Spark provides the tools to connect everyday electronics to the internet over Wi-Fi. After a failed Kickstarter campaign for internet-connected lighting, the team came back with their second product, the Spark Core, and raised over $500,000. We sat down with CEO and co-founder Zach Supalla to discuss their time in a hardware accelerator in China, building a community around Makers and bouncing back from failure.
My background is mostly in business. When I was in college, I studied a mix of computer science, cognitive science, psychology and linguistics; I got into product design too late to realize I should have been an engineer. My senior year I took a design class and I thought, “This is really cool, but it’s my senior year.”
After I graduated, I was trying to figure how to work my way back to product design. I went to work for McKinsey & Company; I was a management consultant for three years, where I primarily did operations work. Much of that work was somewhat engineering related; it was a lot of process improvement and product development. I started doing some design-type work at McKinsey, but it was things like changing a glass cup from frosted to clear—which saves six cents per glass—and because the company makes hundreds of thousands of these things, I saved them $2 million a year. There was a lot of money at stake, but it didn’t feel like I was changing the world.
I ended up going to get a dual degree: an MBA and a Masters in Engineering Management in Design and Operations from Kellogg School of Management at Northwestern. While I was there, that’s when I really started to get into entrepreneurship. I spent my summer internship working at Groupon and worked on a couple of startups in the e-commerce world before jumping in on Spark, which was originally called Switch. When I was in business school, I wanted to make a prototype but I didn’t know how. As a result, I started Googling and watching YouTube videos, and over the course of a few months, went from nothing to a business plan and a working prototype that I built myself. From then, it’s been a gradual evolution from that first prototype into what we have today.
First, I bought an Arduino. I think the Arduino platform and community is such an incredible resource because it makes this stuff much easier and more accessible for someone without a pure engineering background. I knew enough computer science to be dangerous from studying it a while, but I didn’t get that far in my coursework, I used Code School and Codecademy to learn a bit more. I used Codecademy’s Code Year program. I bought books, I watched screencaps and spent a lot of my free time reading and learning and absorbing everything I could.
When we first started Spark, the company was about internet-connected lighting. The inspiration for the Socket was my Dad, who’s deaf. He had this particular problem that the he was very difficult to reach when he was at home because he would take his phone out of his pocket, and then my Mom couldn’t get ahold of him, since he couldn’t feel his phone vibrating. When I was a kid, he had these systems that would flash when he got a phone call on his TTY, or when someone rang the doorbell. He still has those, but they don’t work with a cell phone. Now my parents communiciate with text messages, but there is really no way to get in touch with my dad if his phone isn’t on him. Even when it is, sometimes he doesn’t feel the vibration. My Mom was struggling because she wanted to be able to get ahold of him, and that was the beginning of the Spark Socket.
I wanted to build an internet-connected light that could flash when he received a text message — really simple and straightforward. But as I started building and started thinking about all the other things that it could be used for, I found much broader use cases. Not just for the deaf, but for everybody. Your lights should turn on when the sun goes down, and turn off when the sun comes up. They should know when you’re going to bed and when you’re getting up.
We should be getting to the point where a “smart home” isn’t just about remote control; it’s about real, automated behaviors. It should do what you expect it to do. That was the beginning of Spark — we created what became the Spark Socket. At first it was a wall socket, and then it was a light switch. After that it was a wall socket again and then it turned into a light bulb socket, and that’s where we settled it. I hired a designer to do industrial design. I designed the circuit and the software myself and then started to grow the team when I moved to Minneapolis. Then there were three of us when we launched the Spark Socket.
The Socket was a great learning experience for us, because I think it was a cool product but it was poorly timed. It came out about six weeks after LIFX which was massively successful on Kickstarter and right after the Philips Hue launched in the App Store. When I first started working on the Spark Socket, people told me I was crazy coming up with this stuff, but by the time we actually launched—because we didn’t move quickly enough—people were telling me it was old news. They would say, “People are already doing this, how is it any different?” We were too slow. We came out of it and said, “Okay, well we need to figure out what to do next.”
