According to voicebot.ai research, Siri dominates the Voice Assistant market in terms of actual usage to the tune of 45.6% marketshare compared to Google Assistant on 28.7%, Alexa on 13.2%, Samsung's Bixby on 6.2% and Microsoft's Cortana on 4.9%.
Methinks Microsoft may be inflating Cortana’s figures to look good.
Interesting. This would indeed suggest that despite all the whingeing about Siri being less capable than the other digital assistants, Siri is much more actively used by owners of Apple products than are the digital assistants on other brands.
I use Siri 6-8 times a day, setting timers, asking weather info, manipulating lights. More complex than that is futile, as it will fail a lot of basic requests in my country. The same is more or less true for Google Assistant though, although that one will usually do a lot better. The report simply measured frequency, not quality.
Interesting. This would indeed suggest that despite all the whingeing about Siri being less capable than the other digital assistants, Siri is much more actively used by owners of Apple products than are the digital assistants on other brands.
For Siri and Google Assistant to be tied for usage, a lot of people are talking to Siri on their Apple devices, and a lot of people with Android devices aren’t talking to Google. That pretty much has to mean that, despite the claims that Google Assistant is technically superior or more capable, a lot more of the people who have access to it don’t find it useful enough to bother. So why would that be? Is Google’s AI not actually that great in this context? Do people have security concerns and disable it?
You've drawn a conclusion that can't be supported by the information in the article. The survey says both Siri and GA have a 36% usage rate. If both are at 36%, aren't more people using GA than Siri? 36% of Apple device users = a number people. If there are more Android users to start with, then 36% of their users = a larger number people. Right?
Also, you can't equate usage with the capability of the assistant. There's no corollary that ties the two together. That's akin to saying Windows is better than MacOS because more people use it.
Basically all you can discern from the survey is digital assistants are in use and the each have strengths in different areas.
Guess again. You've misread the information. The published report says, "Which digital assistants are people using?" and answers it with: Cortana, 19%; Siri, 36%; GA, 36%; Alexa, 25%; other, 1%. This means that, of all the respondents to their survey, 36% are using Siri, and 36% are using GA.
This does not mean 36% of Apple device users are using Siri. It means 36% of all people they surveyed are using Siri. (Note that their percentages add up to 117, which means some of their respondents are using more than one brand of digital assistant. This would be expected, as some iPhone and Android phone owners probably have Amazon speakers in the house, etc.)
So assuming their survey's respondents resemble something like the general population, because we know that Android has a much larger share of the market than Apple with regard to devices owned, this means that there is a larger pool of the survey's respondents who own Android devices, yet only the same number of respondents use GA as use Siri. That means that a much lower percentage of Android owners use GA as compared to Apple owners using Siri.
That then is what brought me to the questions about why a larger percentage of Apple device owners use Siri, even though all the online chatter likes to say that GA is a 'more capable' digital assistant.
To quote from your own post: Guess again. You've misread the information.
One of the two surveys was US only, where iOS is the predominant mobile OS. The other used 5 countries, three of which (Australia, US, Canada) Android is behind iOS in user-share, one where iOS and Android are tied (UK) and one (India) where the most of the languages needed for the populace were not yet available for Google Assistant (and Siri?) but Cortana was AFAICT. There's got to be a reason why the specific countries were chosen for inclusion.
So you missed a few details and the surveys really don't say what you think they do. I'm not certain some commenters actually read the entire story, you included.
I'm more than a bit surprised that the author said in his lead: A Microsoft study suggests that Apple's Siri is on par with Google Assistant for worldwide usage, despite the latter being available on a wider range of devices" since it was NOT a worldwide study to begin with. The 5 countries included in the results seem to have been cherry-picked to cast Cortana in a good light.
Interesting. This would indeed suggest that despite all the whingeing about Siri being less capable than the other digital assistants, Siri is much more actively used by owners of Apple products than are the digital assistants on other brands.
