6 ms·

Hardware startups should consider local manufacturing

68 points by twald 12y ago

reformdesigns 12y ago
Great points - thanks Tobias!
teamhappy 12y ago
I've been thinking about this just the other day. The context then was locally grown weed in the US and the implications of that for farmers in South America. Wouldn't we be better off distributing the "workload" evenly across the planet?
- delecti 12y ago
  International drug trafficking laws are almost definitely a bigger factor in the consideration of where to grow legalized medicinal/recreational cannabis than cost, especially considering it grows so readily in so many different climates.
- davidw 12y ago
  Probably not: http://en.wikipedia.org/wiki/Comparative_advantage#Ricardo.27s_example http://en.wikipedia.org/wiki/Comparative_advantage#Ricardo.2... To take your pot example, there are places where it grows well, and places where it's not so easy to grow. It's better for everyone if it's grown in the easy places, and traded for goods that are more easily produced in the 'not so easy' places. That's the theory, at least. Naturally, that specific example also has a lot of legal issues surrounding it!
amckenna 12y ago
I have backed a half dozen Kickstarters and Indigogos that were hardware based (consumer electronics) and chose to manufacture in China. Seeing the time and money they spend in order to fly someone over to China to troubleshoot and issue with manufacturing or testing has left me wondering - why do small startups choose to manufacture in China to begin with? > Is it because they assume that despite the initial cost over local manufacturing the investment will be worth it if their product takes off? > Is it because everyone else seems to do it? > Or maybe because there aren't any good local places to have electronics manufactured? This I would find a little hard to believe given the concentration of hardware startups in SF, NYC, SEA, which have local hardware manufacturers.
- JonFish85 12y ago
  My guess would be that it starts with the assumption that "China is cheaper", and that makes for an easy story to tell to investors "we're making XYZ in China!". The second you say "we're making it in Texas", it opens you up to questions of "isn't that expensive?" On the face of it, I'm sure the up-front quote from a Chinese manufacturer is cheaper than something in the USA. Where costs skyrocket is when something goes wrong: the parts are wrong, the products aren't ready on time, something was done incorrectly, etc. If you don't have a manufacturer you can trust, one of those will happen, and is almost impossible to diagnose from the US. Even if all of that works out, there are other things to consider: costs (and time) of shipping. Want it in a week? Well, get ready to pay. Even if you ship it air freight consolidated, it's still not cheap. If you want it on a boat, there's a good chunk of time to think about. On top of THAT, there are all sorts of fun import tariffs and such to deal with (at least on the US side). Get ready to front money for an import bond, for starters! And there may already be some hidden costs of importing into the US: if you use some TI parts that have to be imported into China, you're going to be paying taxes twice. A lot of those things are likely not looked at in the low quote from a Chinese manufacturer. They're costs that can/will show up later along the line.
- JonFish85 12y ago
  My guess would be that it starts with the assumption that "China is cheaper", and that makes for an easy story to tell to investors "we're making XYZ in China!". The second you say "we're making it in Texas", it opens you up to questions of "isn't that expensive?" On the face of it, I'm sure the up-front quote from a Chinese manufacturer is cheaper than something in the USA. Where costs skyrocket is when something goes wrong: the parts are wrong, the products aren't ready on time, something was done incorrectly, etc. If you don't have a manufacturer you can trust, one of those will happen, and is almost impossible to diagnose from the US. Even if all of that works out, there are other things to consider: costs (and time) of shipping. Want it in a week? Well, get ready to pay. Even if you ship it air freight consolidated, it's still not cheap. If you want it on a boat, there's a good chunk of time to think about. On top of THAT, there are all sorts of fun import tariffs and such to deal with (at least on the US side). Get ready to front money for an import bond, for starters! And there may already be some hidden costs of importing into the US: if you use some TI parts that have to be imported into China, you're going to be paying taxes twice. A lot of those things are likely not looked at in the low quote from a Chinese manufacturer. They're costs that can/will show up later along the line.
jtoberon 12y ago
In NYC, there's http://www.refactory.co/ http://www.refactory.co/. I used to work with these guys, and they're great.
jtoberon 12y ago
In NYC, there's http://www.refactory.co/ http://www.refactory.co/. I used to work with these guys, and they're great.
jtoberon 12y ago
In NYC, there's http://www.refactory.co/ http://www.refactory.co/. I used to work with these guys, and they're great.
Animats 12y ago
Local manufacturing only works if you're in a place where there's a lot of manufacturing, or your product doesn't really require much manufacturing capability. You many need to be in a place where there's someone down the street with a 20 ton press. If you need a 20 ton press, you need to be in a place where at least three shops have one. If you're making things which are basically PC boards, manufacturing is not too hard. Making boards and soldering parts onto them has a well developed workflow. There are lots of board houses and board-assembly services. If you design for what a pick and place machine can do easily, things usually go well. Hint: production is surface mount today. This is a pain for people who are used to prototyping with through-hole parts. Kickstarter-class startups seem to have excessive problems with making cases and panels. Tooling for injection-moulded plastic is difficult and expensive. Once you get it right, the parts just fly out of the machines at a few cents per part. The production process is just getting warmed up on a run of 10,000 parts, and few Kickstarters get there. Basic truth about manufacturing: most of the processes are really cheap if you're making enough items, and far more expensive for short runs.
Animats 12y ago
Local manufacturing only works if you're in a place where there's a lot of manufacturing, or your product doesn't really require much manufacturing capability. You many need to be in a place where there's someone down the street with a 20 ton press. If you need a 20 ton press, you need to be in a place where at least three shops have one. If you're making things which are basically PC boards, manufacturing is not too hard. Making boards and soldering parts onto them has a well developed workflow. There are lots of board houses and board-assembly services. If you design for what a pick and place machine can do easily, things usually go well. Hint: production is surface mount today. This is a pain for people who are used to prototyping with through-hole parts. Kickstarter-class startups seem to have excessive problems with making cases and panels. Tooling for injection-moulded plastic is difficult and expensive. Once you get it right, the parts just fly out of the machines at a few cents per part. The production process is just getting warmed up on a run of 10,000 parts, and few Kickstarters get there. Basic truth about manufacturing: most of the processes are really cheap if you're making enough items, and far more expensive for short runs.