We got accepted into HAXLR8R, which was a great revitalization for us. One of the most important things was that we needed to learn to move more quickly, and we felt like China would be a really important part of that. We would go to China and learn how to apply Lean Startup principles to a hardware startup. We needed to iterate quickly, move fast, and get to market quickly. While we were in China we were still originally planning on making the Socket; we were trying other ways to get to market more quickly. We were getting into LEDs, which is a massive industry.
The original Spark Core was a part of the design of the Spark Socket and we pulled it out. We basically said, “What if we just went to other lighting companies and we sold them the technology they need to make internet connected lights?” Instead of us needing to make a thousand different styles of lamps, we could go to the lamp manufacturers and say, “Here’s a chip, this makes your lamp internet-connected.” We made the chip a little bit more stand-alone than what it was from the Socket, and that was the first iteration of the Spark Core. Then we looked at the chip and thought, “This is way more broadly applicable than just lighting.” Sure, lighting is a great use case, but there are a million other things it could be used for; there was nothing about it that was specific to lighting. That’s when we decided to launch it on Kickstarter as a development platform and a tool that people could use to make internet-connected products more quickly and more easily.
“There was a lot of money at stake, but it didn't feel like I was changing the world.”
One of the things that we learned about Kickstarter—now that we’ve failed once and succeeded once—is that Kickstarter is a very unique crowd. First of all, the people on Kickstarter are phenomenal. They’re so engaged and so supportive. It’s not just a random cross section of the world; it’s a very specific group of people. I think that they tend to be nerdier and they tend to be tech-focused. We were originally planning on going straight B2B with the company. We were going to sell the Spark Core to hardware manufactures with no intention of going into the developer or Maker world. Then we saw some hardware projects doing really well on Kickstarter. It seems like people on Kickstarter get really excited about this stuff. Since we’d done Kickstarter before, we knew how to run a good campaign. We were going to make the product anyways because we were definitely going to go B2B with the product. Because of that, we decided to see how it would fly with the Maker community. That made us change some of our perspectives. We decided to open source everything because we thought that was important and we all believed in the open source movement.
In the end, it turned out that we scratched an itch in the Maker community that has been around for a while. It’s surprisingly difficult to make a Wi-Fi connected device. I think we just hit at the right time with the right product and people piled on. A lot fewer of our backers for the Core came from our proactive marketing than for the Socket; a lot more of them came from word-of-mouth through the Kickstarter. These were things that we had very little influence over or control over other than just making a good product.
We are doing a little bit of everything all at once. Part of turning it into a sustainable business is just figuring out how we sell enough Cores to pay our bills. But also, more broadly than that, we’re trying to figure out how we take a new platform that we are designing for Makers and for hobbyists and turn it into something that is commercially viable for a big company. I think a lot of the big opportunities are not just helping individuals, but helping companies take their existing hardware products and turn them into internet-connected products. A lot of the technology we have built for the Spark Core works for both an individual engineer and for a large industry organization. We’re still learning how we position one product for two markets with a lot of overlap, but also with a lot of differences.
China was an amazing experience. I think anybody who wants to get into hardware start-ups has to go to China, at least for a little while. Look at designing a circuit, for example. We have CAD software to design the circuit boards on our computers and then we have to get them spun up, but we can’t do that ourselves, so we have to go to board houses. In the US, you have a few options. You can go to an American board house, and they can do things relatively quickly—somewhere between two days and two weeks—and they can turn out a board for you and that will cost anywhere between $300 and $1,000, or potentially even more if you want overnight delivery. The other option is to use a service like OSH Park that pulls together a bunch of open source projects or hobbyist projects and sends them out to a board house together. This might cost about $10, but can take three weeks. That really slows down your development cycle if every iteration takes three weeks.