For Siri and Google Assistant to be tied for usage, a lot of people are talking to Siri on their Apple devices, and a lot of people with Android devices aren’t talking to Google. That pretty much has to mean that, despite the claims that Google Assistant is technically superior or more capable, a lot more of the people who have access to it don’t find it useful enough to bother. So why would that be? Is Google’s AI not actually that great in this context? Do people have security concerns and disable it?
You've drawn a conclusion that can't be supported by the information in the article. The survey says both Siri and GA have a 36% usage rate. If both are at 36%, aren't more people using GA than Siri? 36% of Apple device users = a number people. If there are more Android users to start with, then 36% of their users = a larger number people. Right?
Also, you can't equate usage with the capability of the assistant. There's no corollary that ties the two together. That's akin to saying Windows is better than MacOS because more people use it.
Basically all you can discern from the survey is digital assistants are in use and the each have strengths in different areas.
Guess again. You've misread the information. The published report says, "Which digital assistants are people using?" and answers it with: Cortana, 19%; Siri, 36%; GA, 36%; Alexa, 25%; other, 1%. This means that, of all the respondents to their survey, 36% are using Siri, and 36% are using GA.
This does not mean 36% of Apple device users are using Siri. It means 36% of all people they surveyed are using Siri. (Note that their percentages add up to 117, which means some of their respondents are using more than one brand of digital assistant. This would be expected, as some iPhone and Android phone owners probably have Amazon speakers in the house, etc.)
So assuming their survey's respondents resemble something like the general population, because we know that Android has a much larger share of the market than Apple with regard to devices owned, this means that there is a larger pool of the survey's respondents who own Android devices, yet only the same number of respondents use GA as use Siri. That means that a much lower percentage of Android owners use GA as compared to Apple owners using Siri.
That then is what brought me to the questions about why a larger percentage of Apple device owners use Siri, even though all the online chatter likes to say that GA is a 'more capable' digital assistant.
...
One of the two surveys was US only, where iOS is the predominant mobile OS. ...
Really? I don't think so. Apple may currently have a plurality lead in US device sales, that doesn't appear to add up to being the majority or predominant mobile OS in the United States.
Interesting. This would indeed suggest that despite all the whingeing about Siri being less capable than the other digital assistants, Siri is much more actively used by owners of Apple products than are the digital assistants on other brands.
For Siri and Google Assistant to be tied for usage, a lot of people are talking to Siri on their Apple devices, and a lot of people with Android devices aren’t talking to Google. That pretty much has to mean that, despite the claims that Google Assistant is technically superior or more capable, a lot more of the people who have access to it don’t find it useful enough to bother. So why would that be? Is Google’s AI not actually that great in this context? Do people have security concerns and disable it?
You've drawn a conclusion that can't be supported by the information in the article. The survey says both Siri and GA have a 36% usage rate. If both are at 36%, aren't more people using GA than Siri? 36% of Apple device users = a number people. If there are more Android users to start with, then 36% of their users = a larger number people. Right?
Also, you can't equate usage with the capability of the assistant. There's no corollary that ties the two together. That's akin to saying Windows is better than MacOS because more people use it.
Basically all you can discern from the survey is digital assistants are in use and the each have strengths in different areas.
Guess again. You've misread the information. The published report says, "Which digital assistants are people using?" and answers it with: Cortana, 19%; Siri, 36%; GA, 36%; Alexa, 25%; other, 1%. This means that, of all the respondents to their survey, 36% are using Siri, and 36% are using GA.
This does not mean 36% of Apple device users are using Siri. It means 36% of all people they surveyed are using Siri. (Note that their percentages add up to 117, which means some of their respondents are using more than one brand of digital assistant. This would be expected, as some iPhone and Android phone owners probably have Amazon speakers in the house, etc.)
