Feb 22, 2017 ยท 8 minutes

“Hardware is hard. It's called hardware for a reason.”

So said (Pando investor) Marc Andreessen at a PandoMonthly in 2013. He was talking about a mantra that had appeared in Silicon Valley among founders and investors: “Hardware is the new software.” Software, to quote another Andreessenism, was eating the world, thanks to new tools and cloud-based services that were making it cheap to launch a software startup. At the time, some were seeing the same dynamics starting to work in consumer hardware as well.

Five years ago, the New York Times noted a “hardware renaissance in Silicon Valley,” becoming one of the first to use the new mantra and pointing to then promising companies like Nest, Lytro, Pebble and Jawbone. The mantra endured for years, with many others repeating itover and over again. Venture investments in hardware increased. Crowdfunding sites like Kickstarter bootstrapped many ideas for new consumer hardware gadgets like Pebble and Ouya.

This year, it's looking more like Andreessen's cautious assessment of hardware was right. Hardware is fucking hard. Some hardware startups of the past decade shut down operations. Some have effectively done so after being acquired for their employees. Some have pivoted into new lines of business like software, abandoning their hardware products. Some hang on to languishing sales that fail to meet targets every quarter. With a few notable exceptions (more on that later), hardware companies are at best struggling to make a profit and at worst are going out of business.

Here are a few notable examples.

-GoPro, the poster child of the hardware renaissance, is trading a just over a third of its IPO offering price and a tenth of its peak $86 a share, reached in October 2014. This month, after the company warned of weak sales in 2017, several analysts downgraded the stock, with one lowering his target price from $10 a share to $6 a share.

-Pebble said in December it's “no longer able to operate as an independent entity. We have made the tough decision to shut down the company and no longer manufacture Pebble devices.” Fitbit bought its intellectual property and hired some workers.

-Fitbit. Meanwhile, had been seen as a hardware success story, and its consistent profits argue that it still is. But its stock is down 54% from its last earnings report, when it missed revenue estimates and warned of weakness in future quarters. The stock is down 71% from its IPO's $20 offering price.

-Ouya, following lackluster sales of its gaming micro-console, sold to Razer. Or rather, Razer bought its software and employees, leaving the hardware behind.

-Jawbone launched in 2010 with a Bluetooth headset, pivoted to the Jambox wireless speaker, then moved into to fitness trackers with the Up. Earlier this month, reports emerged it may abandon consumer hardware to pivot again into services for health-care clinics.

-Lytro's innovative light-field cameras failed to catch on with consumers, so the company is shifting to technology like 360-degree 3D cameras targeted to film studios.

-Oculus may have a long-term potential in VR headsets, but its near-term progress has been, as Mark Zuckerberg said this month, “delayed, so that was obviously somewhat of a disappointment.” Facebook is closing 200 of the 500 Oculus pop-up demo stations in Best Buy stores.

-Magic Leap, meanwhile, was reportedly scrambling to create a prototype of its augmented-reality glasses.

-Nest's smart thermostat, once seemed as a herald of the Internet of things' consumer market, has along with Dropcam languished inside Alphabet, which folded it into Google's smart-home division last fall.

-Other areas of consumer hardware are also proven to be tough soil to farm. PCs and tablet sales are declining. Most smartphone makers are locked in a cutthroat battle for market share, with Samsung and Xiaomi losing share and even upstarts like Oppo and BBK. 3D printers, drones, the Internet of things - each has so far proven to be faddish markets producing more buzz than profits.

So no, hardware isn't the new software. To paraphrase Andreessen, software may be eating the world, but hardware is slowly eating itself.

This wasn't what proponents expected from the hardware renaissance. One reason for the optimism of the past several years was that economics of hardware business were changing. Batteries and specialized chips were simultaneously growing more powerful and cheaper. Component costs were also falling. Contract manufacturers like Flextronics offered cheap production and design. Individuals and startups could build gadget more quickly than ever. There was reason for VCs to feel like this was a promising new development.

