Skip to main content

Tuning in: Smart speakers and the resilience of radio

The popularity of smart speakers is soaring off the charts, even though they’re an urban-wealthy trend right now. And video hasn’t killed the radio star. In fact, radio is edging out TV in terms of listening time, according to the predictions of Deloitte’s technology leaders Paul Lee and Duncan Stewart.

“We love our smartphones. We love our computers. We are tied to these devices. When you’ve got your hands two feet inside a turkey, you don’t want to be touching your smartphone screen.”

—Duncan Stewart, director of technology, media, and telecommunications research for Deloitte Canada

Tanya Ott: And you don’t have to. Today, on the Press Room, we’re talking about the rise of a new device that makes it even easier to go hands-free … And details on another 100-year-old technology that’s going stronger than many experts predicted.

I’m Tanya Ott … and I’m in my kitchen. My husband Jason and I spend a lot of time in here. We both really enjoy cooking. And we’ve got a new device that’s making it even easier.

Jason Fulmore: Find a recipe for Thai basil pork.

Tanya Ott: We had a smart speaker, which I loved. I’d listen to my podcasts while doing dishes or clearing through email. But recently we upgraded to a smart display. It’s like a speaker, but with a screen so we can watch cooking videos and ask the digital assistant to read us the ingredients, step by step.

Tanya Ott: We can also ask it to convert tablespoons to ounces and suggest substitutions if we don’t have a certain ingredient on hand. It’s awesome! I’m a convert and I’m not alone. Smart displays are pretty new on the market, but smart speakers have been around a while and they’re predicted to be one of the fastest-growing device categories in 2019. Paul Lee and Duncan Stewart have been tracking the industry. Paul is Deloitte’s global head of research for technology, media and telecommunications; he’s based in London. Duncan is director of technology, media and telecommunications research for Deloitte Canada. This is Paul …

Paul Lee: So, this year, we’re expecting a surge in adoption to about an installed base of 250 million units by the end of the year. And also quite a big growth in revenues to about US$7 billion by the end of the year. But just to put that in context, that US$7 billion compares to probably about half a trillion dollars in sales for smartphones and between 1.4 [billion] or 1.5 billion units sold of smartphones versus 164 million units [of smart speakers] shipped. So, it’s a great device. It’s doing really well among certain demographics, but it’s still an emerging device.

Tanya Ott: It’s still an emerging device, but has the potential for real growth. You talk about doing really well with certain demographics. Who is most likely to be using a smart speaker right now?

Paul Lee: At the moment, like with lots of new technologies, it’s wealthier households who are willing to experiment with new types of technology. And the price of smart speakers, particularly with discounts, are quite affordable, let’s say, for [the] top quintiles. And, so, dropping [as little as] US$30 on a new piece of technology which you may use for one of two applications—I think it’s a good match. That’s where we’re seeing very, very strong adoption. But it’s still [the] relatively earlier, more urban, slightly wealthier category which has been buying these rapaciously.

Duncan Stewart: I want to throw up a big stop sign here because when we talk about demographics, we normally mean things like age and gender and income. We’ve got to remember language—and these are very, very language bound. I just got back from a trip where I was in Belgium. The French speakers in the southern part of Belgium were much more likely to have smart  speakers. According to the Flemish speakers—a language very similar to Dutch—in the north, smart speaker ownership was much lower. Many of the people there said they had them, they didn’t like them that much and they were forced to speak to them in English because, apparently, the Flemish-language version was so poor that that they simply saw much less utility. Globally, Mandarin Chinese works apparently very well and [there is] very high penetration in those markets, but I have a friend who is Korean and he said, for example, that’s not working well.

Tanya Ott: So, there’s a real opportunity there to get Flemish working better, to get Korean working better, to grow the market there …

Paul Lee: Yeah, that’s a very good point. Smart speakers or digital voice assistants—currently the language support is limited. If you look at, say, a couple of years ago, it was predominantly focussed on English. And even a year ago, the Alexa service was available in English, German and Japanese. Over the last year, they’ve added in French and Spanish support, but this is all quite recent. And to Duncan’s point—for some of the smaller languages, there is no consistent support available for them. I think this could mean lots of potential, but it is difficult to make voice recognition work at a degree of accuracy that makes it really useful on an everyday basis.

