One of the most important relationships when building a product is with your contract manufacturer (CM). They hold the key to realizing your vision on time and on budget, helping you avoid painful (and often expensive) mistakes as you move from design for manufacturing (DFM) to parts sourcing and final assembly.
Finding a CM that fits your needs is essential to producing a successful hardware product. However, CMs vary about as much as hardware products themselves, from what they are able to make to their expectations about batch size, delivery deadlines, and quality assurance.
Because there are so many factors to consider, choosing the right CM is a hurdle for many first-time hardware creators. To get some insights about how to approach working with a CM, we talked with three creators who launched hardware products on Kickstarter about the challenges and triumphs of navigating the world of manufacturing.
Here’s who we spoke with:
Sam Shames of EMBR Labs
EMBR Labs is creating a wearable thermostat that allows you to regulate your personal temperature. With no existing relationships with manufacturers (the EMBR Wave was the first product this team had brought to market), they relied on referrals to find a CM who could help refine their final prototype and get their device into production.
Henry Haslam of Technology Will Save Us
Technology Will Save Us creates kits that teach kids about STEM topics through fun, creative projects, from the Dough Universe Kit’s conductive clay to the Mover Kit’s programmable motion sensor. This experienced team has launched multiple products, and vetted CMs from several countries to find a manufacturing partner that fit their needs.
Jessica Banks of RockPaperRobot
RockPaperRobot creates kinetic furniture that moves and transforms in unexpected ways. Their most recent creation, the Ollie Chair, is a shape-shifting chair that unfurls and retracts with the pull of a string. It was the team’s first foray into large-scale manufacturing, and they learned many lessons along the way.
How did you find your CM?
Embr Labs: We were lucky to find our CM through the MIT Venture Mentoring Service Network. We spoke to a number of companies who had worked with this CM for many years and got a warm introduction to them.
Technology Will Save Us: We found our CM through a referral from another London-based consumer electronics company. We requested quotes from five to six CMs (based in the UK, Europe, and China), and then went to visit two or three suppliers in China, before picking one.
RockPaperRobot: We started with a manufacturer in China and then switched to our current one in Montreal, Canada. The first was a contact of a business colleague and the second found us after they saw our Kickstarter campaign. Basically the Canadian manufacturer poached us even though we had already spent money on tooling overseas. Doing it again, I would start the process with my first choice CM and then research expanding the supply chain to have backups.
What contracts did you sign with your CM, and when in the production process did you sign them?
Embr Labs: We started with an Non-Disclosure Agreement (NDA) before we shared our design with the CM. We also signed a Master Service Agreement (MSA) after the prototyping stage. Looking back, we probably should have signed the MSA sooner, since it acts as your legal protection in case something goes wrong with the relationship with the CM. We went on trust and didn’t sign before starting because the CM was a referral, however the best practice is always to sign this document before you start working so you have a written record in case of a dispute.
Technology Will Save Us: We have a manufacturing agreement in place with our CM. We used a modified version of a standard template (which we had our lawyers review), and then took our manufacturer through the agreement.
RockPaperRobot: We had NDAs with both the Chinese and Canadian CM, but didn't have contracts with either of them. Our Canadian CM said they only do purchase orders (POs) as contracts at first. After extensive emails delineating our agreements, I felt comfortable enough to proceed — what a mistake. Our CM didn't adhere to many of our "agreeables" and we spent six months and tons of money working with them to get things right.
Manufacturing a first run is tough and neither you nor the CM will have a full understanding of the hurdles from the start. A contract (especially with a CM that’s abroad) isn't likely to cover all issues. However, when you’re in the trenches, it’s sometimes better to keep pushing through rather than stopping to enforce contractual obligations on a partner and stretching a relationship beyond its limit.
What was one of the biggest hurdles you faced in the Request For Quotation (RFQ) process?
Embr Labs: The biggest challenge for us was that our design wasn't finalized so we had to keep revising the RFQ with our CM. As we went though design for manufacturing, we asked our CM for input and made changes to the design. These design changed needed to be reflected in the RFQ, making the entire process complicated.
Technology Will Save Us: Ensuring we were comparing apples with apples was an initial issue with the RFQ process. For example, for plastics, some manufacturers would give you a low per unit cost, but the tooling costs would be very high or vice versa. We did receive an itemized quote from most of the suppliers, which we were able to cross-reference against quotes we had obtained ourselves. The quote also changed after appointment of the CM, due to changes we made in the product as a result of refining the user experience and availability of some components.
RockPaperRobot: We didn't have an RFQ process so this isn't really applicable. But I will say this over and over again: if it looks too good to be true, it is.