So assuming their survey's respondents resemble something like the general population, because we know that Android has a much larger share of the market than Apple with regard to devices owned, this means that there is a larger pool of the survey's respondents who own Android devices, yet only the same number of respondents use GA as use Siri. That means that a much lower percentage of Android owners use GA as compared to Apple owners using Siri.
That then is what brought me to the questions about why a larger percentage of Apple device owners use Siri, even though all the online chatter likes to say that GA is a 'more capable' digital assistant.
...
One of the two surveys was US only, where iOS is the predominant mobile OS. ...
Really? I don't think so. Apple may currently have a plurality lead in US device sales, that doesn't appear to add up to being the majority or predominant mobile OS in the United States.
Mobile Operating System Market Share in United States Of America - March 2019
iOS
54.74%
Android
44.89%
Windows
0.13%
Samsung
0.07%
BlackBerry OS
0.06%
Unknown
0.04%
Another fun fact: In general (Australia and Sweden being notable exceptions) iOS market share v. Google Android has been rising ever so slightly over the past three years. Android in China is an entirely separate market, using forks of the Google developed Android, but for argument purposes may or may not be included in articles touting one OS over another depending on how it benefits the storyline.
A bonus fun fact: Both iOS and Android have been losing market share in India. For what operating system you ask? KaiOS, based largely on the old and now abandoned FirefoxOS. Google is one of their investors.
Interesting. This would indeed suggest that despite all the whingeing about Siri being less capable than the other digital assistants, Siri is much more actively used by owners of Apple products than are the digital assistants on other brands.
For Siri and Google Assistant to be tied for usage, a lot of people are talking to Siri on their Apple devices, and a lot of people with Android devices aren’t talking to Google. That pretty much has to mean that, despite the claims that Google Assistant is technically superior or more capable, a lot more of the people who have access to it don’t find it useful enough to bother. So why would that be? Is Google’s AI not actually that great in this context? Do people have security concerns and disable it?
You've drawn a conclusion that can't be supported by the information in the article. The survey says both Siri and GA have a 36% usage rate. If both are at 36%, aren't more people using GA than Siri? 36% of Apple device users = a number people. If there are more Android users to start with, then 36% of their users = a larger number people. Right?
Also, you can't equate usage with the capability of the assistant. There's no corollary that ties the two together. That's akin to saying Windows is better than MacOS because more people use it.
Basically all you can discern from the survey is digital assistants are in use and the each have strengths in different areas.
Guess again. You've misread the information. The published report says, "Which digital assistants are people using?" and answers it with: Cortana, 19%; Siri, 36%; GA, 36%; Alexa, 25%; other, 1%. This means that, of all the respondents to their survey, 36% are using Siri, and 36% are using GA.
This does not mean 36% of Apple device users are using Siri. It means 36% of all people they surveyed are using Siri. (Note that their percentages add up to 117, which means some of their respondents are using more than one brand of digital assistant. This would be expected, as some iPhone and Android phone owners probably have Amazon speakers in the house, etc.)
So assuming their survey's respondents resemble something like the general population, because we know that Android has a much larger share of the market than Apple with regard to devices owned, this means that there is a larger pool of the survey's respondents who own Android devices, yet only the same number of respondents use GA as use Siri. That means that a much lower percentage of Android owners use GA as compared to Apple owners using Siri.
That then is what brought me to the questions about why a larger percentage of Apple device owners use Siri, even though all the online chatter likes to say that GA is a 'more capable' digital assistant.
...
One of the two surveys was US only, where iOS is the predominant mobile OS. ...
Really? I don't think so. Apple may currently have a plurality lead in US device sales, that doesn't appear to add up to being the majority or predominant mobile OS in the United States.
Mobile Operating System Market Share in United States Of America - March 2019
iOS
54.74%
Android
44.89%
Windows
0.13%
Samsung
0.07%
BlackBerry OS
0.06%
Unknown
0.04%
Another fun fact: In general (Australia and Sweden being notable exceptions) iOS market share v. Google Android has been rising ever so slightly over the past three years. Android in China is an entirely separate market, using forks of the Google developed Android, but for argument purposes may or may not be included in articles touting one OS over another depending on how it benefits the storyline.