Tanya Ott: I mentioned earlier that I use smart speakers and smart displays for lots of tasks around the house. But there is also a lot of opportunity outside home personal use. For instance, in restaurants and, particularly, fast food restaurants. Tell us a little bit about what you could envision there.

Paul Lee: One of the applications where we see smart speakers working really well is with something like drive-thrus. In the US, there are 12 billion drive-thru orders placed every single year. For that kind of the interaction—where you’ve got somebody speaking to basically a microphone and relating to a limited set of options of whatever’s on the menu—that’s the kind of application that you could readily move to a smart speaker. It frees up somebody from what would be a repetitive task.

One other application that we, as a firm, have been involved in is to do with an upgrade to the call button in hospital beds. Currently, in the majority of hospitals, when a patient needs help, he or she presses a button. That button may mean anything from, “I just want a blanket or another pillow,” all the way through, “I've fallen and I need help.” What a smart speaker does in that instance is provide a context to what’s needed. So, if it is just, “I need a blanket,” then an orderly can be sent. And if somebody’s fallen, then that can lead to a quick response from multiple people. The vertical centre applications for smart speakers are really fascinating.

Tanya Ott: The use case in medicine is really pretty interesting. It sounds like one of the other bigger blocks of potential users would be those who are visually impaired, because they could do a lot of tasks more easily by voice command rather than going onto a computer or something like that.

Paul Lee: Yeah, that’s definitely the case. I think we often have overlooked how many people are visually impaired [or] that may be visually impaired at a certain time of the day. You know, once you get to a certain age then we get long-sighted. People don’t always want to be carrying around glasses. So, for people probably in the late 40s, late 50s and onwards, the availability of a smart speaker provides access to computing power without having your glasses on, which sounds quite trivial but that’s one of the benefits of, say, larger smartphones screens. It’s being able to do things without having to put your glasses on. And then, as you mentioned, there are people who are visually impaired. And there are also about 700 million people globally who are illiterate and a smart speaker could provide access to computing power for this very large chunk of people.

Tanya Ott: Interesting. I hadn’t thought about that one. I imagine there are probably listeners who are, at this point in the conversation, saying to themselves—oh great, no one ever has to get off the sofa again. They can do everything they need by yelling at their smart speaker—all the way from getting facts and checking the weather to turning lights on and off. So, you know, they’re just going to live on the sofa. Your response to that?

Paul Lee: Voice can work for certain applications and, in other respects, it can be a lot slower and some of the benefits [can be] quite marginal. For some people, talking to a smart speaker to ask for the weather is something they’d prefer to do. For other people, they’ll just glance at the smartphone or listen to the radio and just have the radio bulletin announce what the weather is going to be and they’re happy with that. What we find quite interesting is at that point in time, when we look at what are the most popular applications of smart speakers, the number one in markets where a range of smart speakers are available is playing music. Which is fascinating because partly what that shows is this is not that disruptive an application; this is continuity. I think this expresses the limitation on the market as of now. I think what you need to have is multiple applications that people are using smart speakers for on a daily basis and we’re not yet at that level.

Duncan Stewart: Just to reinforce Paul’s point, when I was talking with people about their smart speakers, one of the questions I tend to ask is, “Do you have a smart speaker?” The follow-up question is, “Do you love it?” Because we love our smartphones. We love our computers. We are tied to these devices. We love our TVs, in many cases. Many of the people use—the phrase I use for this is “smart clock radios.” If your smart speaker is a device that plays music, news, weather and tells you the time, well, we’ve had that since the 1970s. And the people who use their smart speakers like that [tend to] like the device, they use the device, but they’re not in love with the device. But those who do more, those who are using it for the connected home activities or the cooking example—when you’ve got your hands two feet inside a turkey you don’t want to be touching your smartphone screen—with those applications, people shifted from, “I like my device” to “I love my device.” I think that’s going to be a tipping point to watch going forward.

Paul Lee: And it’s also different phases of life. One thing that we’ve heard from young parents is that if they’re holding a baby with two arms, then the great thing about a smart speaker is it can play music and turn off the TV or turn down the lights without them having to get up. And it’s that convenience which they find makes them love the smart speaker at a point in time. Now, when their children are older, they may feel we don’t need this, but certainly, at that point, a smart speaker is a godsend.

