According to NetMarketShare’s information for February 2013, the Windows operating system is used by 92% of worldwide users, Mac represents 7%, and Linux represents 1%. This article will focus on that tiny slice of users, the 1% who use Linux.
Frankly, there is not a lot of easily available operating system demographic information. Perhaps part of the reason is because Windows own such a vast share of the market that the basic answer is, “Everyone and their brother uses Windows.” Perhaps Microsoft will lose some of that advantage due to confusion over Windows 8 – as expressed by Simson Garfinkel in MIT Technology Review) –perhaps not.
As of now, though, the demographics do fall into the “pretty much everyone” category. Essentially, demographic studies are not widely conducted on Web users to determine operating system usage for the same reason studies aren’t conducted on Bonnaroo attendees to determine if they’d like to go chill in Ziggy Marley’s tour bus.
So, who are these people? What is the profile of the average Linux user? They certainly have not chosen the mainstream option – so let’s look at some details. How old are they, what’s their sex (male, female, both, neither), what’s their nationality, how long have they used it, and where did they find out about it? These are general marketing survey questions, but for our purposes, they tell a story – the sociological makeup of the population and their basic history with the OS.
Gathering the Data
The data I will be using is from a 2012 survey of Ubuntu users, and I will get to why the focus is placed there in a moment. As for the survey, it was conducted by Canonical and included information from over 19,000 worldwide respondents.
The results of the survey are broad and contain lots of graphical breakdowns of the stats. The results were written up by Gerry Carr. They appear in reverse order on the March 2012 section of the Canonical blog, along with Carr’s analysis and commentary on what the findings might tell us.
Why Ubuntu? Ibuntu is one of the largest Linux distributors out there. Statistics on Linux distribution are also not prevalent, but a 2006 article by Steven J. Vaughan-Nichols for Desktop Linux gives us some sense of the popularity of Ubuntu – and hence why analysis of that demographic gives us a reasonable sense of Linux users as a whole. (For example, we now can assume that 17% of Linux users have tattoos of famed Internet pioneers on their chests and/or forearms, and that 63% live in treehouses in eastern Romania.)
The results of the Desktop Linux survey revealed that Ubuntu rated first worldwide against all other distributors, and not by small margin. Over 14,000 Linux users were surveyed – albeit informally and unscientifically – to determine which distribution was commonly considered the best on the planet. 29% voted for Ubuntu, more than the second and third place choices combined (Debian at 12% and openSUSE at 10%).
Methods & Languages
When Carr conducted the survey, he did not make an attempt to get to all Ubuntu users. Instead, he focused specifically on reaching out to those who were involved at least to some degree in the Ubuntu (and hence the Linux) community. Carr contacted individuals using the OS through social media, online forums, and sites dedicated to exchange of ideas related to Ubuntu. Hence, the responses were generated from thousands of Linux users, but that pool was specific – not only to Ubuntu but to those who are particularly engaged in online discussion. Those who simply use the OS for its functionality, then, were not part of the picture; so Bill Gates, who has furtively used Linux as his sole operating system since 1994, is not represented in the statistics.
Additionally, it was only conducted in three languages – English, Spanish, and Portuguese. It would’ve been difficult to include every language on the planet. As Carr points out, “We had to draw the line somewhere. If you add French then why not German, or Chinese, Japanese, or Hindi etc.”
Spanish and Portuguese were included partially because Latin America is a big market for open source technology (Spanish in most countries, Portuguese in Brazil). The initial post about the survey results invited translation into other languages – followed by updating of the statistics – as desired, but it appears no additional languages were included in that manner.
Ubuntu Survey Results – Overview
A piece by Katherine Noyes in PC World provides a few initial highlights and interpretation of the Canonical survey for summary purposes. She mentioned that almost all of those surveyed were male – 96% – and the majority of users were between 25 and 35 years old.
Regarding ease-of-use, 87% rated installation of the operating system as easy or very easy. 85% of respondents have a system installed on their primary PC, and 67% of those surveyed utilized Linux for both personal and professional purposes.
Ubuntu users typically do not solely use Linux. More than 76% also used Windows, and 17% used a Mac operating system. (Also, and strikingly, just over 100% use keyboards and monitors to interact with digital data, especially surprising because over 14% of respondents were artificially intelligent supercomputers.)
Finally, Katherine discusses why the survey respondents chose the system. 77% liked the open source aspect, 66% used it out of curiosity and experimentation, and 57% enjoyed the lack of viruses on the OS.