Animats 12y ago
Local manufacturing only works if you're in a place where there's a lot of manufacturing, or your product doesn't really require much manufacturing capability. You many need to be in a place where there's someone down the street with a 20 ton press. If you need a 20 ton press, you need to be in a place where at least three shops have one. If you're making things which are basically PC boards, manufacturing is not too hard. Making boards and soldering parts onto them has a well developed workflow. There are lots of board houses and board-assembly services. If you design for what a pick and place machine can do easily, things usually go well. Hint: production is surface mount today. This is a pain for people who are used to prototyping with through-hole parts. Kickstarter-class startups seem to have excessive problems with making cases and panels. Tooling for injection-moulded plastic is difficult and expensive. Once you get it right, the parts just fly out of the machines at a few cents per part. The production process is just getting warmed up on a run of 10,000 parts, and few Kickstarters get there. Basic truth about manufacturing: most of the processes are really cheap if you're making enough items, and far more expensive for short runs.
organsnyder 12y ago
I know some people that started a company that makes GPS receivers for avionic use (any pilot will likely be able to name them). I heard them explain their rationale for manufacturing in China, and IIRC it wasn't related to cost—in fact, they said that cost was comparable to domestic production. Rather, the main advantage was the ability to source components, such as the Apple-specific connectors they have to use. Of course, there are difficulties inherent to working with a manufacturer that's half a world away (miscommunications, timezones, travel, etc.), but they determined that it was still the best way to manufacture their product. China especially has manufacturing centers where just about any component can be sourced within a few hours, rather than waiting for parts to be shipped from around the country (if not the world). China has a huge edge in electronics manufacturing, and it's not because of cost (anymore). If we ever hope to compete, we'll likely need to promote such a domestic manufacturing ecosystem via extra-market means.
organsnyder 12y ago
I know some people that started a company that makes GPS receivers for avionic use (any pilot will likely be able to name them). I heard them explain their rationale for manufacturing in China, and IIRC it wasn't related to cost—in fact, they said that cost was comparable to domestic production. Rather, the main advantage was the ability to source components, such as the Apple-specific connectors they have to use. Of course, there are difficulties inherent to working with a manufacturer that's half a world away (miscommunications, timezones, travel, etc.), but they determined that it was still the best way to manufacture their product. China especially has manufacturing centers where just about any component can be sourced within a few hours, rather than waiting for parts to be shipped from around the country (if not the world). China has a huge edge in electronics manufacturing, and it's not because of cost (anymore). If we ever hope to compete, we'll likely need to promote such a domestic manufacturing ecosystem via extra-market means.
organsnyder 12y ago
I know some people that started a company that makes GPS receivers for avionic use (any pilot will likely be able to name them). I heard them explain their rationale for manufacturing in China, and IIRC it wasn't related to cost—in fact, they said that cost was comparable to domestic production. Rather, the main advantage was the ability to source components, such as the Apple-specific connectors they have to use. Of course, there are difficulties inherent to working with a manufacturer that's half a world away (miscommunications, timezones, travel, etc.), but they determined that it was still the best way to manufacture their product. China especially has manufacturing centers where just about any component can be sourced within a few hours, rather than waiting for parts to be shipped from around the country (if not the world). China has a huge edge in electronics manufacturing, and it's not because of cost (anymore). If we ever hope to compete, we'll likely need to promote such a domestic manufacturing ecosystem via extra-market means.
organsnyder 12y ago
I know some people that started a company that makes GPS receivers for avionic use (any pilot will likely be able to name them). I heard them explain their rationale for manufacturing in China, and IIRC it wasn't related to cost—in fact, they said that cost was comparable to domestic production. Rather, the main advantage was the ability to source components, such as the Apple-specific connectors they have to use. Of course, there are difficulties inherent to working with a manufacturer that's half a world away (miscommunications, timezones, travel, etc.), but they determined that it was still the best way to manufacture their product. China especially has manufacturing centers where just about any component can be sourced within a few hours, rather than waiting for parts to be shipped from around the country (if not the world). China has a huge edge in electronics manufacturing, and it's not because of cost (anymore). If we ever hope to compete, we'll likely need to promote such a domestic manufacturing ecosystem via extra-market means.
jtoberon 12y ago
In the NYC area, there's http://www.refactory.co/ http://www.refactory.co/. I used to work with these guys, and they know their stuff. They argue in favor of both designing and manufacturing locally. A quote from their blog: "If you kind of just prototype something and bring it to China expecting that then thousands of these things can be made easily without understanding the processes and materiality of manufacturing, that is gonna be a big problem for you. What works is when the design process is iterated alongside the manufacturing process."
sabalaba 12y ago
Those costs in China are too high. I've lived in Shenzhen for under $70 USD / week in a hostel. If you eat out every meal for 100 RMB, you're not even going to be spending $350 a week for meals. While I agree that if you can do it locally, do it locally, but those numbers for the cost of living and travel in China are inaccurate.
- hawkice 12y ago
  For context: 100RMB is about $16, so you better really be enjoying those meals you're eating 3 times a day, because wowzers that's upscale.
