Nest’s economics don’t work

Why IoT devices should be sold as services.

I’ve been thinking about Nest’s business.

They sell a $249 thermostat. Assuming typical margin structure, when Nest sells one at Home Depot, they probably makes about $66*.

That sounds pretty decent. It’s enough to cover some retail returns, a bit of customer support, and still turn a profit if you sell enough of them.

Except for one thing: that thermostat is going to sit on the wall for 10 years. Maybe 20.

Nest has to spend money to keep it working the whole time. These thermostats connect to cloud servers, which must be kept running. iOS and Android apps need regular updates for security and compliance with ever-changing phone SDKs — if Nest stopped updating those apps, they’d stop working. (Even software systems, somehow, fall under the law of entropy.)

Even worse for Nest, we’ve all come to expect that apps get better over time. We want new integrations with other devices, we want updated design and skinner typefaces. We want swipes in the UI. That takes work, too.

Finally, over the next 10 years, you’re probably going to upgrade your Wi-Fi router at least once, creating a non-zero chance that something will go wrong and you’ll need to call Nest support. That’s also expensive for Nest.

The trouble is: All of Nest’s revenue is up-front, but their costs continue for years.

So instead of thinking of it as $66 in gross margin, we might start to think of it as $6.60 per year. Or 55¢ a month.

Nest better hope people don’t keep those thermostats for 20 or 30 years.

Is there an alternative? What if Nest charged $10/month instead of $249 up-front? Margins look tighter in the first year, but after 10 years revenue would total $1200. That’s a lot more money to support services and upgrades.

Consumers might object, of course. But they might not. Few people pay for their cell phones up front, preferring to finance them over time, often at terrible interest rates. What if the $10/month were part of your utility bill? Since Nest is saving you $140/year, you might not even notice.

We also see plenty of evidence that consumers are getting comfortable with subscription fees. Nest’s camera offers a subscription with additional features, and something like 40% of buyers pay it. (Now that is a great business.)

It’s also worth mentioning that it’s harder to finance a business like this. Unless, of course, your parent company has $83B of cash in the bank.

The point is, consumers are probably going to have to learn to pay for on-going services for connected devices. The economics just don’t work out any other way.


(One exception is worth mentioning: devices that make money over time. Phones are used to buy apps and rent movies, so Apple and Google have an incentive to keep your devices up-to-date (other phone manufacturers do not have the same incentives). Amazon’s Alexa probably won’t ever have a service fee, because you’ll use it to buy more things on Amazon. In commercial IoT, where I work, businesses know that they need ongoing support, and they’re typically happy to pay for it.)

*Assumes 40% margin, or $100 to Home Depot, $76 in COGS including $63 in parts and 20% for assembly, yield loss, delivery, etc… $249–100–76=$66.


Daniel is co-founder of Beep Networks, an enabler of easy-to-deploy long-range IoT networks and systems.