Tanya Ott: Taking it to the other end of the age spectrum, I have a friend whose father is in his mid 90s and he has a smart speaker and he wakes up every morning and asks the smart speaker to play Perry Como. And when the smart speaker plays Perry Como—and it’s Perry Como every day—when the smart speaker plays it, he says “thank you.” And he has this sort of relationship that his son looks at and goes, “Well that’s kind of odd, you realise this isn’t a real person, right?” But there’s the potential that the speakers could get pretty smart in the future and would act sort of like a surrogate friend for someone at the other end of the age spectrum.

Paul Lee: It’s a very good point. One of the roles of technology is to deal with a modern phenomenon—which is loneliness. Smart speakers are one way of dealing with it. Video calls are another and even things like instant messaging.

Tanya Ott: We’ve mentioned radio. We’ve mentioned playing music and things like that, so it’s not hard to make the pivot from smart speakers—which is one of the main things that you examined in your predictions for 2019—to radio, which is another thing you examine in your predictions for 2019. Duncan, I think you’ve made the joke “Video killed the radio star,” but, yeah, not really. What do the stats say on the radio front for 2019?

Duncan Stewart: Well, there’s a couple of ways of looking at it. One of them is the radio absolutely has not gone away. It’s a £30 billion industry. It’s actually growing globally. It is something that has extraordinarily high reach. The numbers vary a bit by country and by language and stuff like that, but [for example], over 90 per cent of Americans listen to radio in a given week and that's true even of young people. Perhaps the most shocking thing that came out of my research on radio this year was not that radio was not dead—I don’t think people worried that much about that—it was the resilience of radio, not just by reach, not by the number of people listening to it, but by the number of minutes.

Now, I need to do a bit of a deep dive on this one. When we measure radio or TV, there’s two things: How many people listen to it and how long do they listen to it on average per day. An interesting thing is going on—not just in the United States, but in other markets—which is that the number of minutes per day of traditional TV that young people are watching is falling and falling at around 10, 15, 20 per cent per year depending on the market. At the same time, radio listening may be declining, but it’s not declining very much. It’s a much lower rate. If you just keep extending the lines, you see that within a few years, in the United States, young Americans will be listening to more minutes of radio per day than they will be watching traditional TV per day. That crossover has already occurred in Sweden and Finland and it’s about to occur this year in Denmark. So, we have multiple points around the world suggesting that if you want to reach young people, radio, over time, may in fact become more effective than television.

Tanya Ott: I want to delve into that a little bit more deeply. When you’re talking about television declining, are you talking about traditional television—like, over-the-air television and maybe recorded- and viewed-within-seven-days sort of television? You’re not talking about all-screen video use?

Duncan Stewart: I’m talking about traditional TV, which I think everybody knows what that means. It’s the stuff out there that’s got ads on it. Right. And that is dropping about 10 per cent per year or more for that younger demographic. Radio is not dropping by the same amount.

Tanya Ott: Interesting. I was at a journalism history conference recently when a young undergraduate student presented a paper about FDR’s fireside chats. And she’s a young person who’s really interested in radio and she made the offhand comment that “of course radio’s not that popular these days because everybody’s listening to podcasts.” And I had to really resist the urge to correct her in the middle of her speech, because the number of people listening to podcasts and consuming audio in that way is still significantly smaller than over-the-air radio. It sounds to me that that’s what you’re seeing as well.

Duncan Stewart: To tie back to the smart speakers or even radio, frequently, when people say, “I listen to a podcast,” they’re actually talking about a show that appeared on the radio, was harvested and turned into a podcast, and they now listen to it that way.

At many levels, that is still radio and, of course, people are listening to podcasts on their smart speakers. But when you size that market, as I said, the weekly reach for radio in the United States is in the 90-ish per cent area. Podcasts are much lower at 20 to 30 per cent. When we look at the global market for radio, it’s about a £30-billion-a-year industry. Podcasts are about a £450-million-a-year industry. Are podcasts growing? Absolutely! They’ve got about a 30 per cent per year compound annual growth rate. So, they’re certainly much better than radio at one per cent per year compound annual growth rate, but you do have to reset. I think you’re running into the same problem I do, which is that when you go to a room of technology early-adopters who don’t have a lot of time, they do listen to podcasts at rates probably 10 times the average and probably listen to less radio as a result. But when you look at the average American, a whole lot of people are still listening to the radio.