But many of the tougher realities of hardware didn't change. The industry still had higher manufacturing costs than software, which still translated into lower margins. It still had longer development times and upgrade cycles. Hardware markets still faced commoditization which erode markets – in fact, this is a phenomenon that seems to occur at a faster rate with every passing year.

Still, some enterprise markets like servers, storage and semiconductors are finding some success, often by waging price wars to increase market share and eking out profits through economies of scale. Some companies are even enjoying a boom time, such Nvidia's graphics chips used to power machine learning algorithms. Nvidia's stock is up 266% in the past year

Consumer hardware is largely another matter. Startups that have emerged during the past decade, often with the backing of venture investments if not Kickstarter campaigns, created new buzz-worthy products and aspired to forge new markets they could in time dominate.

In many cases, though, the products didn't resonate with a mainstream market. Take the wearables market, which generated great expectations only a couple of years ago. Many smartwatches proved to be fads with limited functions, leaving consumers cold. "Excitement for the category was driven almost entirely by vendors of gear and components and not at all by the market,” Jan Dawson wrote on Tech.pinions. “The jobs to be done by these new wearables were far from clear and the offerings in the market did a poor job convincing anyone they were worth having.”

And when these gadgets did resonate with the mainstream, deeper-pocketed rivals stepped in. Smartwatches and fitness trackers had to deal with Apple Watch. Sonos and Jawbone speakers are competing against Echo. And increasingly, those rivals include low-cost manufacturers in China.

Ironically, the same trends that made contract manufacturers cheap outsourcers also created a culture of copycats that can spot a hot Kickstarter idea and quickly produce a lower-priced knockoff. This is shaping up to be a problem for bigger companies like GoPro. “Shenzhen... makes much cheaper cameras that are just as good and drones that are better,” Edison Research analyst Richard Windsornoted recently. “GoPro really has its work cut out for it and we remain unconvinced that it is going to survive on its own.”

Still, there are moderate successes in hardware, starting with the Apple Watch, which has sold tens of millions of units to date even if it hasn't quite lived up to expectations. Roku remains a popular streaming box, yet it's long-expected IPO has yet to happen. Sonos has built up loyalty among its audiophile customers, although it suffered a round of layoffs last year, lost its CEO this month and pushed into new areas like paid streaming and voice control.

And then there are the rare products that, in early 2017, are reaching or showing the potential to reach the broad scale of consumer demand many dream of: Apple's iPhone, Amazon's Alexa, the Square reader and Snap's Spectacles. What those four have in common is that they aren't just hardware but are devices that are connected to a group of fully developed services that make them compelling over time:

-The iPhone's status as a premium smartphone comes not just from its design but from the robust offering of apps that, taken together, offer an experience that lets Apple alone make profit margins on par with a software giant. To the degree that the Watch is successful, it's because it links into the iPhone's world of apps. Some analysts think Apple will keep building an ecosystem of devices that grow interdependent over time, with each device connected to the services business that is now the fastest growing division of Apple.

-Alexa's voice-powered device would be little more than a gimmick if it wasn't connected to the media and other services Amazon has been aggregating into Prime for years. That's strengthened by the skills developers are adding to Alexa, the way the App Store gave Apple an early lead in smartphones. (Google Home may also succeed as it has a similar array of services.)

- Square isn't a consumer hardware company per se, but talk to small-business owners who use its reader or point-of-sales stands and many of them sound like consumers enthusiastic about the gear. Square has build services for businesses on top of the hardware to strengthen those ties, such as inventory and sales analytics.

As for Spectacles, the jury is still out, and yet word of mouth and demand is high, unlike the early reaction to Google Goggles. That may be because Spectacles is a new camera that plugs into the Snap social experience that is already a part of many people's everyday lives.

The lesson here is that, while it may be easier than ever to build a hardware startup, it's also easier than ever to end up with a fading fad of a gadget. To endure, you need more than hardware. Or even the software that's running it. You need a set if immersive services that can handle everyday tasks or connect you with others. What jargonistas might call an ecosystem, but that only begins to describe it. Hardware that works today is a kind of portal that consumers can step through to get into a place where they enjoy spending time.