  - sabalaba 12y ago
    Exactly, and they were suggesting that it would be $200 USD more than that per week!
tlb 12y ago
I agree that manufacturing in-house is great for hardware startups. I've seen it work for several. The insight they gain into the details of design, and the speed they can iterate at, more than make up for higher local labor costs.
- twald 12y ago
  hey, author here. we're thinking about sharing stories about hardware startups that decided for the local manufacturing route including price differences, manufacturing tips, gm contacts, etc. If somebody is interested or wants to get involved, please send me an email: tobias@senic.com
  - Gurkenmaster 12y ago
    I'm not from http://schiit.com/ http://schiit.com/ but they also went with local manufacturing route and their prices are quite competitive.
iandanforth 12y ago
I'll add one word of caution, I've looked into getting a product made that had a textile component and a separate product that had a leather component. Precision, automated stitching is either totally absent or completely unaffordable in the US. There are great shops India and China that can do laser cutting and CNC stitching super cheaply and if you want to make an inexpensive product to tight tolerances this is exactly what you want. Unfortunately much of the consumer level textile industry has completely left the US. Same with leather. You can find boutique manufacturers, luxury manufacturers, and aerospace type companies, but the sweet spot of high quality and a good price just doesn't exist here. (I actually hope I'm wrong about that but I did a lot of searching before abandoning a couple projects.)
- JabavuAdams 12y ago
  That sounds like a business opportunity.
- justizin 12y ago
  Read your own comment. You're basically saying that laser cutting and CNC stitching - operations performed by robots based on files uploaded by computers - are done better in China than here. That's like someone in China saying that we are better at standing. People on both sides of the ocean probably do just fine at both. ;)
  - bsder 12y ago
    > You're basically saying that laser cutting and CNC stitching - operations performed by robots based on files uploaded by computers - are done better in China than here. Not done better. They're not done here AT ALL. That's a big difference.
    - justizin 12y ago
      That's a bold claim. Laser cutting and CNC stitching are not done "here" which I guess is outside China, at all, by anyone?
kcorbitt 12y ago
> If you’re based in Germany, there is a 7 hour time difference to east Asia, if you’re based in California, it is 16 hours. No, in California, it's 8 hours (in the other direction). It doesn't make sense to talk about more than 12 hours of time offset except if you really care about aligning the calendar day instead of just the hour of the day.
- nandemo 12y ago
  Around weekends, it does matter. E.g., when it's Friday afternoon in California, it's already Saturday in East Asia.
kosma 12y ago
As usual, the answers are not simple. At our last hardware startup we ended up manufacturing the mechanical parts (injection moulded case, springs, etc.) in China via our industrial design company but did PCBs, assembly and packaging locally in Poland. We simply didn't have enough prior experience to design a foolproof assembly process (which involved flashing, testing on GSM/GPS repeaters, etc.). We are considering moving the to production to China, but only once we solve any critical hardware bugs, smooth out the wrinkles in the flashing/testing process and reach sufficient quantities. With small orders of 100s of units and unexpected disasters like resistors tombstoning due to small errors in solder paste stencils, a five minute call to the manufacturing plant beats cross-timezone, cross-language and cross-culture debugging every single time.
thruflo 12y ago
As a founder of https://www.opendesk.cc https://www.opendesk.cc I think all startups should consider local manufacturing -- when they're fitting out their work space. </tenuous plug>
terramars 12y ago
Totally agree with this. Have a friend who had a successful kickstarter and decided to go with China - he shipped a year late after spending most of his time in Shenzen and going through multiple suppliers. In the interim, they lost the first mover advantage and there are now multiple similar products on the market. If you're not already well established and making a massive amount of something, it's a really bad idea to outsource.
seddona 12y ago
On the electronics side we (https://circuithub.com/ https://circuithub.com/) have spent a lot of time on automation to bring US manufacturing costs inline with China. I definitely agree with the premise of the OP, local manufacturing has a lot of benefits.
x0x0 12y ago
If you haven't heard of Schiit, it's a fascinating story of a small company that builds high end and low end headphone amps/dacs, for much less money than competitors. In CA. One of the two founders wrote an online book about it here: http://www.head-fi.org/t/701900/schiit-happened-the-story-of-the-worlds-most-improbable-start-up http://www.head-fi.org/t/701900/schiit-happened-the-story-of...
- makomk 12y ago
  There's the minor issue of them not actually testing their designs perform as expected, at least one amp damaging some people's headphones as a result, and them denying the issue existed: http://nwavguy.blogspot.com/2011/07/banned-at-head-fi.html http://nwavguy.blogspot.com/2011/07/banned-at-head-fi.html Though that's not really a problem with manufacturing in the US, it's more of a general audiophile industry problem.
shanecleveland 12y ago
What exists to help entrepreneurs find/pair up with manufacturers locally/domestically? Opportunity there? I work for a US manufacturer (non-tech related), and we frequently get inquiries from folks about making their products. We are primarily focused on producing our own products, but it is nice to fill gaps in production and explore new partnerships. But we don't actively pursue this, and I have no idea how they find us. WOM and random online databases, is my guess.
- aslewofmice 12y ago
  thomasnet.com was decent for finding leads, but we weren't able to find a suitable manufacturer to get past prototype
  - shanecleveland 12y ago
    Interesting. I see that our company is listed in there. That's not voluntary on our part, which may result in a lot of unsuccessful leads.
fest 12y ago
I read the story how "Schiit happened" the other day and the last part of chapter 8 also touched this subject: http://www.head-fi.org/t/701900/schiit-happened-the-story-of-the-worlds-most-improbable-start-up/495#post_10375915 http://www.head-fi.org/t/701900/schiit-happened-the-story-of...
soggypenny 12y ago
