How to Find the Wrong Factory To Make Your Product

Reposted from here.

For a product company, your factory is where all the magic happens. When things go wrong the factory is the place you get on a plane to go and fix stuff. It’s a central source of frustration for many startup CEOs because those relationships sometimes don’t work out.

There are dozens of things that could go wrong with finding the right factory. So let’s chat about all the things you could do that could make things much more complicated. Then, we can talk about the 80/20 optimization that you can do to simplify your life and your production process.

You try and find a factory too early

Timing is critical when it comes to finding a manufacturing partner. You get excited about your product and you want to figure out how much it costs so you reach out to factories early and often. One of two things will happen:

They will turn you away because you will likely have the information they need to provide an accurate quote. (Or that you’re a startup and they don’t deal with startups)

Or

They will realize you’re not ready yet and then subsequently hound you weekly to either to sign a letter of intent, provide a misleading quote using the little information you provided or, sometimes, just keep asking for the things they need to move your relationship forward.

The sales guy just called..

What does this mean for you? It will divert your attention to things you shouldn’t be focusing on like in person meetings, phone calls and lengthy emails. You shouldn’t be focusing appeasing a contract manufacturer, you should be focusing on getting the minimal amount of information together in order to obtain the key knowledge that will help you to make the best business decisions. In other words, if you don’t have the documentation and the (close to) final designs, focus on that first!

Remember, to best utilize your time when building a product, you shouldn’t reach out until you have the proper documentation and product details. I will go in depth on this later down below.

You try and find a factory too late

Did you ever have to take a high school or college course where you stood in front of the room and speak for 5–10 minutes? What do you think would happen if you didn’t prepare for that speech?

The answer is obvious here: You’d likely bomb it and life would be over as you know it. (Not really but you get my point here. 😎)

Bringing on a factory with 1 month to production is the equivalent of making a speech and not knowing what to talk about. Without bringing up your factory, which by the way takes about 4–6 months, has to be done intentionally and with enough buffer time to work out all the kinks in the assembly process. There will be problems you don’t foresee. There will be components that don’t show up on time. Expect it.

It takes time…

Tooling (cutting out a cavity in a piece of metal) for injection molding parts in and of itself takes about 6 months of day in and day out effort. Then, once the tooling is created, tweaks are required to get the injection molding process optimized. Again, one can expect this to happen in one month but you may as well flush all your time and money down the toilet. As it stands right now, creating tooling is tough and requires patience and time. As we move forward with technology, especially in producing plastic or plastic like enclosures, we may see a change in this paradigm. For now though this is the hard truth.

You didn’t create your RFQ package

Going for a detailed RFQ (request for quotation) requires you to be ready to give the details that are needed to know how much a product will cost. Would you ever expect someone who builds homes to know how much it costs without having a detailed blueprint and build plan? The same thing goes for products. Without a bill of materials, details on all the tooling requirements (plastic, metals, etc), all the components (including electronics) all the marketing department at a factory can do is stick their finger in the air and determine your costs by which way the wind is blowing.

What you need:

Checklist

These are the must have documents that you must get together before you start getting quotes from factories. There is no doubt that they will ask for it. Here is an exhaustive list of what you will need:

Electronics Bill of Materials: an electronics BOM is generated typically by exporting the component details from your product’s schematic. A bill of materials typically includes the following info per line:

  • Your internal part number
  • The manufacturer’s part number
  • If relevant, the vendor part number for that part (i.e. The distributors part number)
  • A detailed description of the part. Example: Resistor 0402 1% 100k
  • The reference designator(s). i.e. For resistors these are typically the names of the parts: R1, R2, R3, etc
  • The quantity of this part

Electronics Gerber Files and Schematics: I typically call this the fab package. It consists of the following:

  • Your full schematic. This schematic should have an internal part number
  • Your board file. This is the file that is used to generate the gerber files
  • Your gerber files or ODB++ files. These are the vector ‘images’ of each layer of your printed circuit board.
  • Your stack-up documentation. This indicates how thick your PCB is, what the finish is, etc.

Mechanical Bill of Materials: much like an electrical bill of materials, the mechanical BOM Includes all the part numbers, descriptions and designators for each part. This is often referenced with the assembly plan. I’ll provide some details on the assembly plan down below.

3D CAD Output Files (.stp) : these are the files that the factory will base their tooling costs on. These are the files that make your plastic injected parts real. Your mechanical engineer should be very familiar with these files and should be able to generate them as needed.

An assembly plan: you need an assembly plan to describe what each step is in the process from getting your product shipped and out the door. This typically starts after the printed circuit board testing phase. Here are a few examples of steps in the process:

Step 3: Place PCB in housing and screw the PCB to the housing using 0.5 inch/lb of torque
Step 4: Using an overhead press, push additional battery contacts into the plastic housing

An assembly plan is a highly technical and step oriented document. Be sure to take as many pictures as relevant for each step (something I didn’t do here). This will immensely help the factory calculate costs for you so you have an idea of how much your goods cost ex-factory.

Note: before sending any of these files you should have at least a mutual nondisclosure agreement in place with this potential factory. Also make it clear that these are your designs that you are disclosing to them as intellectual property.

Marketing data: what the factory wants to know is approximately (and realistically) how many products you are going to build that year or how many products you intend on building once at full steam. You don’t have to go too conservative on this number as factories will protect themselves on their end if you provide too big a number. Moral of the story here is to have the data. It will help grease the skids when things start getting sticky.

You went with the best sounding offer even though it was to good to be true

My first company went to quote with several factories located in the Shenzhen/Guangdong Provence area of China. I believe they visited 3–4 factories showing them what our next product was and getting some initial RFQs. One of the bigger factories came in way too expensive while some mid-tier factories came in at about the same and finally a smaller local factory had the lowest bid. The smaller factory seemed to have the experience and the low cost was a plus. But, when we signed on, things started falling apart almost immediately.

