Selling Ubuntu to the "Third World"
Ubuntu adoption for communities in the “Third World” seems like it should be a no-brainer: how could a functional, free operating system not prove wildly popular in developing countries? Nonetheless, I believe Ubuntu use outside rich nations remains limited. Here’s a look at some suggested explanations of that reality, and how to change it.
Counting Ubuntu users by country–like counting Ubuntu users in general–is surprisingly difficult. There used to be a frappr page, a map run by ubuntu-fr.org and a world map hosted on the Ubuntu forums all dedicated to this purpose, but these resources no longer function.
However, there are still a few data sources available suggesting the extent of the disparity between Ubuntu use in the first and third worlds. The Linux Counter has some statistics broken down by country for Linux use in general. Granted, this data is self-reported and represents only a tiny fraction of Linux users worldwide. Nonetheless, it strongly suggests that there are plenty more people using Linux in developed countries than in developing ones.
This difference can be easily visualized using a snapshot from May 2006 of the old Ubuntu Forums worldmap provided by the Internet Archive:
That data is old, and subject to various complicating factors, but the difference between the density of Ubuntu users in North America and Europe on the one hand, and the rest of the world on the other, is striking all the same.
Explaining the Difference
Glyn Moody of The H recently wrote about the surprisingly poor rate of Linux adoption in the Third World. His explanation of the trend echoed that of another recent article on ghabuntu.com that cited lack of affordable bandwidth as the chief obstacle to free-software uptake in developing countries.
Moody accordingly called for private individuals to help Canonical expand its ShipIt program via donations in order to make Ubuntu more accessible to people without the means of downloading an ISO.
That may be part of the answer. The ShipIt program has indeed been in a flux since Canonical announced major changes last October.
But I don’t think expanding initiatives like ShipIt (or the Freedom Toaster project, which provides free Linux CDs in select regions via a different strategy) is the key to bringing Ubuntu to the Third World. Nor am I convinced that bandwidth is the problem–or at least not the chief problem.
Beyond Bandwidth
After all, Ubuntu uptake in developing regions remained limited even when ShipIt was in full swing. There’s no evidence that mailing massive numbers of Ubuntu CDs out for free is going to change things.
In order to increase Ubuntu use in the Third World, Ubuntu needs to become more appealing to the Third World. Lack of attraction, not lack of bandwidth, is the greatest obstacle to surmount.
A good first step towards increasing Ubuntu’s appeal would be improving translation initiatives. According to Launchpad, the Afrikaans translation of Ubuntu, the most complete of all African languages (I know, Afrikaans is not historically African, and South Africa is arguably not part of the Third World, but…), is only about ten percent finished, ranking below more than sixty non-African languages.
Some Asian languages that might be associated with developing countries fare a little better, but still have a long way to go to compare to Ubuntu’s European-language iterations.
Granted, the translation effort depends heavily on volunteers, and represents a sort of chicken-and-egg problem: if Ubuntu is not popular among speakers of a certain language to begin with, there will be few people to translate it into that language.
Nonetheless, recognizing the gross inadequacy of Ubuntu’s translation into the languages of Third World countries is an important component of making Ubuntu more attractive to the people living there.
The second major obstacle to Ubuntu adoption is the prevalence of pirated copies of Windows in the Third World: Ubuntu becomes less attractive when Windows can also be obtained for free. That’s not something Canonical or the Ubuntu community can control.
But what Canonical can do is convince hardware vendors in developing countries to ship low-cost computers with Ubuntu preinstalled, which is exactly what the company, in partnership with IBM, has started doing. That initiative may not make Windows piracy go away, but it will make pirated Windows that much less attractive.
Ubuntu will not take over the desktops of the Third World–or the First World–overnight. But there are some things that Canonical and the community could be doing more of to make Ubuntu more desirable and accessible beyond the borders of rich countries. In order to live up truly to its namesake, these are initiatives that Ubuntu needs to pursue.