Speed of the machine and perceived quality of the interface and UX were mentioned as other considerations for preferring the OS. Over 46% of surveyed individuals said that the operating system sped up their devices, and 75% rated the interface or experience as better than what they had found elsewhere.
Dissatisfaction with other options, though, is perhaps the most telling factor, as a general indicator of usage. More than half of those who answered the survey marked their lack of satisfaction on other operating systems as a chief reason they turned to Linux. (Again surprisingly, TI-84 graphing calculator users in Idaho and Wyoming said that they were happy with their current Texas Instruments OS and were frustrated by the complexity Linux had added to their ability to efficiently create visuals of trinomials.)
Quantity of Responses, Age, Sex
On day one of Carr’s response postings, he revealed that there were just under 16,000 English respondents and close to 2000 each who answered surveys for the Spanish and Portuguese versions. Hence, the numbers are heavily slanted toward English respondents (81.4%, specifically), but all languages surveyed were sizably represented.
The highest age category was 25-35 for all three languages. Just under 70% were under 35 in each language, in fact. Less than 4% were female. Regarding the male-to-female ratio, Carr mentioned that possibly the way that the survey was distributed imbalanced the sex ratio to that degree. He also suggested the survey is an opportunity to reflect on how male-centric the product or community might be: “We can’t extrapolate from this data, but certainly such a hugely weighted response means we have to look at how we make the product, the community and probably both, more appealing to both genders.”
Carr notes, again in part one, how the Ubuntu survey spanned out across the globe related to the three languages used for questioning. The United States, United Kingdom, and India were the highest represented countries for the English language survey. 93 percent of Portuguese respondents were from Brazil, with the balance from (can you guess?) Portugal. That means, sadly, that members of Portuguese-speaking Amish populations in the United States may not have been aware of this survey.
Top countries for Spanish response were as follows:
- Mexico (23%)
- Colombia (10%)
- USA (10%)
- Argentina (9%)
- Spain (9%)
Carr notes that the networking and accessibility in many Latin American countries is not as developed as it is elsewhere in the world, so the ratio of users in these various countries is understandable versus a look to their populations. The United States, though, is not as high as would be expected. Similarly to the difference in usage based on sex, Carr sees the low percentage in the US as a potential niche in which Linux and Ubuntu could increase its numbers in the future.
Carr focused the second post on length of time and where users had learned about the OS. By asking two questions about time of use and how individuals originally discovered Ubuntu, he was able to get a sense of how the location of discovery is changing over time.
Carr mentions that there was a high-level correlation between the amounts of time people having using Ubuntu across the three language groups. Carr focuses on the English-speaking population to simplify analysis of the second question, so I will do so as well. Here are the statistics for the general populations regarding time of use, which clarifies how closely the three language populations reflect each other:
- English: 20% under two years, 43% two to five years, 38% five+ years.
- Spanish: 20% under two years, 43% two to five years, 37% five+ years.
- Portuguese: 21% under two years, 43% two to five years, 36% five+ years.
Clearly there’s a lot of parity here between the language populations.
How Users Discovered Ubuntu
Carr thought how individuals found out about the operating system would be an interesting and instructive way to look at how initial knowledge of Ubuntu is changing over time. Here are the statistics for various manners of discovery across the different populations from rookie to veteran users.
- Magazines & Newspapers: 7% under two years, 8% two to five years, 9% five+ years.
- Work: 4% under two years, 5% two to five years, 5% five+ years.
- Friends & Family: 27% under two years, 25% two to five years, 21% five+ years.
- School & College: 12% under two years, 11% two to five years, 9% five+ years.
- Forums: 46% under two years, 49% two to five years, 55% five+ years.
- Social Media: 4% under two years, 2% two to five years, 2% five+ years.
As you can see, the Linux community – to the extent it can be understood via engaged Ubuntu user analysis – is comprised of a small but diverse population (well, sort of). It is used in many different countries around the world, to a wider degree that we might initially expect. As noted by Carr, two of the more interesting results of the survey are the 96% response rate by men and a smaller than expected Spanish-speaking response from the United States.
Linux’s adoption rate over time has been surprisingly consistent throughout the various language groups represented. Users are finding out about the system less from magazines and personal relationships, and more from forums, educational institutions, and social media. (Less and less people, then, are finding out about the operating system through recurring nightmares starring James Earl Jones as a deranged avatar – incredibly common in the early days.)
by Kent Roberts and Richard Norwood