We (https://supplybetter.com https://supplybetter.com) get asked the China vs. US question all the time and my answer is usually "your mileage may vary". It's a great post by the Flow team, but where to source your manufacturing supply chain is highly dependent on the type of product being built. For instance, the product and vendor certifications required for aerospace, defense, and medical industries are a much larger influencer of where to setup a supply chain than in the consumer hardware space. If you're developing a high-end consumer office product like the Flow team, then it probably makes sense to do the majority of the development domestically. The Bay Area is a great place to prototype hardware, and there's a large fraction of potential beta testers and software talent for them to tap. But like I said, YMMV.
Animats 12y ago
Local manufacturing only works if you're in a place where there's a lot of manufacturing, or your product doesn't really require much manufacturing capability. You many need to be in a place where there's someone down the street with a 20 ton press. If you need a 20 ton press, you need to be in a place where at least three shops have one. If you're making things which are basically PC boards, manufacturing is not too hard. Making boards and soldering parts onto them has a well developed workflow. There are lots of board houses and board-assembly services. If you design for what a pick and place machine can do easily, things usually go well. Hint: production is surface mount today. This is a pain for people who are used to prototyping with through-hole parts. Kickstarter-class startups seem to have excessive problems with making cases and panels. Tooling for injection-moulded plastic is difficult and expensive. Once you get it right, the parts just fly out of the machines at a few cents per part. The production process is just getting warmed up on a run of 10,000 parts, and few Kickstarters get there. Basic truth about manufacturing: most of the processes are really cheap if you're making enough items, and far more expensive for short runs.
- jacquesm 12y ago
  20 Ton presses are $200: http://www.harborfreight.com/20-ton-shop-press-32879.html http://www.harborfreight.com/20-ton-shop-press-32879.html Agreed on the injection moulding, that's expensive (and rightly so). But a simple press is not going to be a showstopper on any budget. If you're going to do sheetmetal manufacturing there are a ton of companies waiting to get your business just about anywhere up to a reasonable volume. Manufacturing hasn't completely moved out (yet). Small series, prototypes, even your first batch of some product can usually (definitely not always) be manufactured locally in just about any country.
dmritard96 12y ago
For some it will make sense. Try iterating on your board designs in CA - I mean like testing your pcb antennas, component swap out etc during development and spend weeks waiting around for you pcbs. Being in shenzhen you can get 1 or 2 day turn around... Even if you don't manufacture there, development is pretty awesome
kweks 12y ago
A lot depends on the product you are manufacturing, and the relationship you have with the manufacturer. Business in China operates via 'Guanxi' [ http://en.wikipedia.org/wiki/Guanxi http://en.wikipedia.org/wiki/Guanxi ] - which losely translates to a heirachy of: family first, then friends, then people from the same town, then of the same province, then of the same country, and then, the lowly gweilo / white devil. If you don't have previous contacts or a solid history in China, you fall right to the bottom of the stack - meaning you'll have a bad price, bad negotiation point, low priority in the factory, and ripe for being 'disappointed' with your experience. Likewise, many people who outsource manufacturing to China hop on alibaba, find a manufacturer with a low per item cost, and think they've cracked East/West trading. Not so. There are many, many hidden gotchyas involved - ranging from child labor, poor conditions, the 'factory' you chose isn't even the manufacturer, to hiden nasties in pricing / shipping, etc. There is a large misconception that products from China are junk. This is partly true. Some products from China are junk - simply because the Chinese can and will manufacture according to the resale price specified by the buyer - that is to say, you. When you buy a $1 item from WalMart, and you know its price 'should be' $5, you contribute actively to this problem. Want to see a great product made in China? Look at your Apple device. If you're contemplating doing business or manufacturing in China, there's a few basic things to do to avoid common pitfalls: 1. Visit your factory. Fly over, arrange meetings with 4 - 5 similar factories, and inspect them all. If something seems suspicious, return unannounced to see the reality. This is really, really important. Meeting clients and relations are very, very important in China. 2. Understand / learn how Chinese do business. If you don't have local contacts, consider finding a local broker to do the hard work on your behalf, and negotiate Chinese/Chinese (See Guanxi above). Due diligence here is important. 3. Plan for 12-month development cycles. If you're in a big factory, you're competing with their big clients. Most manufacturers plan in 24-month development cycles, so plan at least in 12 month cycles. This is just common sense. 4. Factor in shipment prices / shipment terms. Learn the difference between EXW / FOB / etc. These make huge diffrences to your bottom line. Remember for production runs to have shipping insurance / liability insurance. 5. Don't fall into legal issues: ensure you know what and how your goods are manufactured. Making toys / figurines? Make sure your factory is certified to make them, or be able to live with the consequences. 6. Be realistic about your MOQ. If you're looking for 5 items, don't look to China - you'll only waste your time and the factory's time. Like any decision in business, you need to consider what works best for your business. Source: 30+ years of manufacturing hardware [locks, tools, etc], toys [figurines, plush, etc], and electronics [PCBs, injection moulded cases etc] in China.
allworknoplay 12y ago
Local manufacturing is the only way to go if you're developing a reasonably complicated product and don't have a team that's super experienced with international manufacturing. I was the technical co-founder of Brightbox (brightboxcharge.com, us founders are no longer affiliated), we were based in NYC and could never have done the prototyping and iteration required if we couldn't clap eyes on hardware constantly. There's plenty of great manufacturing in the US. Since I'm a software guy, we had one company build us a prototype and handle industrial design (Tomorrow Lab, http://tomorrow-lab.com/; http://tomorrow-lab.com/; Pepin is a friend) and found an integrated mechanical engineering and manufacturing firm just over the river in NJ (Tech Products, techproductsco.com; Bob is an awesome human being) for the final product and first few real runs. Having an actual partner we could meet with all the time, understand clearly, and iterate with was incredibly valuable for us. Plus obviously there are important technical integrations (I was writing all the software myself while the ME and EE was happening) and multiple suppliers, and if you're in charge of managing a few bespoke engineers who sometimes don't see eye to eye, you'd better be able to work with them all carefully and personally.

