Picking a Contract Manufacturer (CM)

In September, 2010 we needed to figure out who was going to make our little robot, Sphero. We had very little experience outsourcing a high tech electromechanical product for manufacturing and needed to find someone to help us. Fortunately for us, we hooked us with a firm called Dragon Innovation and they ushered us through he process of finding a CM and building our beloved Sphero in China. Here are some lessons from our experience:

1. Before you start, ask yourself if you need an overseas CM. Parts are parts - you’ll be able to buy them at roughly the same place anywhere in the world - so don’t think parts are cheaper in China. If you answer yes to some of the following questions going offshore may be a good solution for you. Does your product have a fair amount of manual assembly? Are you going to order north of 10k units to start with an eye to 100k+ units on an annual basis? Do you have any electromechanical actions? Do you have time to dedicate to travel and communicating your idea to a CM?
2. Get an intro to various CMs. We used Dragon but there are other firms that can make an intro. You think you are picking a factory but in reality they are picking you. Factories do not have thousands of clients - they have a handful that make up the majority of their business and every year they take a bet on a couple that they think will turn into something big. You need to convince them you are a worthy bet. While they may cover their costs on your initial product run - chances are they will be losing money until they crank out hundreds of thousands of units. This is why an intro is necessary - the best CMs will want to know there has been some sort of vetting process.
3. Pick a firm / person to help you. When picking a firm or person to make the intros you should look for folks that have built similar products in the past. The Dragon team came from iRobot and was part of the Roomba team - one member of the team spent time at Hasbro earlier in his career. If you were to look at what we are building with Sphero, their past experience lines up directly with our goal to build a sophisticated consumer robot that is also an electronic toy. Some firms may also bring industrial, electrical and mechanical design to help take your idea to a level that can be manufactured. Alternatively you can hire folks to do this in house but that will take time to build. We eventually graduated from using Dragon, as our volume became large enough where we have built our own team of experts, but they were hugely important to our initial success.
4. Travel and meet the CMs. Plan a trip and expect to see one factory a day. There will be a lot of driving. Remember you are selling and being sold. You’ll most likely meet the CEO or owner of the factory. You’ll base yourself out of Hong Kong and travel to China. Dress casually but not like a slob - jeans and a polo or button down shirt are fine. No flip flops (I hate flip flops in the work place).
5. Examine products they have made in the past. Buy them from retailers and look at the quality, ask to talk to some of their clients if able. Every factory will have a showcase in their lobby of what they have built in the past. Look for similar products. Look for products that are original designs vs. knock offs. If you see a bunch of products that are knock offs then you know that this factory might not have a problem sharing IP. If you only see brands you recognize and products that are still current then you know the big companies trust this factory with their IP. Be alert for factories that try to develop their own products and then slap on someone else's brand. This isn’t bad but remember if it is their design, they will try to sell it to as many companies as possible. Our factory only builds products that are for major brands and does not design generic products for others to label.
6. Don’t pick the factory based on the CEO. In our case, we met a very outgoing CEO of a factory that was super excited about our product. He took us to a fancy lunch and had some great ideas. They made a lot of robot construction kits which spoke to our DNA. But Dragon recommend that we choose the factory that was more down to earth. We ate lunch with their employees in their canteen. Their attitude was very calm and professional and their pride was more subdued. They were a bit larger than the other factory and worked with major toy brands. The back channel from the CM was very positive and Dragon felt like they had the best R&D group for our type of product.
7. Contracts are not for your legal protection. They are great for finding a meeting of the minds but are nearly impossible to enforce without a great expense on your end. Bottom line - you have to trust the people. If the trust is violated you will have to walk. Don’t expect great payment terms as a startup - trust is earned both ways so pay on time and once you are humming along making product reopen the conversation about terms. It isn’t worth the energy to fight for it up front. Don’t pick a factory based on terms.
8. Meet the team. Ask to meet the team(s) managing the project. Chances are you will have one team to build the prototypes and then you will be handed off to another team to actually make the product. You may even be moved to a different facility (many CMs have multiple factories).
9. Get a price quote. Asking for a quote is fair game but remember it is a quote. If the product is loosely defined the quote will just be an estimate. If you have spec’ed every part down to the resistors and the exact packaging, then the quote will be more accurate but still may fluctuate due to changing labor costs and part availability. If you are reading this far then you probably not well spec'ed. Determine what is the most important metric to hit - price, design, or features and focus on that. What is quoted is not the real price - plan for variability. If you hold their feet to the fire on price to the point it is a losing proposition, they will fire you as a client.
10. Be present. Selecting a CM requires active management and plenty of face time. If you think you can do it all by Skype - you're wrong. You need to travel and spend significant time at the factory working with the various teams. Senior management needs to show up as well - you can't delegate the creation process especially if you are a startup. If your product is super simple like a plush toy or a simple IC - you may get away without a ton of travel. But if you are building a product no one has ever built, you'll want to be there to make sure things are going well.

Here is a video I made to share with our company about my most recent trip to China.