First, let me congratulate YCombinator and the founders they’ll fund, for putting forth an initiative to support the until-very-recently-unsexy segment of hardware startups. This is great news for the anyone and everyone in the “Internet of Things” and “Wearable” and other related segments. I myself am a mentor to Highway One (another such incubator) and know there are others in the space as well.
By way of background, for those who don’t know me, I’ve been involved in hardware startups since 1999, as a cofounder of Mediabolic (acquired by Rovi in 2006), VP of Products for Sling Media (acquired by DISH in 2007), and a former product consultant to great companies like Bug Labs, Boxee, D-Link, DivX, Dropcam, and many others (I’m getting all nostalgic here!). Having spent almost 20 years building mostly successful gadgets of one kind or another, I thought it might be good to chime in on the topics raised by some of the YC HW founders, and add my own “recipes for success” to share with other founders.
Hardware is still very hard.
Seems amazing to me that, while things have gotten notably easier, anyone could possibly come to the conclusion that hardware startups are easy. You need expertise in so many fields, from distribution to marketing to manufacturing to support, and have so little wiggle room (more in a moment), it’s almost mind-boggling. And when things go wrong (more on that coming too), and things always go wrong, it’s entirely possible that there’s not enough resources/knowledge/etc to ever recover. TL;DR: there’s no pivots or growth hacking in hardware.
HW requires some fundamentally different skill sets than Software.
In software/app/web startups, there are a lot of skills that transfer easily, whether across platforms, segments, borders, etc. This is rarely true for a hardware startup where, for example, a very experienced customer acquisition/marketing specialist may find themselves completely in unfamiliar territory when building a distribution strategy. And from what I’ve seen across my career, not a single HW startup comprised of highly competent founders with no hardware background has shown tremendous success. TL;DR: make sure you have domain experts in your team.
You need to know what can go wrong with HW.
I’ll address the “what can go wrong” topic one more time below, but I guess I can’t emphasize enough: more can and will go wrong than you’ve ever thought possible. Ever have your manufacturer swap out specified memory chips in your device, not tell you, and not QA them prior to shipping to customers? Check. Ever have your CM have their assets seized mid-way through a production run, putting all your pre-paid inventory into a massive governmental lawsuit? Check. Ever have your rep tell you the design is being met to 100%, only you can see with plain eyes that they are cutting circles into the mold by hand? Check. Ever have your packaging fail to meet a retailers’ drop-test, one week after loading up endcaps? Check. Ever have a chipmaker massively exaggerate a platform’s capabilities, and not be able to learn the truth until 60 days before shipping? Check. TL;DR: start with the expectation that some unimaginable thing will go awry; never forget it.
You should raise more money / funding than you plan for.
Here’s a mind-boggler: success can bankrupt your HW startup as easily as failure. How, you might ask? Because with every month/quarter’s sales, you must order and plan for the next month, and do so without necessarily seeing revenue. Basically if you’re getting orders from retailers, you need to plan for growth. And order the parts for future orders. And what happens when volume increases? So do cost of goods (even if the per-unit cost is dropping due to scale). I’ve seen this multiple times before: companies scramble to project ahead, order either too much or too little inventory, and run out of money along the way. TL;DR: you will incur costs prior to revenue and need oodles of cash on hand to manage!
HW requires a deeper understanding of customers / markets.
Its fine/great to start a software company and slowly learn the features that drive adoption, or discover hidden market opportunities. The ability to tweak products and meet different opportunities is the beauty of the modern startup. But this doesn’t work in hardware – you can’t add a button, change a component, etc to a product in the market. Sure if it’s a “headless” device (like a Slingbox or Dropcam) you can always improve the end-user software experience. But need more memory, or an extra port of some kind? Welcome to 2.0. TL;DR: there’s no such thing as a lean hardware startup.
You need a better “crystal ball”.
For the most part, since you are starting 6-18 months away from first customer ship, you need to pull a Wayne Gretzky. It’s not about where the puck is, it’s about where is the puck going? What technology/infrastructure will change in the interim? A product I had designed for a huge consumer electronics company won accolades and awards from retailers, industry professionals, etc. And then the world turned HD in an amazingly short window, and the product got killed. Done, game over. TL;DR: your HW startup vision should make bankable assumptions about the world 18-36 months from now.
Solid backup plan for when things go wrong.
So I’ve identified a plethora of things that can go wrong. Now what happens when something delays your device by 4 months (a fairly reasonable timeframe, if not longer)? What do you do with the ad inventory you’ve pre-agreed to? How about your marketing team you’ve been recruiting and hiring? Or the conference you paid to launch at? How do you keep up morale? TL;DR: have a plan in place, from day one, assuming a multi-month delay will occur at some point prior to launch.
It’s a very exciting time for hardware entrepreneurs, and it’s very true that the resources available to them today are far superior than anything before. But in all candor, most of this new support and infrastructure doesn’t actually fix the fundamental struggles that come along with making things. I’m looking forward to the next crop of must-have gadgets, and hope the above tips and thoughts help any readers who come along.