As we continued our development process the factory soon realized that we were still very early in flushing out the bugs. Therefore, as we found issues or had to change components it would cause the factory to start the quoting and ordering process from scratch. It got to the point where every time we had to make a change to the bill of materials we would find that some components would be grossly marked up comparing to distribution pricing here in the USA. At the time Octopart was just launching and I had used it extensively to ensure we were getting the best pricing we could be. We soon realized that we weren’t the magical client that they had expected and they were not the magical factory that we expected.

Rather, if we had taking the time to do our due diligence and provide as much complete information to the factories as possible we would have had a far better idea of who was the right group of people to work with.

You chose the wrong location

Let’s say you decided to pick a factory that happens to be one hour away from your office. It’s convenient and you think the value of traveling back and forth between the factory and your office is much more than spending several weeks at a time somewhere in Asia. This is great logical thinking until something else fails.

Sometimes they have good excuses..

Say the majority of your design is actually manufactured in China. All the mechanical parts, the printed circuit boards, screws etc. because we all know that things are cheaper there. You think you’re all set but soon realize you need to order most of these parts 8 weeks ahead of your scheduled production start. 8 weeks! I chuckle every time I see that phrase because most vendors, when they quote, they indicate that a part has an 8–10 week lead-time. I think it’s those vendors sandbagging more than anything. I digress, as a small customer there is little you can do about it!

To give you some perspective, here’s what the timeline for shipping parts from China to the USA looks like:

Factory builds the parts ➡️ Packs them up/creates the invoices and pays the necessary taxes for the parts to leave the factory ➡️ Gets a shipping company to likely ship those items across the border into Hong Kong ➡️ Parts get loaded on a container with other things ➡️ Transit via boat ➡️ Boat arrives and starts getting unloaded ➡️ Wait a few days for customs release ➡️ Items get picked up by a shipping company ➡️ Items arrive at your factory where they need to be inventoried ➡️ Parts are used for production ➡️🎉

Complicated, yes? What if, at any point in time, the chain breaks? No parts. No production. (Alternatively, you can ship via air but you may as well say goodbye to any profit margin you have. Especially if you’re shipping piece parts)

Every link is important. Don’t break the chain!

Sometimes I wish it was as easy as this..I definitely don’t want to naysay having production in the United States. It is completely doable and at affordable rates but your supply chain needs to be optimized for domestic assembly not assembly in Asia. Focusing on local vendors will far outweigh the benefit of saving 10¢ and 20¢ there. Remember receiving something on the back of a truck the same days is always more effective and likely cheaper than getting it sent overseas via plane or boat.

Here’s another though provoking exercise: if you were Apple and your whole supply chain is in Asia, why the hell would you want to go anywhere else? They would be at the whim of people shipping theirs parts to make phones, computers, etc. (which, in some cases, they are) If one shipment is late or the manufacturer is having an issue with one of their parts there is no chance to get things done faster. Granted if you’re scrappy, there are ways, but if you’re Apple you can’t just hand carry parts for your production build on an airplane. We’re talking about millions and billions of parts here not a couple thousand!

You chose a factory based on invalid marketing data

Marketing data can be useful but it can also be a death knell. Let me give you an example:

We all make mistakes..

Company A wants to build a product. There seems to be lots of hype about the product and they decide to run a pre sales campaign for the product. They have a successful campaign and everyone seems to be happy. Let’s say, as a ballpark, that they estimate that they can sell around 100k units their first year. That’s about 2000 units per week. Or a little more than that when you remove 2–3 weeks for vacation or holidays like Chinese New Years. They use this data to convince a factory to give them special pricing at the 100k quantities. Things seem to be going well until things start falling apart.

Unfortunately, they soon realize that they cannot make their product they originally intended to. After some flailing and internal struggle they have to change the design significantly in order to ship anything within the next year or so. The problem? The market validated the original product not the ‘pivot’ product.

The company works tirelessly to create this product. They polish it as much as possible and get it to production within a reasonable amount of time. The factory soon realizes though it takes 2x as long as they originally anticipated based on the original design. The Company A’s marketing team realize that they’re not going to sell nearly as many units. The factory immediately determines this from the drastically reduced initial order volume and raises their prices accordingly.

Don’t drink the cool-aid. Be real.

Think before you jump.This is a complex scenario but it goes to show that a few series of events can alter how a factory will treat you. For instance very large factories like Foxconn, Foxlink, etc charge hefty fees for smaller customers to build at their facilities because you are taking up space that would otherwise be occupied by an iPhone assembly line. The only way it makes sense is if they’re making more money off you than the other guy and thus why your cost of goods sold shot through the roof and you’re out of business.

So, when going into a factory deal sometimes it pays to be both optimistic but also conservative. There are factories out there that give you the option to start at low quantities (3–5k) and move to much larger ones. But it requires transparency at the beginning and it needs to make sense for both your company and the factory. Otherwise, no one wins.

Conclusion

Wrangling your factory can be a life altering experience. If you don’t mind the extra grey hairs take on but be ready for some hard work. 👴

There are technologies out there though that make things much easier. PLM systems often take lots of the stress and hurt out of the equation. I personally used Aligni, which is still a great low cost option for companies of all sizes. Put one of these in place before you get building if you can.

I digress, these stories should help you make some actionable decisions before ever signing a letter of intent or factory contract. Every product and company is different so your mileage may vary.

Looking to learn more about hardware startup life? Check out my blog where I go into vivid detail on how not only to save you time but save you money when bringing products to market.