Smart Speaker Ownership Hits 19% Globally, 35% In U.S.

Ownership of home digital assistants has been growing dramatically over the last year but it looks like there will be some holdouts in adopting the smart speakers.

Digital assistants like Amazon Echo, Google Home and Apple HomePod are owned by 19% of people, according to a new global study.

However, a year from now, that number is projected to be 57% globally and 65% in the U.S., based on plans to purchase. More than a third (38%) of consumers plan to purchase a smart speaker within the next one to two years and of those, 21% plan to purchase one within the next six months.

The study comprised a survey of 5,000 consumers, 500 in France, Germany, India, Italy, Japan, Malaysia, Singapore, South Korea, the U.K. and the U.S. conducted for Limelight Networks by Cint, a global research firm. Respondents were adults who had downloaded software or streamed online video or music during the last month.

Despite the large adoption of smart home speakers, 43% of consumers globally and 35% in the U.S. have no plans to purchase one.

Ownership of Internet-connected speakers is highest in the U.S. and lowest in Japan. Here’s the current ownership level by country:

  • 35% -- U.S.
  • 31% -- India
  • 23% -- U.K.
  • 22% - -Malaysia
  • 16% -- Germany
  • 16% -- South Korea
  • 15% -- France
  • 14% -- Singapore
  • 12% -- Italy
  • 7% - -Japan

By age group, the highest ownership is by those 26 to 35 (25%), followed by those 36-45 (23%), 18 to 25 (18%), 46 to 60 (16%) and those over 60 (12%).

Voice control at home is here to stay.

19 comments about "Smart Speaker Ownership Hits 19% Globally, 35% In U.S.".
Check to receive email when comments are posted.
  1. M Cohen from marshall cohen associates, June 27, 2018 at 9:59 a.m.

    Chuck,  It would be really helpful to know: how many interviews were conducted per country (10 x n=500?) who actually did this research?  (you say it was "for" Limelight Networks but not who did it), how was this research actually done? online? how were the potential respondents "invited" to participate? were they invited "online?"

    So, does the sample represent people who are online, not people in general? 

    Without any of the necessary information about the methodology employed here, no insights can be gained by this research.  

    I really wish that publications like MediaPost, that aim to serve the media industry would consider asking about these important methodological details, before you simply write from the press release, and if they are not supplied, refusing to publish.

  2. Ed Papazian from Media Dynamics Inc, June 27, 2018 at 10:47 a.m.

    In line with Marshall's comments, it's really a stretch to accept the statement that with 19% of the world now owning a "smart speaker"--assuming you believe this figure----that by next year this will rise to 57%---based on "plans" to purchase. Nothing, in my experience, has shot up so fast in such a short time---not even smartphones. Why doesn't MP question such claims while reporting on them---or ask someone with knowledge of the subject --not me----to offer a comment to give some balance to the article?

  3. Chuck Martin from Chuck Martin replied, June 27, 2018 at 10:49 a.m.

    There were 500 in each country, M Cohen.

  4. Chuck Martin from Chuck Martin replied, June 27, 2018 at 10:50 a.m.

    The U.S. numbers are consistent with other studies, Ed.

  5. Deborah hohler from Limelight Networks, June 27, 2018 at 10:56 a.m.

    M Cohen, 

    This survey was fielded online by a third-party company with access to consumer panels in France, Germany, India, Italy, Japan, Malaysia, Singapore, South Korea, the United Kingdom, and the United States. 500 responses were collected from each country for a total of 5,000.  Respondents were 18 and older who had downloaded software or streamed online video or music during the last month. You can view the full report here

  6. M Cohen from marshall cohen associates, June 27, 2018 at 11:01 a.m.

    Being consistent with other studies is, unfortunately irrelevant. All researchers want to know is the details of how the research was conducted. That’s why research companies join and support CASRO, founded in 1975, (now called who have guidelines for all publication of research findings, including all the details on study methodology -- like was it a "random" sample or a "convenience" sample, etc.

    What are the the people who served this up to you, or what are you afraid of, in asking them for details on who did this study? Your readers deserve that.

  7. Laura Cowan from N/A, June 27, 2018 at 4:36 p.m.

    I'd love to learn everything in the comment above, and also ask whether they factored multiple digital assistants per HH (for example, I've seen data that people have an Echo in the kitchen, then several Echo Dots in other rooms) - so is it number of units/total US HH or factored for multiple units in one HH?