Tanya Ott: And a whole lot of them are listening to the radio while they’re driving, so that leads me to another question: When we have widespread use of semi-autonomous or autonomous vehicles and you don’t necessarily have to be restricted to just consuming audio or looking at the scenery around you when you’re in your car, what prediction do you have on that? Because I’ve talked with some industry professionals who are kind of freaked out about it.

Duncan Stewart: Correct me if I’m wrong, but you’re actually trying to score a bonus—a free prediction off of me here—not just on radio but on self-driving cars (laughter). But that’s ok. Yeah. The data that we have shows that of the people who do listen to radio, over 90 per cent still listen in their car. So, it's a big driver, pun intended. Almost certainly, at one point, we will see the radio numbers be affected by the growth in the autonomous vehicles, but it is a long way off.

Paul Lee: Just one comment on that... if you go back to the 1990s, that was when there was a surge in the number of cars with CD players. They had CD players with cassettes that you put in the boot. So, you load up six CDs and that was meant to be the big clash for radios. People could choose their own music. And what they tended to find was that for people who had that kind of setup, the only times they changed their CDs that they had in the back of their boot was when they were selling the car. Basically, radio completely wiped out the need for that because a morning show or an afternoon show when you’re driving is a great way to accompany your drive. It’s a blend of music, conversation, weather, news. It’s a great way to start the morning and I think, by comparison, just a straight play of tracks that you’ve chosen can be a little bit repetitive. Can be a little bit bland.

Tanya Ott: Any other major findings from your research on radio that you wanted to share?

Duncan Stewart: The biggest surprise on radio for me was the demographics. Not the young people—although that was a surprise—but the bigger one was that I assumed radio is free and there are a lot of people driving around who maybe haven’t got as much money and so they’re listening to the radio because it’s free and you just turn it. You don’t need a technological device or a data subscription in order to stream music to your vehicle. We found the exact opposite. In our survey in the United States, people who listened to the radio were more likely to have a job by a lot. They were more likely to have higher education and, most strikingly, they were much more likely to be higher income. So, if I’m an advertiser—and let’s be honest, at the end of the day radio does depend on advertising—for a lot of radio around the world, that ability to reach the employed, higher income, higher educated [audience] is a compelling story and one that I think radio, by and large, hasn’t done a great job of spreading perhaps as widely as it might [have].

Tanya Ott: I’m just wondering what do you think is behind that? I’ve spent my career in public radio so I know my audience on public radio tends to be higher educated and wealthier. But that is kind of a surprising result.

Duncan Stewart: It’s actually pretty simple if you combine it. You’re much more likely to listen to the radio if you have a job and can afford a car, right? That’s the tipping point. And, by the way, that’s not just a North America finding. We saw exactly the same thing in data I did in the Nordics and in Europe. So, you know, have a job, have a car and you listened to the radio. If you exclude those who don’t have jobs and aren’t driving to work, you change the demographics a lot.

One of the final things I guess I might mention is where people listen to the radio. We mention the car, but the other thing is—and I don’t know if this backs into anybody’s experience here but it does mine—when I was 20 or 21, I put myself through school one summer by working in the back room of a pizza restaurant and there were eight or nine of us making pizzas for eight hours at a time. And there was a radio playing for my entire shift and nobody ever changed the station and nobody ever skipped a commercial. Instead, we were the ultimate captive audience. We were young. We were buyers. Our preferences had not yet been set. And from an advertiser perspective, the fact that 30 per cent of 18- to 24-year-olds in the United States listen to radio while they’re in the workplace, whether it’s a pizza restaurant or a construction site or in a factory or warehouse—that's an enormous captive audience that radio is capable of reaching, perhaps as well or better than most other media.

Tanya Ott: I’ve got to say, as a long-time radio producer, I am really happy to hear that! Paul Lee and Duncan Stewart’s reports on smart speakers and radio are available for download at They’re part of a series of predictions for 2019 including things like quantum computing, 5G connectivity and China’s plans in the semiconductor industry.

Subscribe to the podcast—it’s free and you can have these conversations delivered directly to your device. You can also connect with us on Twitter at @deloitteinsight (no S) and I’m on Twitter @tanyaott1. Catch you again in fortnight.

This podcast is provided by Deloitte and is intended to provide general information only. This podcast is not intended to constitute advice or services of any kind. For additional information about Deloitte, go to

Did you find this useful?

Thanks for your feedback

If you would like to help improve further, please complete a 3-minute survey