How To Kill A Product Gracefully

Nest caused an uproar by abruptly disabling its smart home hub Revolv. What’s a better way to discontinue products? Designers weigh in.

This week, the home automation company Nest announced that it would disable the Revolv smart home platform, because it no longer fits with the company’s product line. So as of May 15, if you own a Revolv–which lets you control various smart home devices through a central hub–it will simply go to sleep, never to run again. And users are pissed. One customer likened the gadget to a tub of hummus.


But sooner or later, every product dies. So what should the tech industry do differently when these Internet of Things things, released with the flash-in-the-pan fervor of any other technology, reach the end of life? What do we do about connected products when they’re no longer profitable or feasible to sustain?

Talking to IoT design experts, we found ideas that really aren’t so difficult to implement–but the Valley may need government regulation before it learns real customer service.


The Valley Needs To Shift To A 10-Year Mentality

We have an expectation for things in our home–water heaters, air conditioners, KitchenAid mixers–that we don’t have for our smartphones. We expect them to last. And yet Valley companies, with a fail-fast mentality, are building Internet of Things devices without thinking about long-term life spans.


So the first thing IoT companies need is a fundamental perspective shift.

“Obviously, people investing in smoke alarms assume they’ll work for many years–5, 10, 20. These are eons in technology terms,” says Gadi Amit, founder of NewDealDesign. “Will Nest support their smoke alarm, guaranteed, for the next 10 years? I don’t have a clear answer to that, but it’s definitely an answer that someone at Nest should be looking into.”

John Kestner knows this point well. He runs the IoT company Supermechanical. His first product, Twine, was a sort of do-anything environmental sensor you could place in your home to track stats such as heat and moisture. It was an influential product with a loyal cult following, but it wasn’t a massive success. Now, five years past his Kickstarter campaign, he’s paying out server fees that equate to nothing more than lost profits.


“The expense of making it was high. And we weren’t going to continue plowing money into something that wasn’t necessarily developed for the long-term future,” Kestner says. “But we feel a sense of obligation. The majority of [Twines] in active service are doing important things in people’s houses–like monitoring floods. Sure there were tinkerers, and theirs are on the shelf. The ones that are active are doing important things.”

Range Dial

Importantly, Kestner planned for this scenario. Twine’s back end was coded in such a way that costs him a relatively inconsequential sum to maintain, especially as server prices drop over the years. Basically, even though he didn’t have a lot of money, he knew he’d have the money to support Twine for a long future.

“I don’t fault anyone who plans for failure,” he says. “Does that mean you design the thing so it’s easy to open source? Do you make sure a certain amount of money is put into escrow so it can maintain servers?” Truth be told, there are a lot of solutions, if companies think at the 10-year limit, rather than the two.


Build Smart Devices Without The Dumb Internet

Of course, an even simpler approach to smart devices that can stand the test of time is to build them to be self-sufficient, and operate without the Internet. Maybe that sounds like a paradox–the Internet of Things, no Internet required! But what sounds better to you? A smart lightbulb? Or a smart lightbulb that won’t turn on because it requires a firmware update?

Plus, when you remove the Internet from the equation, you remove the long-term maintenance of servers and code. It’s no surprise that, after learning this lesson with Twine, Kestner developed a smart thermometer that only needed an app–not the Internet. And then his next thermometer didn’t even require the app; it has a hard dial, like good old-fashioned kitchen electronics. You can still go on the Internet with his new products, but it’s not mandatory. Naturally, he thinks his early learnings will become an industry-wide trend.

“I think there’s a good chance that most of this [Internet of Things] stuff gets rolled back into making things more like the Clapper,” Kestner says. “Something that is a convenience, and makes life easier, but it’s a completely self-contained service. It doesn’t gratuitously require the Internet.”


Consider The Comcast Model–Don’t Groan!

But technology is improving faster than ever, you say. No one wants to wait 10 years between new Clappers, you say. Can’t we have Valley products that are both bleeding-edge and reliable?

Maybe. As John Edson, president of Lunar, points out, a company has already done this. But a word of warning, you might not like this company very much.

“If you think about Comcast and their home security, or DVRs, if they were to discontinue supporting something, I’m paying less per month, or they’re replacing it with another thing,” Edson says. “I think that’s an interesting model.”


Specifically, that model is something like hardware as service, rather than hardware as product. Rather than paying $300 for a thermostat that might become a brick in two years, you’ve subscribed to a service, with hardware that can be swapped in or out as necessary to provide that service.

“It’s less about, ‘I need another camera’ than ‘I need to see in my driveway at night.’ Then Comcast makes that happen,” Edson says. “That model is much more understandable from a consumer side, and it feels more like I’m getting a smart home service than I bought all this equipment, and one day one item stopped working and I lost that money.”

And If No One Complies, Just Regulate The Industry

Even with service models or better self-regulation, it’s possible that the hardware industry, rewarded for big gambles and breakout products more than anything else, won’t change the way it approaches Internet of Things devices.


And if that’s the case, there’s one last solution. “Frankly, as much as I don’t like a lot of government intervention, and I really believe in free markets, there are aspects of technology and business behaviors that should be regulated by some degree,” Amit says. “In the connected home . . . there should be some legislation on that. And, it’s not unfamiliar to some industries.”

A decent analog might be the telecommunications industry, which ensures licensees provide emergency communications for the public, or better still, the auto industry, for which manufacturers are tasked with countless safety tests and emissions standards. That level of bureaucracy doesn’t always ensure consumer safety or corporate transparency, but it operates as a check-and-balance system (one the tech industry largely evades right now). “Let’s say you’re driving your Toyota, and Toyota decides they don’t want to support that engine anymore,” Amit says. “I think if a car company tried this tomorrow morning, things would be a lot more complicated.”

It’s Really Not Hard; It’s Just Customer Experience

The Internet of Things may be technology’s Wild West, filled with nascent hardware and competitive software. But that’s no reason for companies to treat customers any differently.


In Revolv’s case, millions of people weren’t directly affected, but the hasty announcement from Nest–which Google acquired in 2014–doesn’t exactly endear the company to customers. “People talk. That’s the era we’re in. You don’t buy any product over $50 or maybe less without going online to see what people say about it,” Edson says. “This is going to create some kind of impact on what people say about Nest, surely.”


About the author

Mark Wilson is a senior writer at Fast Company who has written about design, technology, and culture for almost 15 years. His work has appeared at Gizmodo, Kotaku, PopMech, PopSci, Esquire, American Photo and Lucky Peach