Hardware startups are not exactly the same as your usual software startup, and the conversation at the Generator v 1.2 conference in Shenzhen has been a great place to know where the differences lie.
The Hardware Renaissance is a positive context for hardware startups
This trend is powered by new technologies of fast prototyping (3D printers, laser cutters), new ways to get funding with crowdfunding (top Kickstarter projects are almost all hardware stuff, from OUYA to Oculus), and a first series of success stories (Makerbot bought $600M by Stratasys, Nest bought $2bn by Google, Oculus bought $2bn by Facebook).
The business model of hardware startups is quite different from software startups. Most successful ventures will operate a combination of hardware and software, both to avoid commoditization and to keep the possibility of selling also part of the software. As hardware requires a lot of cash, margins should also be high in order to reach profitability, and high margins are better found with niche products with a bold and great vision, typically such as the Oculus headset.
Shenzhen, a top hub for hardware startups thanks factories and entrepreneurship culture
Back to Shenzhen, the talks of the Haxlr8r startups and alumni also help to understand why Shenzhen is the new Mecca for these makers and hardware startupers.
- First, for the cost of labor. Hiring a decent engineer in the Silicon Valley will require about $120k/y, doing so in Singapore about $70k/y, while in Shenzhen, it’s possible to find a good one at $15k/y.
- Then, because of the proximity of factories. Shenzhen has long been the “factory-city of the world”, and although it’s changing fast (see our report on the Shenzhen startup ecosystem), it is still home to Foxconn’s 400 000 employee factory, where most smartphones are still made. A myriad of smaller, easier to work with factories are also in Shenzhen.
- Last but not least, thanks to a culture of entrepreneurship and creativity proper to Shenzhen. The city and area are famous for shanzhai, an art of counterfacting products in a smart way that never starts from scratch and adds features users actually want (check out this interesting mouse/phone, for instance). Shenzhen is also a city of migrants, which has gone from the mythical fishing village with 300 000 inhabitants three decades ago to the gigantic 15m metropolis it is now.
Hardware startups 101: do’s and dont’s
So if you want to build your own hardware startups, here are a few tips from the speakers:
- Most important: get in touch with a factory FIRST, and co-create the product with them. There’s no planning/designing on paper and then ordering any factory to produce your ideas. You need to go hand in hand with them. Cyril stresses that unlike the Valley, where an engineer will typically spend 80% of his time on software and can work in a relatively isolated way, hardware startups are 80% about operations, so the engineers need to be in the factory, and vice versa.
- Then, the design of your product must do with the right components. Are they the best of the market? Have past versions changed a lot (and, for instance, can’t be found anymore for replacement)? Is there any supply of it? Can the factory arrange for some customization of it? These are as many factors to bear in mind.
- The business model should also be precise and realistic. Cash is burnt fast in most startups, but it’s even truer in the hardware world where, says Bunnie Luang, “a mistake would cost about $10k/day”. Anticipation of the cashflow is key as prototyping and then producing may take each 3 months to 2 years. You cannot build something credible as fast as in the software world, so a good question would be to ask yourself if your idea would still be innovative 2-5 years down the road.
- Be memorable, because you’re not going to spend years (not months), money (yours, your family’s, your Kickstarter’s backers, your investors’), and brainpower to mimic something existing or sell 10k units. Be bold, be it only for people to know exactly what you mean. Makerbot, for instance, is both an awesome brand name (“a robot which makes things”) and, at the time, a bold vision of a world where anyone could print affordably stuff and build actual things at home.
As Haxlr8r is opening applications for a 5th batch of 10-ish startups this July, their last piece of advice is to avoid the “bad” -ware you can see below. From vaporware (cool idea but not doable) to boreware (cool idea but doesn’t manage to change an habit), you will see that there is a LONG and HARD way to go to have a successful hardware startups. At least you now know where to go and who to get in touch with.
In any case, should you want to launch a hardware startup, there’s probably no greater place than Shenzhen to do it, and Cyril Ebersweiler and its general partner Benjamin Joffe are two persons you would want to meet and discuss with.
You will find the full Slideshare from Haxlr8r on Hardware Startups here