Smart home gadgets suck. Consumer IoT has been a massive fad for a couple of years now but good products are few and far between. Most either don’t deliver on their promises or try to solve problems that aren’t even there. So let’s look at the few examples from the likes of iRobot, Sonos, and Amazon that shine through the pile of crappy products in their category.

But first, how do the vast majority fail? In two ways: wrong purpose and wrong mindset. Let’s start with purpose…

Novelty is not value

One of the most revered products among the smart home preachers are the Philips Hue bulbs. Lists of best gadgets for home feature it all the time (CNET, Business Insider, Independent). But when you look at the issues people struggle with, you would not accept it from an off-the-shelf, ‘dumb’ bulb. 

Hue reviews

Users are facing many problems: wrong colors, buggy app, settings changing after power outage, you name it. And while no product is free from bad apples and occasional issues, Hue comes with a ridiculous amount of problems when you consider the standard set is three smart bulbs priced at $180. You could get three decent lamps and still have enough money left to buy standard bulbs for them. And the overall experience wouldn’t suffer because: a) bugs are a foreign concept in lamps; b) how much do you really need a smartphone-operated bulb that can change colors?

Focus on novelty lead the industry to believe that there’s an app for that (or, you know, a whole OS) is an answer to everything. Being connected to an app has become a purpose onto itself. And it makes sense on the surface: after all, we carry the phones all the time, right? Except that at home, we don’t. And even if we did… unlocking my phone with a passcode is already more complicated than hitting the light switch, and I’m not even in the app yet.

Novelty for the sake of novelty is the opposite of value. Take smart door locks. They come with additional conveniences like key-sharing and remote control. But they’re based on Bluetooth, an interface that doesn’t really have a flawless history. And while BT is just right for most uses, I don’t want to pay $200 for a lock that works eight out of 10 times. My analog lock that came with the door, might not support multiple digital keys but it opens every time I need. Classic appliances have another advantage: they don’t need to be connected to Wi-Fi to work. A relief, judging from the fact that iKettle 2.0 has about 10 questions in its rather large FAQ related just to connectivity. And I don’t even want to think what are the less common issues if ‘My iKettle 2.0 keeps beeping randomly?’ made it to the FAQ.

Using and setting up connected objects is a chore, compared to the simplicity of good ol’ analog appliances, be it a bathroom scale or a coffee machine. You pay premium, but what for? The only return is novelty. Not much to compete with established designs that just work.

Move fast, break home

Nest, though more appreciated than Hue, represents another set of problems typical for smart home. Battery-killing software bugs in their thermostats, smoke alarms that could turn off by accident, stopping support for the Revolv hub, Nest has been through all of that. In fact, after being acquired by Google, they saw so much trouble that the founder, Tony Fadell, had to leave. And that’s the poster child for consumer IoT.

The thing is, a lot of smart home products are created by startups or companies trying to act like startups. And startups are often a mix of two things: software background and the move fast, break things mindset. Sadly, neither works well with home appliances.

Physical world is messy space, where your elegant code might not be the best solution. The best solution might not require code at all. Also, users are familiar with how the things are in software. They know Windows or iOS need updates. They’re used to occasional bugs in apps that are fixed with new releases. But that’s not how we think about home appliances. If my bathroom scale is crippled by a fatal error or you tell me I can’t turn on lights because of a firmware update, I’m never spending a dime on your future products.

Reid Hoffman, LinkedIn’s co-founder, has once said that If you are not embarrassed by the first version of your product, you’ve launched too late. It may appeal to software developers with their MVPs and continuous integrations, but if you’re embarrassed by a home appliance you designed, please stay embarrassed without selling it to people. Home is where I live. I want many things to happen here, but disruption isn’t one.

Purpose and mindset done right

Luckily, there are gems proving that smart home can indeed be smart. My favorite is iRobot Roomba that single-handedly created the space of autonomous vacuum cleaners. Roomba never had the only purpose is novelty problem. It was first released in 2002, well before the consumer IoT bonanza started. And even today only the newest, most expensive models of Roomba have a companion app, which is just a handy extension of an already great device.

In 2003, Bloomberg ran a story about bringing Roomba to the market. If you ever decide to make consumer hardware, read it. A lot of what you’ll learn runs counter to what you’d find in a startup playbook. First of all, iRobot was not a bunch of software guys that just happened to stumble upon an idea. It was founded in 1990 by experts in robotics and artificial intelligence. Only after a decade of working military contracts they moved to design their first consumer device. They spent months talking with industry veterans and even hired a vacuum cleaner consultant to learn about the market. Still, their product started taking shape only after talking to focus groups. And even then, it took iRobot another 20 iterations to make first Roomba production-ready.

But hey, you might be saying, we don’t have funds to hire consultants, run focus groups, and go through 20 iterations of our product.