A bonus fun fact: Both iOS and Android have been losing market share in India. For what operating system you ask? KaiOS, based largely on the old and now abandoned FirefoxOS. Google is one of their investors.
Statcounter tracks page views on partner websites to determine what browsers and, ostensibly, devices are visiting those sites. That could be an indicator of OS market share, or it could be seriously skewed if certain browsers tend to be used more for surfing the web. Consumer Intelligence Research Partners (CIRP), the source of the data used in the AI article I linked, creates their data by directly surveying consumers. Without being a CIRP subscriber, I can't get more details of their methodology (perhaps an AI moderator could shed light on that). Given that the discussion here is regarding the idea that some significant percentage of AndroidOS device owners do not use Google Assistant, it may also be the case that some significant percentage of AndroidOS device users also don't surf the web much. If that's true, it would significantly skew data derived from counting browsers used for page views on the web.
I would offer a hypothesis that since Apple does not pursue the "low end" of the phone market, but many Android device makers do pursue that market, it's likely that iPhones will tend to be tied to phone plans with higher data allowances, while many more AndroidOS devices will tend to be connected via phone plans with much more limited data allowances. If that's the case, then the average iPhone owner probably clicks on a lot more web pages than the average Android phone owner. It's also reasonable to think that for every pricey iPhone sold, several cheap or free Android phones (attached to those less expensive, limited data plans) land in a consumer's pocket.
So what would that mean? That would mean that getting at accurate market share data is highly dependent on methodology. If CIRP does its due diligence in sample selection for its direct consumer surveys (again, maybe an AI writer can get more details about that), then directly asking respondents "what device are you using," and "what OS is it running" will get a more direct answer about market share. Statcounter, on the other hand, specifically says "We provide independent, unbiased stats on internet usage trends. We do not collate our stats with any other information sources. No artificial weightings are used." That sounds great, but what it means is that they are only measuring page hits, and if they aren't "collating their stats," then they really can't claim to be measuring device market share. Page views are not a direct proxy for device market share. Given that serious deficiency in Statcounter's methodology, I'm more inclined to accept the CIRP data that says Apple has a minority share in the US mobile device market.
Interesting. This would indeed suggest that despite all the whingeing about Siri being less capable than the other digital assistants, Siri is much more actively used by owners of Apple products than are the digital assistants on other brands.
For Siri and Google Assistant to be tied for usage, a lot of people are talking to Siri on their Apple devices, and a lot of people with Android devices aren’t talking to Google. That pretty much has to mean that, despite the claims that Google Assistant is technically superior or more capable, a lot more of the people who have access to it don’t find it useful enough to bother. So why would that be? Is Google’s AI not actually that great in this context? Do people have security concerns and disable it?
You've drawn a conclusion that can't be supported by the information in the article. The survey says both Siri and GA have a 36% usage rate. If both are at 36%, aren't more people using GA than Siri? 36% of Apple device users = a number people. If there are more Android users to start with, then 36% of their users = a larger number people. Right?
Also, you can't equate usage with the capability of the assistant. There's no corollary that ties the two together. That's akin to saying Windows is better than MacOS because more people use it.
Basically all you can discern from the survey is digital assistants are in use and the each have strengths in different areas.
Guess again. You've misread the information. The published report says, "Which digital assistants are people using?" and answers it with: Cortana, 19%; Siri, 36%; GA, 36%; Alexa, 25%; other, 1%. This means that, of all the respondents to their survey, 36% are using Siri, and 36% are using GA.
This does not mean 36% of Apple device users are using Siri. It means 36% of all people they surveyed are using Siri. (Note that their percentages add up to 117, which means some of their respondents are using more than one brand of digital assistant. This would be expected, as some iPhone and Android phone owners probably have Amazon speakers in the house, etc.)