Did the project fall behind schedule at all? If so, what was the main reason? If you managed to stay on schedule, what was your secret?
Embr Labs: Our project did fall behind schedule because of design-related issues. We had to go through several more iterations of DFM than we expected. This was a result of needing to decrease the failure rate (i.e. number of units that weren’t up to our standards) and reach a final design that had high enough yield to be manufactured at scale.
The biggest design issue was hand soldering lead wires. Because of the size of our product, we could not reliably solder lead wires to the PCBA despite three different attempts. Ultimately, we switched to a cable harness, and that solved the issue so we could produce reliably at a larger batch size.
Technology Will Save Us: We have had a mixture of minor delays on our Kickstarter projects, including design issues, component availability and slower than anticipated production. The best thing we have done to reduce the impact of these issues is to arrange visits during the most important parts of the process. We have also implemented regular conference calls with our suppliers to reduce the time it takes to report and address any issues that come up.
RockPaperRobot: We ended up six months behind the schedule that the manufacturer had sold us on when we signed up (though still right on time for our Kickstarter ship date). I could say that this delay was because our product was their first business to consumer (B2C) inventoried product and that they just weren't set up physically, operationally, or managerially to handle such a thing — but really, it was my fault. I was a first time buyer — wide-eyed and bushy-tailed and inexperienced with (or blind to) what, in retrospect, were clear indications that the CM may not be the right partner. These clear indications came in a few forms:
The CM specialized in B2B products, however the support required to do a B2C product is much different, as are the repercussions for getting something wrong. For instance, if a manufacturer has a quality issue with a B2B order, the whole order can be remade and replaced. For our business, this situation would be catastrophic – customers might leave bad reviews, or worse, get hurt by the product.
The factory was messy and there were neglected parts laying around. The workers didn't have optimal conditions which could have been a sign of them having a lot of work — but more likely, their overall disorganization.
They didn't have a dedicated space for warehousing, even though it was promised.
They just didn’t have the infrastructure and personnel for shipping a B2C product.
They didn’t have rigorous quality control methods or the knowledge of how to execute such a thing in terms of assembly line efficiency, fabrication standards, or accountability.
At a certain point, our priority became making sure that the product quality was high enough, no matter how long it took. A secret of success is sometimes changing the definition of success.
How often did your team visit your CM?
Embr Labs: We visited three times in the first year of working together: once during the prototype stage, once during the engineering sample stage, and once during the engineering pilot stage. We learned which aspects of the design were causing yield issues; we visited suppliers to audit their factories; we learned about the test fixtures and quality-test plan issues. We learned about the assembly line workflow and how it limited throughput.
Technology Will Save Us: We typically visit our contract manufacturer three times during production. The initial visit, to hand over the product before tooling, reduces the risk of wasted time due to misalignment of expectations. A second visit, to check the initial tooling and see the first shots out of the mold, allows us to review the plastics in detail, finalising fits, shapes, colors, and tolerances. The final visit, at the stage of mass manufacturing, lets us review whether the production is being done to spec, and solve any last minute problems that may occur.
RockPaperRobot: Initially we had three planned trips to: assess the first samples; sign off on tooling and warehouse; instantiate a quality control (QC) process; and onboard them to our fulfillment system, Shipstation. It didn’t go quite as planned though and we ended up visiting more than ten times in six months for two to ten days at a time. Days were often twelve to sixteen hours, most of them actually doing physical quality control, setting up the plant, and even overseeing the shipment of our Kickstarter units. Most of all, I learned that, as the CEO, I shouldn't be the one there for most visits. We should be sending a dedicated operations person or project manager, ideally someone who can oversee the relationship with the CM and isn’t as stressed about what it means to lose a day of revenue. Ultimately some of the most important skills a CEO can learn are to delegate and distance.
I’d also like to raise another issue: I was the only woman ever in the whole factory aside from the factory assistant. That’s including the fabricators, support staff, management, and clients. A woman can obviously handle the task of building the assembly line and getting the product shipped, however these cultural or subtle gender biases had to be overcome to reach our goals. At first, I let it go, rationalizing that women are just not often on the factory floor, which is incorrect and destructive logic.
I think it’s important (for me) to look for a level of gender balance at a CM, where women, not just men, are in charge of day-to-day operations. This can be hard to achieve depending on location, but it can be helpful as a female business owner to assess where extra energy may be required in the relationship in order to facilitate the most efficient communication and frictionless rapport.
Considering working with a CM? Already seasoned when it comes to negotiating timelines and sourcing components with your CM? We’d love to hear about your experience manufacturing or any questions you have on working with a CM below in the comments.