Life in a Google Home and what it says about the smart home

In mid-December, my family went cold turkey on Alexa in both our kitchen and living room. Alexa still reigns in the bedrooms, but we removed the Echo Show to try out life with a Google Home, as we had been growing increasingly frustrated with Alexa’s inability to answer our dinnertime questions. The results of the Google Home experiment, however, have been even more frustrating.

Google is awesome at answering weird questions. And it does a better job than the Echo Show of “hearing” us when we make requests from across the room. But it has become the bane of my existence because it is glitchy as all get-out. For example, when I say, “Hey Google, turn on television,” sometimes it can turn on the TV. And sometimes it can’t.

When it can’t, it reports back with, “Sorry something went wrong. When you’re ready, give it another try.” If instead, I say, “Hey Google, ask Harmony to turn on the television,” it will often work. But why will one command work only part of the time? I have no idea. This seemingly random ability to work also extends to other connected devices in my home.

When I decided to go all in on the Google Home in our main living area, I unlinked all of my linked accounts and relinked them to give the system a fresh start. Yet, just a week into the experiment, Google stopped being able to reach my Lutron lights. One day they worked, and the next day Google was telling me that it had trouble reaching the “Lutron Caseta RA2 system.”

After hoping this would fix itself (yes, I know, but that’s what normal people would do), I went back into the app and unlinked the account and relinked it again. Now it works. But for how long? And why did it stop in the first place? Ugh.

So integrations need work. Or maybe they just need regular attention. In the meantime, there are two other areas where Google Home has let me down. One is around voice recognition. Google wants users to train the Assistant on the use of their voice. So when I ask for my calendar, the Home shows mine; but when my husband asks, it shows his. This is a really cool feature. If I suggest that Google call Mom and it recognizes my voice, it calls my mom. If my husband does the same thing, it uses his contacts to call his mom.

That’s magical — when it works. But it doesn’t always work. Apparently, Google doesn’t like my voice, because roughly 30% of the time it won’t recognize me. If I take a particular tone with Google I can be recognized more often, but sometimes that doesn’t work, either. And that means that I don’t have permission to do certain things around my home, such as turn on the television. Google lets you train the Assistant as many times as you want, and I have done so countless times, but it still presents a hurdle.

Yes, I could turn the voice permission feature off, but getting my calendar when I ask for it sure is handy, so I keep trying. And that leads me to the second problem. We have a 12-year-old child. Which means that with voice permissions turned on, she’s also incapable of doing a variety of things around the home.

The catch is she doesn’t have her own smartphone, which she needs to train her voice and control the Google Home. To get around that, we used a Gmail address I created for her and let her sign in using an app on a spare iPhone. Another way to do it would be to download the Google Family Link app and create a Google account and permitting for specific devices. But if I went this route, only certain permissions and apps would be available to her. Additionally, at the age of 13, when the law in the U.S. deems her no longer a child (in 8 months), the permissions I set would go away.

Given her age and my trust in her judgment, the best way to go was to set her up as a third adult member of the household. Which means that — just like my husband — she got an invite to train her voice and control the Google Home. As one of my Twitter followers pointed out, those permissions are also a pain to set up.

There are a few other niggling issues that frustrate me with the Google Assistant set-up, such as the inability to put one device into multiple rooms and the fact that when I put several devices in a room and then ask Google to turn on Kitchen, everything in the kitchen turns on when I really just wanted to the overhead lights. Ironically, naming each device for the room it’s in works well for the Amazon Alexa ecosystem but is frustrating for the Google Home.

But I promised to tell you what Google’s challenges here say about the smart home. Anyone who gets deep into this space begins to recognize that smart homes are still stupid, requiring the user to program the devices or command them via voice. The smart home, despite five years or more of effort, is still not intuitive. We are starting to see what we need to make it intuitive though.

To make a home intuitive, digital assistants need to understand who people are in the home because those people have different needs and requests. They may also have different rights, such as a parent’s right to turn a thermostat up while a kid might be told to “put on a sweater.”  An assistant also needs to understand where people are in the home. So far, some devices have presence sensors, and facial recognition to offer this, but digital assistants aren’t yet taking advantage of it.

For example, a few days ago on a whim, I asked the Google Home if it had seen Anna on the Nest Hello doorbell this morning. I didn’t think it would work, and it didn’t, but it’s an example of information that requires both the context of where a person is and who they are. The Nest Doorbell can actually show me this in the app, but it can’t yet offer the information verbally through Google Home.

Finally, these devices have to be much easier to program. They should do a better job of helping people understand how to name a device and why they should place it in a room. If you look at the two leading companies in this space, Google, and Amazon, it’s clear the things Google has decided to tackle are very different than the problems Amazon has tried to tackle.

Amazon is working hard to make it easy to set up a smart home using your voice. It makes it easy to control devices and bring them onto your network and even allows you to build really complex routines if you know what you want and you’re to invest time in programming them. It’s also trying out Hunches, which will offer a suggestion if you make a request that is normally paired with another request or action that Amazon deems obvious. For example, if you tell Alexa Goodnight and it runs the command but also sees another light on, it might suggest you turn that other light off.

Hunches are still rolling out, so we’ll have to see if they are a gateway to the intuitive home, or merely an annoyance.

Other than a pretty easy-to-bypass security code, Amazon doesn’t provide much in the way of barriers to controlling the device from a permissions and identity perspective. You can train Alexa to recognize your voice, but it only uses that recognition for message playback and calling. On business accounts, you can restrict calendar access by voice, but on home accounts, my husband can ask to see Stacey’s calendar and get it.

Google cares more about the identity of people in the home and is clearly envisioning far more complicated interactions. Unfortunately, as it tries to showcase those use cases — such as the ability to call the right mom depending on who is asking — it has let some important usability elements slide. As such, it’s really hard to set up a Google Home device and quickly experience any of its benefits. The barrier to entry is higher. You do get access to some cool features once you’ve breached that barrier, but you’re also limited in scope by fewer apps and glitches that make commanding a smart home a challenge.

I suspect that for many people the Google Home Hub, the JBL Google display, and the Lenovo Google Smart Displays may be fancy picture frames that let people ask questions, play music, and see shared Google albums. Meanwhile, Amazon is letting Alexa run away with the smart home and a digital assistant that provides a lot of assistance when it comes to taking basic tasks and making them hands-free.

It’s a shame because I see where Google is trying to turn its devices into assistants that can do a lot more than pre-heat your oven or turn off your upstairs lights. It wants them to notify you if your commute will be longer or help you spend less time on your phone by scheduling meetings and reminders with your voice. It wants to make an intuitive home. But while Amazon is trying to address the difficulty of setting up and programming these devices, Google is tackling user identity and rights. Both may get us to the intuitive smart home, but Amazon’s approach definitely feels more user-friendly.