So assuming their survey's respondents resemble something like the general population, because we know that Android has a much larger share of the market than Apple with regard to devices owned, this means that there is a larger pool of the survey's respondents who own Android devices, yet only the same number of respondents use GA as use Siri. That means that a much lower percentage of Android owners use GA as compared to Apple owners using Siri.
That then is what brought me to the questions about why a larger percentage of Apple device owners use Siri, even though all the online chatter likes to say that GA is a 'more capable' digital assistant.
...
One of the two surveys was US only, where iOS is the predominant mobile OS. ...
Really? I don't think so. Apple may currently have a plurality lead in US device sales, that doesn't appear to add up to being the majority or predominant mobile OS in the United States.
Mobile Operating System Market Share in United States Of America - March 2019
iOS
54.74%
Android
44.89%
Windows
0.13%
Samsung
0.07%
BlackBerry OS
0.06%
Unknown
0.04%
Another fun fact: In general (Australia and Sweden being notable exceptions) iOS market share v. Google Android has been rising ever so slightly over the past three years. Android in China is an entirely separate market, using forks of the Google developed Android, but for argument purposes may or may not be included in articles touting one OS over another depending on how it benefits the storyline.
A bonus fun fact: Both iOS and Android have been losing market share in India. For what operating system you ask? KaiOS, based largely on the old and now abandoned FirefoxOS. Google is one of their investors.
Statcounter tracks page views on partner websites to determine what browsers and, ostensibly, devices are visiting those sites. That could be an indicator of OS market share, or it could be seriously skewed if certain browsers tend to be used more for surfing the web. Consumer Intelligence Research Partners (CIRP), the source of the data used in the AI article I linked, creates their data by directly surveying consumers.
So what would that mean? That would mean that getting at accurate market share data is highly dependent on methodology. If CIRP does its due diligence in sample selection for its direct consumer surveys (again, maybe an AI writer can get more details about that), then directly asking respondents "what device are you using," and "what OS is it running" will get a more direct answer about market share.
Personally I'm always suspect of results based on asking someone what they use, or what they plan to do, or pretty much anything where the honesty and understanding of a question of it posed to a respondent is necessary. Even the wording can give different results. Look no further than "Hillary is going to win in a landslide" according to consumer surveys. The coronation was indefinitely delayed.
I've had too many people refer to some Android phone as an iPhone (aren't all smartphones an iPhone to some people? Kinda like Googling and Kleenex) to trust the accuracy of a consumer response. I would tend to trust Statcounter more, but that's me.
Interesting. This would indeed suggest that despite all the whingeing about Siri being less capable than the other digital assistants, Siri is much more actively used by owners of Apple products than are the digital assistants on other brands.
For Siri and Google Assistant to be tied for usage, a lot of people are talking to Siri on their Apple devices, and a lot of people with Android devices aren’t talking to Google. That pretty much has to mean that, despite the claims that Google Assistant is technically superior or more capable, a lot more of the people who have access to it don’t find it useful enough to bother. So why would that be? Is Google’s AI not actually that great in this context? Do people have security concerns and disable it?
You've drawn a conclusion that can't be supported by the information in the article. The survey says both Siri and GA have a 36% usage rate. If both are at 36%, aren't more people using GA than Siri? 36% of Apple device users = a number people. If there are more Android users to start with, then 36% of their users = a larger number people. Right?
Also, you can't equate usage with the capability of the assistant. There's no corollary that ties the two together. That's akin to saying Windows is better than MacOS because more people use it.
Basically all you can discern from the survey is digital assistants are in use and the each have strengths in different areas.
Guess again. You've misread the information. The published report says, "Which digital assistants are people using?" and answers it with: Cortana, 19%; Siri, 36%; GA, 36%; Alexa, 25%; other, 1%. This means that, of all the respondents to their survey, 36% are using Siri, and 36% are using GA.