I’m afraid there’s no ‘don’t worry’ coming. Before releasing Roomba, iRobot had a wealth of R&D experience. What they didn’t have was knowledge of the market and consumers: so they spent time acquiring it. Autonomous vacuum cleaner was not a fantasy conjured by someone on the team and legitimized with a Kickstarter campaign. It was exactly what people wanted and needed. iRobot invested enough in research to know that. And they had a team that could deliver it without compromising on quality. Thanks to that, consumers got a vacuum cleaner blowing other models out of the water.

None of this comes cheaply or quickly, but nothing in consumer hardware really does.

Sonos, even though the company is struggling today, set another example for how to create a successful smart home product. In case of Sonos, however, what they built is as important as how they did it. Their flagship product is a high-quality sound system that’s wireless and works over Wi-Fi. But first and foremost, it’s a high quality sound system, period. Sonos Sub subwoofer has a 4.8 rating on Amazon: you’ll be hard pressed to find a higher rated consumer electronics product. ArsTechnica called the Play:5 model the best-sounding wireless speaker system they ever used. Sonos speakers would hold their ground against any sound systems in their price range. If it was a wired system, people would still love it for sound quality. And only upon a foundation this strong, the company added the true selling point: Wi-Fi support.

Don’t mistake it for an afterthought, though. John MacFarlane and his partners founded Sonos in 2002, when iTunes Store wasn’t a thing, much less Spotify, and they were ready to make a bet on wireless music and streaming. The purpose was there from the very beginning: add wireless connectivity to speakers. It makes home audio system simpler to use, not more complex like in case of connected bulbs. But it’s the mindset that allowed Sonos to truly deliver. Tagline for the Play:5 speaker is ultimate listening experience. Because listening to music is defined by music. Obvious? Tell that to the manufacturers of smart door locks, who seem to have forgotten that a lock is defined by locking and unlocking the door, not key-sharing and remote control.

Last example, Amazon Echo, is important because it shows a completely different approach from the previous two. iRobot and Sonos took existing appliances with a purpose to make them much more convenient with technology. And they followed a mindset of never compromising on quality and always having technology serve the customer, not the other way around, to succeed. Amazon took another route: they invented a whole new category of product.

Amazon Echo is a digital assistant locked in a speaker. It’s like someone moved Siri from your iPhone to your sound system. Amazon was one of the few companies that saw through the mobile-first mania and decided an app is not an answer to everything. As mentioned before, we don’t really carry phones around the house. And by the way, is tapping on a screen really the least friction way to check weather or order an Uber? So instead of improving existing products, Amazon’s purpose with Echo was to create a better interface for the same tasks.

This purpose is best defined with Echo’s response time. The threshold was set at 1.5 seconds, even though initially it seemed technically impossible. But only that would ensure a seamless conversation with Alexa, the assistant’s identity. And only a seamless conversation was a big enough improvement over tapping on a phone to make Echo a worthwhile investment. And this is where the mindset kicks in: the speaker/assistant combo spent four years in development to achieve the desired technical sophistication. There was no MVP, beta release, or an explorer program. Sure, Amazon keeps expanding Alexa’s skillset with updates, but they knew they had to ship something that feels like a complete product from day one. If you’re designing for home, you need to delight users from the beginning. No one is going to wait for you to update and bugfix a home appliance, because there are so many others that just work.

Takeaways

Every smart device needs a purpose. And no, making it connected is not a purpose, it’s a slogan. A dumb one at that. Roomba isn’t connected and in smart home it plays in a league of its own. The only way to define the purpose is to figure out what would make everyday objects more useful. Is there a way to improve vacuum cleaners? Make them autonomous. Is there a way to make sound systems more convenient? Connect them via Wi-Fi. Is there a way to access information and complete simple tasks faster? Bring a digital assistant to user’s houses.

But it’s not enough to set a lofty goal to improve something. Electric home appliances have been around for a century. Things like door locks or kettles go way back to the antiquity. We had a lot of time to figure them out and in most cases, we did. They are already smart. They last long and work as advertised 99% of the time. Plus, people already know how to use them. Think of this as your entry level. If you can’t jump over that bar, if you can’t deliver core functionality on par with what’s already there, don’t even think about slapping Bluetooth on it. If you start selling a coffee machine with a companion app that’s buggy or has confusing UI, you’re only making it dumber.

Many people think you can enter home appliance market just on novelty of apps and connectivity. But at home, you’re competing with established designs that have proven to work time and again. To succeed, you’ll need a lot of technical know-how and a deep understanding of how and why people use a given item. Be aware, it might take you years to get there. Because fuck it, ship it is not the mindset that scores any points here.

About Wojciech Borowicz

Wojciech is a regular contributor to the Mokriya blog. He reads more than he writes, but in both cases it’s mostly about how technology and interfaces are changing: us and themselves.