5 reasons I’m still not using Google Home’s powerful new automations

For the last week, I’ve been testing out Google Home’s powerful new automation editor and trying to see if it could perform the kind of routines I need in my smart home. I was happy to see a few interesting additions, like adding time delays, but Google is still shooting itself in the foot by artificially restricting so many possible automations and actions.

As a Home Assistant and Google Home dual-platform user, I find the difference between what I can do on both laughable. Of course, I expect to do more in Home Assistant — that’s the whole point of a powerful and open platform like that — but so much of Google’s Home power is tied and limited, for no reason whatsoever.

Building a routine in Google Home has been a pretty barebones, yet somehow confusing process for seven years now, since the feature first launched. It used to be a matter of choosing one sentence and the action it triggered, then Google added options for more trigger types (date and time or device statuses), multiple actions, and even some conditions.

The whole interface was a bit confusing, specifically when adding conditions to a trigger, because you could only do it inside the trigger. Now, the interface is more straightforward. The logic is “when this happens, and if this is true, then do that and that.” I think separating triggers and conditions makes them a lot more obvious for anyone building a new routine. Google also went beyond date and time, so conditions can also be device statuses or linked to your home/away presence.

I can finally write a routine that turns on my air purifier at night when I’m home, or turns off my office desk switch when I’m away, or turns on the fan in the living room when the indoor temperature is too hot, but only if the light is on. Simple things like that used to be impossible (unless you used the Help Me Script YAML editor); now they can be easily done in the basic interface of Google Home. I can even add multiple conditions, thus granularly controlling when each automation runs or not. I love this because the more routines you create, the more you realize that there are many situations where you don’t want them to run.

Actions got overhauled, too. There’s less focus on the Google Assistant capabilities, like adjusting phone settings, playing the news or radio, and getting the weather, and more focus on automations per se. You can still do all of those, but you’ll have to write the command yourself under an Ask Google action. That’s a painful downgrade. What Google has added, though, are two crucial options: notifying house members and adding delays before an action is executed. Those two are essential for me, since I like getting reminders when I leave home: run the robot vacuum, double-check that the alarm is armed, but I don’t want to be notified the moment I leave. A few minutes delay is important since I’m usually busy checking bus times or directions, and I don’t want to be distracted by the vacuum reminder notification or swipe it away by mistake.

All of these are excellent improvements to the routines editor in Google Home, but they’re still not enough to sway me to rely on it for all of my automation needs. Here’s why.

What has always annoyed me about Google Home’s routines is that Google artificially restricts the conditions and actions I can choose for most of my smart devices. This is still the case now, despite all the other improvements. For example, I have an air quality monitor that measures the CO2, particles, and volatile organic compounds in my bedroom, but I can’t use those values to trigger any automation. Turning the light red when the CO2 level is too high to remind me to aerate the room is impossible with the new visual automation interface (unless I mess with the YAML editor).

The same limitation is there with my fans, air purifiers, air conditioner, and so on. I can only turn on/off lights and control their brightness, but I can’t change their colors. I can only turn on/off the fan, but I can’t change its speed or toggle its rotation. I can only power on/off the A/C, but I can’t change the mode or temperature anymore. All of these are available in my device’s control page in Google Home, but they’re not available as actions in the automation builder. Why? If Google knows those actions are there, why not surface them in automations? That baffles me.

Instead, I have to manually write the voice command under an Ask Google action, and hope that I chose the right words to trigger the automation. For fan speeds, this has been an exercise in frustration and futility. One day, “Set the bedroom fan to four” works, the other it doesn’t, and I have to change my command to “Set the bedroom fan to speed four.”

For many devices, automations aren’t even possible yet. They appear greyed out in the automation builder with a “Support coming soon” notice. My Google Nest speakers and hubs, robot vacuums, washing machine, dryer, and several sensors are not available as triggers. I can’t set up an automation to announce that the washer cycle is done, for example. Nor can I use my outdoor humidity sensor to warn me of impending rain.

For executing actions, the new routine builder doesn’t support my vacuums, security cams, washer, or dryer. Once again, I can use Ask Google and write the command, but somehow, the visual automation builder says support is coming soon. That’s dumb.

Worse yet, scenes are nowhere to be found in the Google Home app or automation builder. All of my Philips Hue scenes are somewhere in the ether — Google Assistant understands them if I use the correct voice command, and I can add them to automations under Ask Google, but there’s no way to pick them in the app’s visual interface.

Add to this that the default Home and Away routines are so hamfisted for no reason. They still use the old routine interface, too. Why can I only turn on and off lights when I come home or leave? Why not surface the option to toggle the fans, TV, or A/C? Why not  set the thermostat to Away or Home mode? Instead, I have to create a separate routine and use presence as a trigger or condition to choose those actions.

At this point, I feel silly mentioning the lack of nested conditions and actions. I’d love to be able to set up routines that go like: do this, wait, and if this, then do that. That’d be perfect when I leave home: Turn off the lights, TV, switches, fans, and air purifiers, then wait two minutes, and if the door is not locked or the alarm is not armed, then send me a notification and lock the door. (My alarm only supports showing arming status, not executing actions in Google Home, and I like it like that.) Clearly, I can’t do that in Google Home’s routines. It’s a miracle that we got standalone conditions to begin with.

Look, I love that Google is paying more and more attention to Google Home; there’s even a whole new app redesign on the horizon with Gemini integration. But many of these powerful new features are basic things any smart home user wanted last decade, not in the mid-20s. The fact that I can finally use presence as a trigger or condition is nice, but it’s too little and way too late. And the fact that many of the possible triggers and actions are still artificially hidden and restricted is crazy.

I took my smart home into my own hands earlier this year and adopted Home Assistant. The learning curve has been steeper than steep, and the way Home Assistant handles some aspects is beyond counterintuitive, but once I learned the ropes, I was up and running with more powerful automations than Google Home could ever give me. My office light switches to red when the washer or dryer is done. It turns yellow when the CO2 level is too high. All my devices turn off when I leave home — not just my lights. I can set the air purifier and A/C to the exact mode and speed I need them to be in different situations. And so on. The only routines that remain in Google Home are the ones that require a voice trigger or that interact with my Google smart speakers and hubs, and I plan on migrating those to Home Assistant soon.

Google Home lost me to Home Assistant for almost everything because of its super slow and inconsistent updates. I’ve even built my own, more powerful dashboards that I could never set up in Google’s app. The only things I still use Google for are some random voice commands and accessing my devices from my Google TV or Pixel Watch. For everything else, Home Assistant became my go-to, and the fact that it’s received one monthly update like clockwork since I set it up, with a bunch of excellent new features that show active development, gives me more hope than relying on Google’s whims to keep its hobbyist Google Home project alive.

I’m sorry, Google, but I can’t trust you with my smart home anymore.