This does not mean 36% of Apple device users are using Siri. It means 36% of all people they surveyed are using Siri. (Note that their percentages add up to 117, which means some of their respondents are using more than one brand of digital assistant. This would be expected, as some iPhone and Android phone owners probably have Amazon speakers in the house, etc.)
So assuming their survey's respondents resemble something like the general population, because we know that Android has a much larger share of the market than Apple with regard to devices owned, this means that there is a larger pool of the survey's respondents who own Android devices, yet only the same number of respondents use GA as use Siri. That means that a much lower percentage of Android owners use GA as compared to Apple owners using Siri.
That then is what brought me to the questions about why a larger percentage of Apple device owners use Siri, even though all the online chatter likes to say that GA is a 'more capable' digital assistant.
...
One of the two surveys was US only, where iOS is the predominant mobile OS. ...
Really? I don't think so. Apple may currently have a plurality lead in US device sales, that doesn't appear to add up to being the majority or predominant mobile OS in the United States.
Mobile Operating System Market Share in United States Of America - March 2019
iOS
54.74%
Android
44.89%
Windows
0.13%
Samsung
0.07%
BlackBerry OS
0.06%
Unknown
0.04%
Another fun fact: In general (Australia and Sweden being notable exceptions) iOS market share v. Google Android has been rising ever so slightly over the past three years. Android in China is an entirely separate market, using forks of the Google developed Android, but for argument purposes may or may not be included in articles touting one OS over another depending on how it benefits the storyline.
A bonus fun fact: Both iOS and Android have been losing market share in India. For what operating system you ask? KaiOS, based largely on the old and now abandoned FirefoxOS. Google is one of their investors.
Statcounter tracks page views on partner websites to determine what browsers and, ostensibly, devices are visiting those sites. That could be an indicator of OS market share, or it could be seriously skewed if certain browsers tend to be used more for surfing the web. Consumer Intelligence Research Partners (CIRP), the source of the data used in the AI article I linked, creates their data by directly surveying consumers.
So what would that mean? That would mean that getting at accurate market share data is highly dependent on methodology. If CIRP does its due diligence in sample selection for its direct consumer surveys (again, maybe an AI writer can get more details about that), then directly asking respondents "what device are you using," and "what OS is it running" will get a more direct answer about market share.
Personally I'm always suspect of results based on asking someone what they use, or what they plan to do, or pretty much anything where the honesty and understanding of a question of it posed to a respondent is necessary. Even the wording can give different results. Look no further than "Hillary is going to win in a landslide" according to consumer surveys. The coronation was indefinitely delayed.
I've had too many people refer to some Android phone as an iPhone (aren't all smartphones an iPhone to some people? Kinda like Googling and Kleenex) to trust the accuracy of a consumer response. I would tend to trust Statcounter more, but that's me.
Every methodology has its weaknesses, but I think Statcounter's jump from 'device page views' to 'device market share' is an unsubstantiated leap that ignores potential biases that would be baked in. As for survey respondents thinking all smartphones are iPhones (like all tissues are Kleenex), that would tend to cause the survey to overstate iPhone market share, and the survey already says that iPhones hold a minority of the market. So correcting for that sort of error would only lead to the conclusion that Apple has an even smaller market share. So then you can go back to the original thing I was commenting on, which was: if Android has the bulk of device marketshare, but similar numbers of people are using Siri as are using Google Assistant, why is that? Maybe GA isn't so much more capable after all, or maybe all those Android users with limited data plans don't want to spend their data resources sending voice queries to Google. I don't know.
Interesting. This would indeed suggest that despite all the whingeing about Siri being less capable than the other digital assistants, Siri is much more actively used by owners of Apple products than are the digital assistants on other brands.