I live in central Brazil and pirated Windows is rife. Vendors in the streets of Sao Paulo and Rio sell pirated versions of every software title (and Movie) that you can think of.
Virtually any small PC repair shop or service centre will install or sell pirated versions of XP, Vista or 7.
This means that Windows has an equal footing (price) to Ubuntu but everyone naturally goes for Windows as that is what is available and familiar.
I recently bought a new boxed Acer Aspire One in Paraguay that had WinXP installed. It was only after I got it home I found out it was the Linux version that had a pirate copy of XP installed over the top.
Fake Steve Jobs already covered this:
1. Windows is better than Linux
2. In the third world, both are free of charge.
Naturally, with two choices that are free of charge, people are going to choose the better one.
http://www.fakesteve.net/2007/07/freetards-in-deep-denial.html
No amount of marketing, ShipIt CDs, partnerships with hardware manufacturers, or revamped themes will fix the fundamental problem with Ubuntu: Nothing works right.
Make Ubuntu run smoother and interface better than the alternatives, and you’ll have a winner without even trying.
I come from a third world country called Sri Lanka. Not surprisingly, it hasn’t been marked in the map as a country that has adopted Ubuntu. The main reason why my fellow countrymen don’t like to use Ubuntu is the lack of a good offline software installing system as you get in Windows. As many people don’t have access to the Internet(or even if they do it’s mostly either gonna be a dial-up connection or a limited broadband connection of 1GB) they prefer to go with Windows, of which you can find pirated copies aplenty along with software that works on it.
I once tried to find a good solution for this problem so that I can convince my friends and family to use Ubuntu. But that was in vain. Neither did Ubuntu guys out there were willing to help me when I told them about my intentions to make an online website that would help users to download software from Ubuntu repositories (along with the dependencies)for offline installation. As I’m not a good coder I gave up. But I still wish there’s someone out there who’s capable of providing a good solution.
immy: APTonCD sounds like it’s what you’re looking for in order to be able to download software more easily for offline installation. If you haven’t seen the project’s website, I’d take a look: aptoncd.sourceforge.net
Soviet Union still exists on that map. So unprofessional.
I think a project like apt-sync can help the users that use a low bandwidth infrastructure. A feature like in fedora that permit to download only the differences from an old package to the new (delta).
https://wiki.ubuntu.com/apt-sync
Thanks for the info guys. I’ll sure have a look at them. 🙂
Hi. Thanks for the mention in your post. I think I disagree with your point of language being a barrier to Linux adoption here in Africa. If people, and I mean educated people, do not even know that a computer can run anything other than Windows, then translation would not help in anyway whatsoever.
I can of a surety say that almost 99% of all the software and OSs here are in English, French or Arabic, the three major languages of Africa.
You see, unlike Hmm above is claiming, Ubuntu is good, but the problem of it not doing great in places like here in Africa is because it is relatively unknown and bandwidth restriction.
Though I’m an Ubuntu proponent, for first timers, I give them Mint because it needs no more work to get multimedia to play. Then like Immy said, you would need to be online to install software, that is really a no go zone.
If I have 1GB/month, I doubt if I would be willing to spend some downloading and installing software. For Ubuntu to make any gains here, it would need to
1) be on the ground here. Example, you would have more people interested in Ubuntu should we see on the news that Canonical has donated X number of computers to some deprived school here. You would then have more people beginning to ask what is Ubuntu.
2) software installation should be online independent as much as possible. That would go a long way to help
3) the install fests should be encouraged in our tertiary institutions here as much as possible, obviously with some visible backing from Canonical.
If I maybe permitted, it is in this regard that I have set up http://www.omgafrica.net to help bring together people both from your end and here to deliberate over the spread of Linux in particular and Open Source in general here in Africa and other developing countries.
Open Source has a great role to play in all spheres of life here, but a lot of work needs to be done before its dominance (at least reasonably) can be realized.
I don’t agree with you on a lot of your points. I am a linux advocate living in a “third world” country and the fact that Pirated copies of Windows are even more widely available than Ubuntu discs is definitely the most significant factor affecting the popularity of Linux or any other Open-Source O.S.