68 points by twald 12y ago

Great points - thanks Tobias!

I've been thinking about this just the other day. The context then was locally grown weed in the US and the implications of that for farmers in South America. Wouldn't we be better off distributing the "workload" evenly across the planet?

delecti 12y ago
International drug trafficking laws are almost definitely a bigger factor in the consideration of where to grow legalized medicinal/recreational cannabis than cost, especially considering it grows so readily in so many different climates.
davidw 12y ago
Probably not: http://en.wikipedia.org/wiki/Comparative_advantage#Ricardo.27s_example http://en.wikipedia.org/wiki/Comparative_advantage#Ricardo.2... To take your pot example, there are places where it grows well, and places where it's not so easy to grow. It's better for everyone if it's grown in the easy places, and traded for goods that are more easily produced in the 'not so easy' places. That's the theory, at least. Naturally, that specific example also has a lot of legal issues surrounding it!

amckenna 12y ago

I have backed a half dozen Kickstarters and Indigogos that were hardware based (consumer electronics) and chose to manufacture in China. Seeing the time and money they spend in order to fly someone over to China to troubleshoot and issue with manufacturing or testing has left me wondering - why do small startups choose to manufacture in China to begin with? > Is it because they assume that despite the initial cost over local manufacturing the investment will be worth it if their product takes off? > Is it because everyone else seems to do it? > Or maybe because there aren't any good local places to have electronics manufactured? This I would find a little hard to believe given the concentration of hardware startups in SF, NYC, SEA, which have local hardware manufacturers.