For Siri and Google Assistant to be tied for usage, a lot of people are talking to Siri on their Apple devices, and a lot of people with Android devices aren’t talking to Google. That pretty much has to mean that, despite the claims that Google Assistant is technically superior or more capable, a lot more of the people who have access to it don’t find it useful enough to bother. So why would that be? Is Google’s AI not actually that great in this context? Do people have security concerns and disable it?
You've drawn a conclusion that can't be supported by the information in the article. The survey says both Siri and GA have a 36% usage rate. If both are at 36%, aren't more people using GA than Siri? 36% of Apple device users = a number people. If there are more Android users to start with, then 36% of their users = a larger number people. Right?
Also, you can't equate usage with the capability of the assistant. There's no corollary that ties the two together. That's akin to saying Windows is better than MacOS because more people use it.
Basically all you can discern from the survey is digital assistants are in use and the each have strengths in different areas.
Guess again. You've misread the information. The published report says, "Which digital assistants are people using?" and answers it with: Cortana, 19%; Siri, 36%; GA, 36%; Alexa, 25%; other, 1%. This means that, of all the respondents to their survey, 36% are using Siri, and 36% are using GA.
This does not mean 36% of Apple device users are using Siri. It means 36% of all people they surveyed are using Siri. (Note that their percentages add up to 117, which means some of their respondents are using more than one brand of digital assistant. This would be expected, as some iPhone and Android phone owners probably have Amazon speakers in the house, etc.)
So assuming their survey's respondents resemble something like the general population, because we know that Android has a much larger share of the market than Apple with regard to devices owned, this means that there is a larger pool of the survey's respondents who own Android devices, yet only the same number of respondents use GA as use Siri. That means that a much lower percentage of Android owners use GA as compared to Apple owners using Siri.
That then is what brought me to the questions about why a larger percentage of Apple device owners use Siri, even though all the online chatter likes to say that GA is a 'more capable' digital assistant.
...
One of the two surveys was US only, where iOS is the predominant mobile OS. ...
Really? I don't think so. Apple may currently have a plurality lead in US device sales, that doesn't appear to add up to being the majority or predominant mobile OS in the United States.
Mobile Operating System Market Share in United States Of America - March 2019
iOS
54.74%
Android
44.89%
Windows
0.13%
Samsung
0.07%
BlackBerry OS
0.06%
Unknown
0.04%
Another fun fact: In general (Australia and Sweden being notable exceptions) iOS market share v. Google Android has been rising ever so slightly over the past three years. Android in China is an entirely separate market, using forks of the Google developed Android, but for argument purposes may or may not be included in articles touting one OS over another depending on how it benefits the storyline.
A bonus fun fact: Both iOS and Android have been losing market share in India. For what operating system you ask? KaiOS, based largely on the old and now abandoned FirefoxOS. Google is one of their investors.
Statcounter tracks page views on partner websites to determine what browsers and, ostensibly, devices are visiting those sites. That could be an indicator of OS market share, or it could be seriously skewed if certain browsers tend to be used more for surfing the web. Consumer Intelligence Research Partners (CIRP), the source of the data used in the AI article I linked, creates their data by directly surveying consumers.
So what would that mean? That would mean that getting at accurate market share data is highly dependent on methodology. If CIRP does its due diligence in sample selection for its direct consumer surveys (again, maybe an AI writer can get more details about that), then directly asking respondents "what device are you using," and "what OS is it running" will get a more direct answer about market share.
Personally I'm always suspect of results based on asking someone what they use, or what they plan to do, or pretty much anything where the honesty and understanding of a question of it posed to a respondent is necessary. Even the wording can give different results. Look no further than "Hillary is going to win in a landslide" according to consumer surveys. The coronation was indefinitely delayed.
I've had too many people refer to some Android phone as an iPhone (aren't all smartphones an iPhone to some people? Kinda like Googling and Kleenex) to trust the accuracy of a consumer response. I would tend to trust Statcounter more, but that's me.