  8. Chuck Martin from Chuck Martin replied, June 27, 2018 at 4:52 p.m.

    Good point, Laura, as in owned per household...other studies also have used "owned or used". I will ask.

  9. Fred Jacobs from Jacobs Media, June 27, 2018 at 5:02 p.m.

    On-board with Marshall's comments here.  This sample is a subset of the population - people who downloaded software (huh?), streamed video or streamed music (and nothing else, like spoken word, podcasts?)in the past month.  This is a big number, and while you've provided some information about the sample, it's not enough to reach firm, reliable conclusions.   The Amazon Echo is only available in English, German, and Japanese, making it more challenging for those in most of these other countries to purchase one. 

  10. Michael Milligan from Limelight Networks replied, June 27, 2018 at 5:05 p.m.


       I helped field the survey for Limelight, so I can answer this. 

       The question that was asked was "Do you have or are you planning to purchase an internet-connected digital assistant such as Amazon Echo, Google Home Assistant, Apple HomePod)?".  Respondents had the option of choosing, "I own one today", "I plan to purchase one in the next 6 months", "I plan to purchase one in the next 1 or 2 years", or "I have no plans to purchase this item".  The data for this question is in figures 30 and 31 of the report which is available at

      So, the responses reflect households, not the number of devices sold.  As you mentioned, some people have multiple devices.  I have two Echo Dots and one Google Home Mini in my house.  :-)

       I hope that information was helpful.  Thanks.


  11. Chuck Martin from Chuck Martin replied, June 27, 2018 at 5:43 p.m.

    Good point about languages, Fred. While Amazon Echo is available in more than 85 countries, including all the countries in the study but one, and Google Assistant speaks English, German, Japanese, Italian and Spanish, the globalization of the devices has been a gradual rollout.

  12. John Grono from GAP Research replied, June 27, 2018 at 10:32 p.m.

    Ed, I plan to be a millionaire by the end of this year.   Add 1 to the Millionaires Club..

  13. John Grono from GAP Research replied, June 27, 2018 at 10:42 p.m.

    Thanks Michael   The survey question asked seems fair, but many other questions remain unanswered.

    So could you please elucidate further on Page 44 - Methodology.   It states "...fielded by a third-party company with access to consumer panels...".   Were these online-onlypanels as Marshall (correctly) asked?   If so were the results downweighted at least to factor in internet penetration in each country?   Was the 'reward' to the respondent considered (i.e. were they 'professional' survey monkeys)?   How did the third-party recruit and maintain their panel.   Was any consideration given to include propensity weighting in the results?

    It's a pity you didn't include Australia as I have access to lots of data on device penetration.

  14. Dominic Sunnebo from mr, June 29, 2018 at 10:32 a.m.

    Unless the study was conducted amongst Google and Amazon employees in India, I can tell you without conducting any research that Smart Speaker penetration in India is not 31%. 

  15. Michael Milligan from Limelight Networks, June 29, 2018 at 11:06 a.m.

    John - Our target audience for the survey was people who use online digital content.  The survey results were collected online.  The results weren't weighted to try to reflect overall internet usage in each country. We just focused on people who use online digital content.  I don't have specific details about how the third-party recruits or maintains their panels.  

    BTW - I love Aussies.  We didn't have a large enough budget to include Australia in the survey, but we will certainly consider it in the future!

  16. M Cohen from marshall cohen associates, June 29, 2018 at 5:19 p.m.

    One more little question. And then I am done with this.
    Why are you so reluctant to reveal the name of the research supplier you used?

  17. Chuck Martin from Chuck Martin replied, July 1, 2018 at 12:24 p.m.

    The name of the company is in the story, M Cohen.

  18. David Mercer from Strategy Analytics, July 5, 2018 at 11:33 a.m.

    These smart speaker ownership levels are certainly high according to our estimates (based on supply-side research as well as user surveys), most likely because of the content-user focus of this study:
    Strategy Analytics mid-18 estimates (% of total households)
    US - 23% 
    UK - 15%
    Germany - 6%
    South Korea - 6%
    France - 4%
    Italy - <1%
    Worldwide, if it really means anything, penetration is currently around 2% of households. 

  19. Chuck Martin from Chuck Martin replied, July 5, 2018 at 11:43 a.m.

    Thanks, David, good data.

Next story loading loading..