Smart speakers with voice assistants have accomplished one key goal. At a low cost, consumers were able to bring some smarts into their homes on the cheap. And that has led to a surge in the sales of other connected devices to actually control by voice. Think how many more smart outlets, light bulbs, switches, and cameras were bought because they work with smart speakers.
But aside from that, have any companies offered truly compelling reasons to chat more with our in-home digital assistants? So far no, except when it suits them. Hey Alexa, notify me when that changes.
This situation is manifesting itself in low engagement with Alexa, according to a recent Bloomberg report. According to what it says is internal Amazon data, Bloomberg tells us that Amazon’s effort to flood the market with low-cost smart speakers isn’t equating to revenue stream growth based on our fewer and fewer conversations with Alexa. Amazon may be surprised by this, but I’m not.
Goodbye smart speaker sales growth
It’s not a stretch to say that the growth of the smart speaker hardware market has basically ground to a halt. We experienced a few years of sales growth well into the 30 and 40 percent range. Amazon now forecasts a meager 1.2 percent annual sales increase in smart speakers for the next few years, according to the internal documents cited by Bloomberg.
I expected this: Low-cost speakers and deals that include a free speaker with some connected device purchase made it easy to saturate the market. In many homes, the number of these speakers have multiplied like Tribbles in Star Trek. (Or “like rabbits” if you’re not a Trekkie). Most of us simply have all of the connected speakers we need, if not more than that.
Amazon doesn’t mind if we don’t buy more Alexa devices though. Here’s why, again from the Bloomberg article:
In 2018, Amazon projected it would lose $5 per device in 2021 and said it hoped to improve that to a $2-per-unit profit in 2028. The company says its goal is to make money when people use Alexa to access other Amazon services.
So the company built up its Alexa footprint at a loss in order to gain future revenue on services. That’s not an uncommon strategy, particularly for Amazon. But there’s a problem: That pesky data showing that engagement with Alexa is very low.
The reported internal Amazon data representing the past few years notes that 15 to 25 percent of new Echo device owners stopped speaking to Alexa in just the second week of device ownership. That’s no way to treat a new houseguest.
People mainly use Alexa for three things
More alarming is the data on reported use-cases. I bet you can guess them: Mainly to voice-control connected lights in the home, set timers, and play music. There’s little to no additional revenue in such services. But wait, you say: What about the massive number of Alexa Skills available that could generate income? Back in 2019, there were a reported 80,000 such skills, and this year that number is estimated to be more than 100,000.
Like I said in 2019: There are a few reasons that this growth in skills hasn’t equated to similar engagement we’ve witnessed in the mobile app explosion. The reasons range from limitations of voice-only apps to all of us collectively trying to figure out what we need smart speakers to actually do for us; a question similar to “what do we want our smart homes to do?”
Discovering features on a smart speaker is hard, with little benefit
Discoverability of skills is a major sticking point as well. How easy is it to find and install an Alexa Skill that does what you want? (Hint: It’s not easy, or at least not as easy as it is for finding good mobile apps for your phone.) You might disagree with me on that, but Amazon does. Why else would it be proactively surfacing more and more capabilities on its smart speakers? Stacey has been getting annoyed with those “By the way, did you know I can…” nags from Alexa, and Google Home devices, since February. It’s clearly an approach to boost engagement.
This strategy isn’t a compelling one though. When chatting about this with Stacey, she said none of these “by the way” chats have shown a potential benefit to her. Instead, it’s Amazon that gains value because the suggested services only benefit it.
After buying something on Amazon, for example, her Echo has prompted her to review the product. That’s great for Amazon, but what does she get out of it other than another voice in her home doling out tasks for her to complete?
Smart displays have a better interface for many
Due to the limitations of a smart speaker, there’s no interface to “see”. Yes, using natural language, you can ask the digital assistant whatever you want. But it gets back to discovery again: How do you know if the speaker is capable of handling your voice request?
For this reason, I stopped accumulating smart speakers a few years ago. Instead, I turned to smart displays, which complement the existing voice interaction with a visual interface. And kudos to Amazon and Google: Both have improved that interface over the past year or two with cleaner menus, leveraged existing screens in the home such as FireTVs, and more relevant “at a glance” data.
By the way, did you realize I didn’t mention Apple?
That’s because it’s not likely losing revenue on a per-device basis since the Apple HomePod mini costs $99. And you’re not getting one free unless you spend hundreds, if not thousands of dollars on some other Apple products. Plus it has no smart display: Something I think it will rectify in 2022 with what I’m calling the “Apple HomePad”. You don’t need much more hardware capabilities for such a smart display than you can already find in a $329 iPad.