Every methodology has its weaknesses, but I think Statcounter's jump from 'device page views' to 'device market share' is an unsubstantiated leap that ignores potential biases that would be baked in. As for survey respondents thinking all smartphones are iPhones (like all tissues are Kleenex), that would tend to cause the survey to overstate iPhone market share, and the survey already says that iPhones hold a minority of the market. So correcting for that sort of error would only lead to the conclusion that Apple has an even smaller market share. So then you can go back to the original thing I was commenting on, which was: if Android has the bulk of device marketshare, but similar numbers of people are using Siri as are using Google Assistant, why is that? Maybe GA isn't so much more capable after all, or maybe all those Android users with limited data plans don't want to spend their data resources sending voice queries to Google. I don't know.
No, what it would indicate is CIRP's data is even more suspect. I know you'd like to keep relying on it to form the basis for your presumption and followup "if this then that", but it's not reliable data in the first place.. "If" is a perfectly valid question, but positing it is not proven by CIRP data for obvious reasons which you've now understood.
Comments
One of the two surveys was US only, where iOS is the predominant mobile OS. The other used 5 countries, three of which (Australia, US, Canada) Android is behind iOS in user-share, one where iOS and Android are tied (UK) and one (India) where the most of the languages needed for the populace were not yet available for Google Assistant (and Siri?) but Cortana was AFAICT. There's got to be a reason why the specific countries were chosen for inclusion.
So you missed a few details and the surveys really don't say what you think they do. I'm not certain some commenters actually read the entire story, you included.
I'm more than a bit surprised that the author said in his lead: A Microsoft study suggests that Apple's Siri is on par with Google Assistant for worldwide usage, despite the latter being available on a wider range of devices" since it was NOT a worldwide study to begin with. The 5 countries included in the results seem to have been cherry-picked to cast Cortana in a good light.
http://gs.statcounter.com/os-market-share/mobile/united-states-of-america
Another fun fact: In general (Australia and Sweden being notable exceptions) iOS market share v. Google Android has been rising ever so slightly over the past three years. Android in China is an entirely separate market, using forks of the Google developed Android, but for argument purposes may or may not be included in articles touting one OS over another depending on how it benefits the storyline.
A bonus fun fact: Both iOS and Android have been losing market share in India. For what operating system you ask? KaiOS, based largely on the old and now abandoned FirefoxOS. Google is one of their investors.
I would offer a hypothesis that since Apple does not pursue the "low end" of the phone market, but many Android device makers do pursue that market, it's likely that iPhones will tend to be tied to phone plans with higher data allowances, while many more AndroidOS devices will tend to be connected via phone plans with much more limited data allowances. If that's the case, then the average iPhone owner probably clicks on a lot more web pages than the average Android phone owner. It's also reasonable to think that for every pricey iPhone sold, several cheap or free Android phones (attached to those less expensive, limited data plans) land in a consumer's pocket.
So what would that mean? That would mean that getting at accurate market share data is highly dependent on methodology. If CIRP does its due diligence in sample selection for its direct consumer surveys (again, maybe an AI writer can get more details about that), then directly asking respondents "what device are you using," and "what OS is it running" will get a more direct answer about market share. Statcounter, on the other hand, specifically says "We provide independent, unbiased stats on internet usage trends. We do not collate our stats with any other information sources. No artificial weightings are used." That sounds great, but what it means is that they are only measuring page hits, and if they aren't "collating their stats," then they really can't claim to be measuring device market share. Page views are not a direct proxy for device market share. Given that serious deficiency in Statcounter's methodology, I'm more inclined to accept the CIRP data that says Apple has a minority share in the US mobile device market.
I've had too many people refer to some Android phone as an iPhone (aren't all smartphones an iPhone to some people? Kinda like Googling and Kleenex) to trust the accuracy of a consumer response. I would tend to trust Statcounter more, but that's me.