JonFish85 12y ago
My guess would be that it starts with the assumption that "China is cheaper", and that makes for an easy story to tell to investors "we're making XYZ in China!". The second you say "we're making it in Texas", it opens you up to questions of "isn't that expensive?" On the face of it, I'm sure the up-front quote from a Chinese manufacturer is cheaper than something in the USA. Where costs skyrocket is when something goes wrong: the parts are wrong, the products aren't ready on time, something was done incorrectly, etc. If you don't have a manufacturer you can trust, one of those will happen, and is almost impossible to diagnose from the US. Even if all of that works out, there are other things to consider: costs (and time) of shipping. Want it in a week? Well, get ready to pay. Even if you ship it air freight consolidated, it's still not cheap. If you want it on a boat, there's a good chunk of time to think about. On top of THAT, there are all sorts of fun import tariffs and such to deal with (at least on the US side). Get ready to front money for an import bond, for starters! And there may already be some hidden costs of importing into the US: if you use some TI parts that have to be imported into China, you're going to be paying taxes twice. A lot of those things are likely not looked at in the low quote from a Chinese manufacturer. They're costs that can/will show up later along the line.
JonFish85 12y ago
My guess would be that it starts with the assumption that "China is cheaper", and that makes for an easy story to tell to investors "we're making XYZ in China!". The second you say "we're making it in Texas", it opens you up to questions of "isn't that expensive?" On the face of it, I'm sure the up-front quote from a Chinese manufacturer is cheaper than something in the USA. Where costs skyrocket is when something goes wrong: the parts are wrong, the products aren't ready on time, something was done incorrectly, etc. If you don't have a manufacturer you can trust, one of those will happen, and is almost impossible to diagnose from the US. Even if all of that works out, there are other things to consider: costs (and time) of shipping. Want it in a week? Well, get ready to pay. Even if you ship it air freight consolidated, it's still not cheap. If you want it on a boat, there's a good chunk of time to think about. On top of THAT, there are all sorts of fun import tariffs and such to deal with (at least on the US side). Get ready to front money for an import bond, for starters! And there may already be some hidden costs of importing into the US: if you use some TI parts that have to be imported into China, you're going to be paying taxes twice. A lot of those things are likely not looked at in the low quote from a Chinese manufacturer. They're costs that can/will show up later along the line.

jtoberon 12y ago

In NYC, there's http://www.refactory.co/ http://www.refactory.co/. I used to work with these guys, and they're great.

jtoberon 12y ago

In NYC, there's http://www.refactory.co/ http://www.refactory.co/. I used to work with these guys, and they're great.

jtoberon 12y ago

In NYC, there's http://www.refactory.co/ http://www.refactory.co/. I used to work with these guys, and they're great.

Animats 12y ago

Local manufacturing only works if you're in a place where there's a lot of manufacturing, or your product doesn't really require much manufacturing capability. You many need to be in a place where there's someone down the street with a 20 ton press. If you need a 20 ton press, you need to be in a place where at least three shops have one. If you're making things which are basically PC boards, manufacturing is not too hard. Making boards and soldering parts onto them has a well developed workflow. There are lots of board houses and board-assembly services. If you design for what a pick and place machine can do easily, things usually go well. Hint: production is surface mount today. This is a pain for people who are used to prototyping with through-hole parts. Kickstarter-class startups seem to have excessive problems with making cases and panels. Tooling for injection-moulded plastic is difficult and expensive. Once you get it right, the parts just fly out of the machines at a few cents per part. The production process is just getting warmed up on a run of 10,000 parts, and few Kickstarters get there. Basic truth about manufacturing: most of the processes are really cheap if you're making enough items, and far more expensive for short runs.

Animats 12y ago

organsnyder 12y ago

I know some people that started a company that makes GPS receivers for avionic use (any pilot will likely be able to name them). I heard them explain their rationale for manufacturing in China, and IIRC it wasn't related to cost—in fact, they said that cost was comparable to domestic production. Rather, the main advantage was the ability to source components, such as the Apple-specific connectors they have to use. Of course, there are difficulties inherent to working with a manufacturer that's half a world away (miscommunications, timezones, travel, etc.), but they determined that it was still the best way to manufacture their product. China especially has manufacturing centers where just about any component can be sourced within a few hours, rather than waiting for parts to be shipped from around the country (if not the world). China has a huge edge in electronics manufacturing, and it's not because of cost (anymore). If we ever hope to compete, we'll likely need to promote such a domestic manufacturing ecosystem via extra-market means.

organsnyder 12y ago

jtoberon 12y ago

In the NYC area, there's http://www.refactory.co/ http://www.refactory.co/. I used to work with these guys, and they know their stuff. They argue in favor of both designing and manufacturing locally. A quote from their blog: "If you kind of just prototype something and bring it to China expecting that then thousands of these things can be made easily without understanding the processes and materiality of manufacturing, that is gonna be a big problem for you. What works is when the design process is iterated alongside the manufacturing process."

sabalaba 12y ago

Those costs in China are too high. I've lived in Shenzhen for under $70 USD / week in a hostel. If you eat out every meal for 100 RMB, you're not even going to be spending $350 a week for meals. While I agree that if you can do it locally, do it locally, but those numbers for the cost of living and travel in China are inaccurate.

hawkice 12y ago
For context: 100RMB is about $16, so you better really be enjoying those meals you're eating 3 times a day, because wowzers that's upscale.
- sabalaba 12y ago
  Exactly, and they were suggesting that it would be $200 USD more than that per week!

tlb 12y ago

I agree that manufacturing in-house is great for hardware startups. I've seen it work for several. The insight they gain into the details of design, and the speed they can iterate at, more than make up for higher local labor costs.

twald 12y ago
hey, author here. we're thinking about sharing stories about hardware startups that decided for the local manufacturing route including price differences, manufacturing tips, gm contacts, etc. If somebody is interested or wants to get involved, please send me an email: tobias@senic.com
- Gurkenmaster 12y ago
  I'm not from http://schiit.com/ http://schiit.com/ but they also went with local manufacturing route and their prices are quite competitive.

iandanforth 12y ago

I'll add one word of caution, I've looked into getting a product made that had a textile component and a separate product that had a leather component. Precision, automated stitching is either totally absent or completely unaffordable in the US. There are great shops India and China that can do laser cutting and CNC stitching super cheaply and if you want to make an inexpensive product to tight tolerances this is exactly what you want. Unfortunately much of the consumer level textile industry has completely left the US. Same with leather. You can find boutique manufacturers, luxury manufacturers, and aerospace type companies, but the sweet spot of high quality and a good price just doesn't exist here. (I actually hope I'm wrong about that but I did a lot of searching before abandoning a couple projects.)

