From a spate of wearables startups to Google's recent $3 billion acquisition of thermostat and smoke alarm maker Nest, Silicon Valley has rediscovered the virtues of actually making stuff.
Don't Miss: This is How to Find a Nintendo Switch in Stock
But it’s still not easy, least of all for consumer products. Fitbit, the pioneering maker of health and fitness tracking gear, today provided some insight into just how hard hardware is for a startup. Fitbit cofounder and CEO James Park talked about the company’s difficult experience in an interview with one of its investors, Jeff Clavier of SoftTech VC, at the Startup Grind entrepreneur conference at the Computer History Museum in Mountain View.
To hear Park tell it, the challenges of making stuff almost put the company out of business–seven times. Park recalled that he started the company in April 2007 after realizing the potential of sensors such as accelerometers paired with smaller and smaller devices. The company raised $400,000 from Park and cofounder and CTO Eric Friedman ourselves and investors in their last company. “We had a couple slides at the time,” he said. “We actually thought we could get to market on $400,000. A year later, we weren’t even close.”
So they had to go into fundraising mode. SoftTech and True Ventures stepped up. “We had a prototype up and running, and a very rudimentary website,” Park recalled. “Eric said I don’t think more than five people will be stupid enough to give us their credit card. I said you’re being too pessimistic, there will be 50 people.”
In September 2008, they got a great reception from a presentation at the TechCrunch 50 conference, getting about 2,000 pre-orders by the end of the day. ”We did announce onstage we were going to ship at Christmas, but we didn’t say which year,” Park noted.
But it turned out to take another year to get to market, as the company encountered what Clavier called the Manufacturing Valley of Death. “Onstage, we literally had a circuit board and a balsa box,” Park recalled. “We had no experience in manufacturing. We probably spent about three months in Asia looking at suppliers, bringing up production lines. Several times, we were pretty close to being dead. Seven times we were close to death.”
No. 2 was a problem with the device’s antenna. “In my hotel room, I was thinking, this is it, we’re done. We literally took a piece of foam and put it on the circuit board to fix an antenna problem.”
So Fitbit finally launched around Christmas 2009, by which time the company had more than 25,000 orders. Why the relatively huge interest in a product that didn’t yet exist? “It was the right product at the right time at the right price point,” Park said. What’s more, the 5,000 units they could ship were all sold direct, so “the margin was pretty darn good.”
Even so, it was still hard to sell investors on a bigger round of funding. “That round, we talked to about 30 investors.” Finally, Clavier persuaded VC Brad Feld to take another look and he invested.
The early focus paid off. Park said Fitbit is the No. 1 app for fitness in Apple's App Store, which he said is amazing because you need a $100 device to use it. NPD now says Fitbit has 77% of the market for full-body activity trackers despite potent competition from Jawbone, Nike’s Fuelband, and others, he said, as well as 50% market share for digital fitness devices such as those made by Garmin and others. Park noted that Nike spent more at the South By Southwest conference than Fitbit’s entire two previous years of marketing spend.
But with Samsung, Apple, and others eyeing the market, how will Fitbit compete? Park said they’re betting that a single form factor such as a watch won’t take over. That’s why Fitbit has multiple models.
Despite the obvious hardware angle, Park said the key remains good software to make it work well. “Two-thirds of our engineers are in software, not hardware,” he said. “The lack of focus on hardware was what allowed us to take over the market very quickly. Garmin and others that were already doing hardware gave software short shrift.”
Park also said there are different philosophies to getting distribution for hardware products that can work. Fitbit decided on a national-account strategy to control pricing better–in contrast, for instance, to GoPro, which started with mom-and-pop stores. Best Buy reached out to Fitbit not long after its launch, and the startup first did a four-store pilot, then expanded to 40 and eventually 650. Fitbit is now in 30,000 stores overall in 28 countries. “Most companies, it’s going to take many, many years to build up distribution, especially for a new category,” Park said.
Park said it’s easier today to do a hardware startup, for several reasons. “There’s a lot more resources and experts to help you out,” he said. “When we started, there was no public information on how to make hardware. The key to success in hardware is access to information–about suppliers, manufacturers, techniques. The thing we spent the most time on was gaining access to that knowledge.”
Second, it’s easier to get market validation. “Kickstarter makes it much easier to raise funds,” he said.
Still, there’s little room for error. “You have to be involved in all the details because one mistake in hardware and you’re done,” he said. “One mistake can set you back months.”
Indeed, it seems that most entrepreneurs still aren’t willing to go the hardware route. Clavier asked the audience of several hundred how many are building a hardware startup. I saw only one hand. To which Park said: “You’re crazy.”
Clavier’s parting advice: “Manufacturing is a bitch. Remember that.”