When Alexa says “By the way…” I shout “Alexa, STOP!” There is nothing more annoying. I want to throw it through the window. Also, for more than a year multi-room music has not worked properly or dependably. I’ve spent a lot of time with tech support without a solution. Most often, the device I spoke to doesn’t play but the other two do. Or only one does. Or they all do for awhile, and then one stops. All on the same network – I check!
Why would Apple need to make the “Homepad”? Siri already exist on the iPad. You could just set an iPad up in a room to actively listen for the catch phrase and you’re done. It’s also an active home hub. So unless the iOS is modified there is zero chance such a device would exist.
Just using an iPad is fine if you’re going to be within about 8 inches of the device when you’re speaking to it. but if you want voice recognition that works from across the room, while the television is on, and while you are placed at any angle from the device, you have to go to the kind of microphone array that you will find in echo and other Smart speakers designed for this purpose.
I’m quadriparetic, so I’m heavily reliant on voice technology, and I have tried most of everything that’s out there for home use. The echo was a huge innovation because its voice recognition worked so well under normal conditions at a distance. Most systems which were based on a tablet or laptop prior to the echo required that the user wear a headset microphone in order to get acceptable voice recognition at a distance.
In my opinion the HomePod is much more responsive than any echo I’ve owned. I have two paired in stereo with my Apple TV 4K and even during a loud show the HomePods pick up very quickly.
Side note, I used to work for Amazon Lab126 as an engineer. I have an early Echo (pre production gen 1) that my mother now uses. Apparently the device is starting to pipe in adds / services to hear to “improve the amazon experience”. I find that to be disturbing as well as extremely annoying when you just want to do something quick. I’ve stripped out all the Madame A stuff outa my own home but may need to do the same with my mother. I have much more trust in Apple vs Google or Amazon due to their business practices.
“back in my day…” we didn’t have voice control. Using voice to control things was is the holy grail of smart home. I love looking for new ways to talk to Alexa. But It is true, its mostly home control, timers, and lists. I can barely get Dominos to deliver using their own app, I wouldn’t imagine using a skill.
I do have a couple smart displays, but if I use them for anything other than a picture frame with google and alexa lurking inside, I feel like I am in a 1985 episode of Star Trek The Next Gen. I don’t want to get touchy with my smarthome, if I did I would use a switch.
What I have learned from following some of the Alexa users’ groups on FaceBook is that not everyone who buys Alexa products is interested in home automation. The discussions on many of these groups suggest that many buyers simply buy the speakers for the novelty of having a device with which to hold simple back-and-forth conversations, play music and to play simple-minded practical jokes on friends and family. That seems to provide entertainment for a year or two or less, and when the novelty wears off, many of their posts seem to veer into little more than complaints about petty annoyances. Within this group, and at this juncture, there is little or no interest in learning and employing the ins and outs of the Alexa app. But considering the fact that many buyers of smartphones and other home appliances also vastly underuse the capabilities of those products, I doubt that this is viewed as a discouraging development by Amazon, and it is highly likely that it had already been factored into their long-range plans for Alexa. My hunch is that Alexa development is on course and that it will continue to be a major player in home automation as the latter becomes more commonplace and increasingly intuitive.
Unpopular opinion: Microsoft had the correct concept with Cortana. They weren’t trying to get you to buy aything then. They knew it was just an assistant; to get you to open apps or turn on devices. But by the time they realized they needed stand alone devices, Amazon and Google had already cornered that market. So they moved it to the board room where it lived happily ever after.
What does the data say about voice use in the car? That stat had been trending up earlier
I went into more detail in a comment on the podcast, but as I look at the Bloomberg article, it appears to be looking at interaction “per device,“ but I suspect that in many homes where people really like Alexa they end up with lots of echo devices that don’t get used very often as individual devices.
For example, we have one in the laundry room which mostly serves as a clock, but if someone happens to be in the laundry room when a visitor comes to the door, they can interact with the visitor through the ring/echo combination. We really like this even though we might only use it once or twice a month.
Our household uses Alexa every day, many times a day, but that doesn’t mean we’re using every echo device in our house every day.
In fact we have one which we got free with the purchase of something else which is put in the basement to act as part of the security features. It acts as a glassbreak acoustic sensor and it allows us to check the area visually if we want. It’s never been used after it was first set up, and I hope it never has to be, but that doesn’t mean we’re unhappy with it.
So I’m not sure the data they reviewed leads to the conclusion that they drew for the article. Not using any one particular device doesn’t mean you’re not using the service as a whole.
Indeed. These days I find myself in casual conversation “oh you don’t have alexa? Here take this one.” one less old echo sitting in a drawer. Better than the $5 trade in.