Being in China, we could get a board spun up from a board house that was around the corner. We could get it turned around in two to five days and it would cost somewhere between $15 and $100. $15 if we are willing to wait five days and $100 if we want it the next day. The design of the Core that was on the Kickstarter campaign was the tenth iteration, and we did those ten iterations in a month. If we had done it in the US—at ten iterations a month—it would have cost us tens of thousands of dollars.
“I expect that the time it takes to get the product out the door will come down a lot and the cost will come down a lot.”
There are a lot of things that are hard about building a commercially viable hardware product and there are a lot of companies that are solving different parts of the equation. I think right now it’s still a mess of tools — you find the things that work for you, and you don’t use the things that don’t. I image over the course of the next few years, a lot of this will come together and it will be much easier to transition from prototype to a product that’s mass manufacturable. I expect that the time it takes to get the product out the door will come down a lot and the cost will come down a lot. All of a sudden you’ll see a lot more start-ups building their own hardware.
A lot of them are backers. There are quite a few people that we’re talking to that backed us on Kickstarter and are planning on launching their own products on Kickstarter or Indiegogo. Anyone who comes us and says, “Hey, I’m building a thing using the Spark Core,” we do whatever we can to help them. The Choosatron is the first and there are a few more in the pipeline. You should expect to see a couple more Spark-powered products coming out in the next couple of months. Anyone we can help, we want to help. It’s not just about giving them Spark Cores, it’s about our knowledge of Kickstarter and marketing and China.
“The things that make you move quickly in hardware aren't the same as the things that make you move quickly in software, and we come from software backgrounds.”
I would tell them to make something small: start a small business. One of my greatest learning experiences was before I started Spark. I started with a little e-commerce retailer I made called Hex Goods that was selling designer products, which I built myself. It was easy; it was dropshipping from some designers I found on Etsy. I learned so much about how start-ups work just by doing it. I did it with something small enough that I could do it myself and do it on my own time; it was stressful at times, but it was manageable. When it wasn’t working I shut it down — I shut off the website.
I think the best advice I’ve been given is to move quickly. I’ve gotten it many times from many people in many different contexts. It was a lesson that we learned the hard way from the Spark Socket, because we didn’t move fast enough, and we weren’t first to market. I think that really killed us. With the Spark Core, we got it right. We were first to market with a new technology that’s better than everything else on the market and it worked. I think that it took us a while to understand how to move fast in hardware. It’s hard. The things that make you move quickly in hardware aren’t the same as the things that make you move quickly in software, and we come from software backgrounds.
What’s your current state of mind?
Excited. We're so close to delivery, and I can't wait to get our product out into peoples' hands.
What is your idea of misery?
Helplessness. Being unable to control ones' own destiny.
When and where were you happiest?
In business school I was in a cover band, and we closed out a packed show at The Metro in Chicago, where I crowd surfed off the stage at the end of the set.
What is your greatest extravagance?
I'm a huge coffee snob, and I have way too many coffee gadgets around the house.
Which words or phrases do you most overuse?
Massively. As in, "This is a massively complex project."
Which talent would you most like to have?
I really wish I could draw.
What’s your favourite quote?
"Any sufficiently advanced technology is indistinguishable from magic." - Arthur C. Clarke
Which living person do you most admire?
My parents. I suppose that's a cheesy answer, but they're both professors that have changed the world with their research, and also did a pretty solid job raising me and my sister, and I think that's worthy of admiration.
What is your greatest fear?
That after having a taste of entrepreneurship, I'll have to give up and go back to the corporate life.
What do you consider your greatest achievement?
Learning electrical engineering. I'm entirely self-taught.
In the office we have a board where we keep quotes from our backers. One of them said, “You’re helping to make Minneapolis a better place to live.” That really meant a lot to us because we get a lot of people telling us, “Oh, this product is cool. I can’t wait to use it. I’m really excited about.” And of course that’s wonderful. But also to believe that, beyond just our product, we are doing something good for the community is wonderful.
What question would you pose to our next interviewee?
What's been the most exciting moment in the journey to getting your product to market?