English query and turn it into Chinese. Copy and paste that into the search box, and away you go. If you use Google Chrome, it will automatically translate the results back into English so you can navigate and understand what you're being shown. It isn't perfect, but it gets you 80% of the way and that is enough to get a price estimate.
Taobao has a few features that are useful for doing price estimation:
- Sort by price, filter by price. Obviously you are looking for a price estimate, so sorting your results by price is a natural way to go about it. Be careful that you don't get unrealistic prices by sorting lowest to highest. Make sure you do a reality check.
- Sort by popularity and sales. This is my favorite way to do it. There is safety in numbers: if lots of people are buying this product, chances are that its legit. Check the reviews too as they will often have valuable feedback on the product quality.
- Related products from seller. Oftentimes you'll search for something and it isn't quite what you want. The merchant may have what you're looking for, just not where you're looking. At the bottom of the page, check out the related products and see if any catch your eye. This has definitely clued me into the proper part (or even a better suited part) in the past.
Pricing on electronics is pretty straightforward, and consists of 5 different cost categories:
1. Setup / NRE Fees
These fees consist of things like engineering setup, machine programming, SMT stencil creation, etc. Your solder stencil should not be over $100, and if they're hitting you on fees here you should reconsider your choice of supplier.
2. PCB Fabrication
The PCB cost is almost entirely based on size. Certain options will increase the price per square cm such as finish, copper thickness, # of layers, etc. There are tons of options and it can get pretty pricey, pretty quickly. For that reason, it is best to use an online board cost calculator. If you have a good PCB manufacturer, they will have one for you. I use HQPCB, but their stuff is Chinese only and you need to log into get to it.
Here are some online calculators:
- Gold Phoenix PCB - Relatively well known online PCB shop.
- Bittele - divide by half to get a more accurate China price.
- HQ PCB - Local Chinese PCB manufacturer, site in Mandarin. A bit on the high side as its intended for prototype quantities.
This is the cost of all the actual chips that are assembled onto your board. Don't just go with the Mouser or Digikey price, because its much more than it costs in China. A great way to get pricing is to use Taobao to find out what it really costs.
Some rules of thumb:
- $0.005 for each passive (resistor, capacitor, diode, etc.)
- $0.001 / pin for connectors or headers
There are 2 main phases to PCB assembly: automated SMT assembly and manual through-hole assembly. Obviously the automated assembly is less expensive than the manual assembly because robots. This is one of the reasons why SMT is incredibly popular.
Some rules of thumb:
- $0.01 / pad for SMT
- $0.025 / pad for through-hole
It is highly recommended that you test every single board that comes off the line using a functional test fixture. Any defective product that makes it to your customers should give you a sick feeling in your stomach. Therefore, testing needs to be baked into every step. Unfortunately that costs money.
Estimate about $0.06 / minute required to test your boards. This will be less if one person can test multiple boards in parallel.
Quoting Injection Molding
If you're building a consumer product, chances are you'll need some sort of plastic part, and that plastic part will be made using injection molding. The costs for injection molding are very straightforward: the mold (aka tooling), and raw materials. Your factory may break it down even further into labor costs and cycle times to run the machine.
Quoting injection molding can be rather tricky. Luckily for you, there are a couple sites that will do that for you: Protomold.com and QuickParts.com. Keep in mind that these sites are based in the USA and charge USA prices. Estimate 20-30% of that price, you will end up with an accurate estimate of what it will cost you in China.
Another great resource that really lets you dig into the details is this mold cost estimator at Chinaplastic.org. It requires you to input a ton of information such as size, mold material, resin type, gate styles, and cycle times. However, the information you get out of this will be much more accurate.
Assembly of your product can be a significant factor, especially if you have a complex product. The best way to get this information is to work directly with a CM so that they can understand what needs to be done. Keep in mind that it is not just putting parts together, but also testing your final product, packing it up, and all the other logistics of running a factory that go into labor costs.
The minimum wage in Shenzhen is roughly $2.50/hour, but keep in mind that your line will have multiple workers, as well as supervisors who will be making more per hour. You can either estimate how long it will take in man-hours to manufacture your product and multiply by the wage above, or you can use a simple multiplier of your total BOM cost to estimate. 10% of your BOM cost will get you into the ballpark.
Quoting Shipping Costs
Shipping your product is entirely based on weight and volume. If you are shipping large volumes (enough to fill a 20' container) then volume is the main factor to determine your price. First, you'll need to estimate your product weight and volume. Once you have the volume, you can estimate how many will fit into a container.
- A standard 20' container will fit 1,360 cu ft (38.5 m3) of goods. The max weight of your goods is 21,600 kilograms (48,000 lb).
- A standard 40' container will fit 2,720 cu ft (77 m3) of goods. The max weight of your goods is 26,500 kilograms (58,000 lb).
A 20' container from Shenzhen to LA will cost about $3000, a 40' will cost about $5000. These rates change often based on market fluctuations. It takes about 4-5 weeks to arrive.
Xpressrate.com has an online system to get quotations for shipping containers to anywhere in the world.
For air freight, your best bet is to use the calculator of a site like Fedex, UPS, or DHL to determine how much it will cost you. Cypress Industries has a useful table on shipping costs.
The devil is in the details, and there are lots of things that go into making a mass scale product that are simply not a part of small batch kit making or the DIY marketplace. Things like certification, engineering costs, customer support, software development, and returns all cost money and need to be accounted for in a successful business.
To put some numbers on the table, you can expect EMI certification to cost anywhere from $5,000 to $20,000 depending on the complexity of your product and if there are any issues encountered.
Your employees need to be paid and will not work for free. If you plan on having a sustainable business that is around for years to come, you'll need to factor that into the price of your product. Estimate the salary of all the engineers, programmers, and customer support reps you'll need in order to operate and then divide this by the number of products you plan on selling in a year.
Last, but not least: returns. It is a fact of life in the world of hardware that you **will** ship a defective product at some point in time. All of your quality programs are geared towards making sure that this does not happen, but inevitably it will. You need to recognize this in your product cost so that when you get returns it will not sink you. Add in 1-5% based on how confident you are in your product.
It's a good idea to add in a safety factor during the beginning stages to account for all of life's little surprises. As you get closer to the final product and more things transition from guesses to facts, you can whittle away at this safety factor. I typically start out with a 10% safety factor in the beginning and then nothing by the end of a project.