JabavuAdams 12y ago
That sounds like a business opportunity.
justizin 12y ago
Read your own comment. You're basically saying that laser cutting and CNC stitching - operations performed by robots based on files uploaded by computers - are done better in China than here. That's like someone in China saying that we are better at standing. People on both sides of the ocean probably do just fine at both. ;)
- bsder 12y ago
  > You're basically saying that laser cutting and CNC stitching - operations performed by robots based on files uploaded by computers - are done better in China than here. Not done better. They're not done here AT ALL. That's a big difference.
  - justizin 12y ago
    That's a bold claim. Laser cutting and CNC stitching are not done "here" which I guess is outside China, at all, by anyone?

kcorbitt 12y ago

> If you’re based in Germany, there is a 7 hour time difference to east Asia, if you’re based in California, it is 16 hours. No, in California, it's 8 hours (in the other direction). It doesn't make sense to talk about more than 12 hours of time offset except if you really care about aligning the calendar day instead of just the hour of the day.

nandemo 12y ago
Around weekends, it does matter. E.g., when it's Friday afternoon in California, it's already Saturday in East Asia.

kosma 12y ago

As usual, the answers are not simple. At our last hardware startup we ended up manufacturing the mechanical parts (injection moulded case, springs, etc.) in China via our industrial design company but did PCBs, assembly and packaging locally in Poland. We simply didn't have enough prior experience to design a foolproof assembly process (which involved flashing, testing on GSM/GPS repeaters, etc.). We are considering moving the to production to China, but only once we solve any critical hardware bugs, smooth out the wrinkles in the flashing/testing process and reach sufficient quantities. With small orders of 100s of units and unexpected disasters like resistors tombstoning due to small errors in solder paste stencils, a five minute call to the manufacturing plant beats cross-timezone, cross-language and cross-culture debugging every single time.

thruflo 12y ago

As a founder of https://www.opendesk.cc https://www.opendesk.cc I think all startups should consider local manufacturing -- when they're fitting out their work space. </tenuous plug>

terramars 12y ago

Totally agree with this. Have a friend who had a successful kickstarter and decided to go with China - he shipped a year late after spending most of his time in Shenzen and going through multiple suppliers. In the interim, they lost the first mover advantage and there are now multiple similar products on the market. If you're not already well established and making a massive amount of something, it's a really bad idea to outsource.

seddona 12y ago

On the electronics side we (https://circuithub.com/ https://circuithub.com/) have spent a lot of time on automation to bring US manufacturing costs inline with China. I definitely agree with the premise of the OP, local manufacturing has a lot of benefits.

x0x0 12y ago

If you haven't heard of Schiit, it's a fascinating story of a small company that builds high end and low end headphone amps/dacs, for much less money than competitors. In CA. One of the two founders wrote an online book about it here: http://www.head-fi.org/t/701900/schiit-happened-the-story-of-the-worlds-most-improbable-start-up http://www.head-fi.org/t/701900/schiit-happened-the-story-of...

makomk 12y ago
There's the minor issue of them not actually testing their designs perform as expected, at least one amp damaging some people's headphones as a result, and them denying the issue existed: http://nwavguy.blogspot.com/2011/07/banned-at-head-fi.html http://nwavguy.blogspot.com/2011/07/banned-at-head-fi.html Though that's not really a problem with manufacturing in the US, it's more of a general audiophile industry problem.

shanecleveland 12y ago

What exists to help entrepreneurs find/pair up with manufacturers locally/domestically? Opportunity there? I work for a US manufacturer (non-tech related), and we frequently get inquiries from folks about making their products. We are primarily focused on producing our own products, but it is nice to fill gaps in production and explore new partnerships. But we don't actively pursue this, and I have no idea how they find us. WOM and random online databases, is my guess.

aslewofmice 12y ago
thomasnet.com was decent for finding leads, but we weren't able to find a suitable manufacturer to get past prototype
- shanecleveland 12y ago
  Interesting. I see that our company is listed in there. That's not voluntary on our part, which may result in a lot of unsuccessful leads.

fest 12y ago

I read the story how "Schiit happened" the other day and the last part of chapter 8 also touched this subject: http://www.head-fi.org/t/701900/schiit-happened-the-story-of-the-worlds-most-improbable-start-up/495#post_10375915 http://www.head-fi.org/t/701900/schiit-happened-the-story-of...

soggypenny 12y ago

We (https://supplybetter.com https://supplybetter.com) get asked the China vs. US question all the time and my answer is usually "your mileage may vary". It's a great post by the Flow team, but where to source your manufacturing supply chain is highly dependent on the type of product being built. For instance, the product and vendor certifications required for aerospace, defense, and medical industries are a much larger influencer of where to setup a supply chain than in the consumer hardware space. If you're developing a high-end consumer office product like the Flow team, then it probably makes sense to do the majority of the development domestically. The Bay Area is a great place to prototype hardware, and there's a large fraction of potential beta testers and software talent for them to tap. But like I said, YMMV.

Animats 12y ago

jacquesm 12y ago
20 Ton presses are $200: http://www.harborfreight.com/20-ton-shop-press-32879.html http://www.harborfreight.com/20-ton-shop-press-32879.html Agreed on the injection moulding, that's expensive (and rightly so). But a simple press is not going to be a showstopper on any budget. If you're going to do sheetmetal manufacturing there are a ton of companies waiting to get your business just about anywhere up to a reasonable volume. Manufacturing hasn't completely moved out (yet). Small series, prototypes, even your first batch of some product can usually (definitely not always) be manufactured locally in just about any country.

dmritard96 12y ago

For some it will make sense. Try iterating on your board designs in CA - I mean like testing your pcb antennas, component swap out etc during development and spend weeks waiting around for you pcbs. Being in shenzhen you can get 1 or 2 day turn around... Even if you don't manufacture there, development is pretty awesome

kweks 12y ago

A lot depends on the product you are manufacturing, and the relationship you have with the manufacturer. Business in China operates via 'Guanxi' [ http://en.wikipedia.org/wiki/Guanxi http://en.wikipedia.org/wiki/Guanxi ] - which losely translates to a heirachy of: family first, then friends, then people from the same town, then of the same province, then of the same country, and then, the lowly gweilo / white devil. If you don't have previous contacts or a solid history in China, you fall right to the bottom of the stack - meaning you'll have a bad price, bad negotiation point, low priority in the factory, and ripe for being 'disappointed' with your experience. Likewise, many people who outsource manufacturing to China hop on alibaba, find a manufacturer with a low per item cost, and think they've cracked East/West trading. Not so. There are many, many hidden gotchyas involved - ranging from child labor, poor conditions, the 'factory' you chose isn't even the manufacturer, to hiden nasties in pricing / shipping, etc. There is a large misconception that products from China are junk. This is partly true. Some products from China are junk - simply because the Chinese can and will manufacture according to the resale price specified by the buyer - that is to say, you. When you buy a $1 item from WalMart, and you know its price 'should be' $5, you contribute actively to this problem. Want to see a great product made in China? Look at your Apple device. If you're contemplating doing business or manufacturing in China, there's a few basic things to do to avoid common pitfalls: 1. Visit your factory. Fly over, arrange meetings with 4 - 5 similar factories, and inspect them all. If something seems suspicious, return unannounced to see the reality. This is really, really important. Meeting clients and relations are very, very important in China. 2. Understand / learn how Chinese do business. If you don't have local contacts, consider finding a local broker to do the hard work on your behalf, and negotiate Chinese/Chinese (See Guanxi above). Due diligence here is important. 3. Plan for 12-month development cycles. If you're in a big factory, you're competing with their big clients. Most manufacturers plan in 24-month development cycles, so plan at least in 12 month cycles. This is just common sense. 4. Factor in shipment prices / shipment terms. Learn the difference between EXW / FOB / etc. These make huge diffrences to your bottom line. Remember for production runs to have shipping insurance / liability insurance. 5. Don't fall into legal issues: ensure you know what and how your goods are manufactured. Making toys / figurines? Make sure your factory is certified to make them, or be able to live with the consequences. 6. Be realistic about your MOQ. If you're looking for 5 items, don't look to China - you'll only waste your time and the factory's time. Like any decision in business, you need to consider what works best for your business. Source: 30+ years of manufacturing hardware [locks, tools, etc], toys [figurines, plush, etc], and electronics [PCBs, injection moulded cases etc] in China.

allworknoplay 12y ago

Local manufacturing is the only way to go if you're developing a reasonably complicated product and don't have a team that's super experienced with international manufacturing. I was the technical co-founder of Brightbox (brightboxcharge.com, us founders are no longer affiliated), we were based in NYC and could never have done the prototyping and iteration required if we couldn't clap eyes on hardware constantly. There's plenty of great manufacturing in the US. Since I'm a software guy, we had one company build us a prototype and handle industrial design (Tomorrow Lab, http://tomorrow-lab.com/; http://tomorrow-lab.com/; Pepin is a friend) and found an integrated mechanical engineering and manufacturing firm just over the river in NJ (Tech Products, techproductsco.com; Bob is an awesome human being) for the final product and first few real runs. Having an actual partner we could meet with all the time, understand clearly, and iterate with was incredibly valuable for us. Plus obviously there are important technical integrations (I was writing all the software myself while the ME and EE was happening) and multiple suppliers, and if you're in charge of managing a few bespoke engineers who sometimes don't see eye to eye, you'd better be able to work with them all carefully and personally.