The cost of accessing the Internet is also another factor, at about $0.10 per MB, few people are willing to pay that much to be able to install extra software.
This is right, in developing economies the alternatives are: do you get a pirate version of windows from any of your friends, and get help from any of them to fix the infinite number of problems you’ll have? Or … well, or you have an OSS guru around, which is a lot less likely.
MS’s strategy has been a no-brainer: look the other way. And wait for the economy of the country to pick up to crack down on people. In many cases they do go after commercial users (businesses). It is in their best interest to keep the monopoly going, whatever it takes.
The big missing factor here is Governments and businesses. These should be the targets for Ubuntu/Linux. Governments need access to the source, they need openness, transparency. They get all that with Linux. Business need to look at costs even more than in developed economies. Again, Ubuntu should have an edge.
Does anyone who is responsible for generating the map image Chris has used in this article. There are better ways to visualize density which I can help them with.
For example of how I think you can get a better visualization look at this:
fedoraproject.org/maps/all.png
In Fedora we are generating unique IP address density maps..heat maps..from our MirrorManager logs in an effort to give a global picture of where Fedora is being actively used. I’ve gridded unique ip address clients counts based on a lat/lon grid and then projected that back over a global map.
One of the reasons I did the client density heat map was because the simple scatter plot of geoip markers was too crowded to really say anything significant.
If the Ubuntu forum people are interested adapting our map generation script.. it can pulled from our Fedora infrastructure git repository so they can produce similar heat maps from their forum connection logs. I’m more than happy to help with the adaptation if they run into any problems.
The 2006 map is probably a very out dated picture. I’m much more confident in the Fedora usage map image I’ve included, which tells a somewhat different story. We are seeing significant pockets of usage in India and in South America, to the extent that we can trust the geoip conversion. I’d very much like to see the Ubuntu community use a similar approach so we can get a more comprehensive view of global linux adoption that we can trend over time.
-jef
Once again the problem is third party applications and developers. OpenOffice.org and Firefox are excellent professionally developed and finished applications. However that’s not enough.
If Linux is ever going to take over the desktop then all the applications need to be up to scratch. There is absolutely no point in the Ubuntu repositories being chock full of half baked not even nearly finished games if the alternative accountancy software doesn’t do everything the proprietary Windows based counterpart does.
It’s not enough for free or open source software alternatives to do nearly everything or mostly everything the Windows based competition does. Free software alternatives must do it all, do it better and do more.
That is the nature of competition. People will not turn from their expensive ball and chain unless the competition is better. That’s how Apple captured the MP3 download market. They produced a far superior product than the competition.
Open Source and Free Open Source Software developers clearly don’t have the resources to do everything on their own. Companies like IBM are not interested in developing a vibrant ecosystem. Their interests lay closer to home. Not that their contributions aren’t welcome.
If Canonical really wants to conquer the desktop it will need to seriously invest in applications software development and marketing. It will need to develop partnerships with third party developers and OEMs. IBM and Dell are good starts. But they’re playing the wrong end of the field.
Everybody involved in OSS and FOSS should ditch the fanatical ethos that a Linux OS must run on a 286 with 64KB of RAM. We live in the modern world. People who live in the third world want to live in our modern world. They don’t want to be patronised with something designed for yesterdays technology. Even if that is all they have. They want to feel like they’re on par with the rest of us. Which is why they foolishly hold up Windows as the gold standard to beat. So we need to stop selling Linux short.
Yes it can run on old machines. That’s not such a great marketing slogan though. It might be what many people want. But it doesn’t inspire people to buy into the dream.
It might not strictly be a third world concern. However the gaming industry has repeatedly expressed an interest in Linux. Every now and then a games developer dips it’s toes in the Linux waters. But there’s no corporate entity for them to engage with. To partner with. Canonical certainly aren’t embracing them. So they wander off to another platform.
They adopt the Xbox or the Playstation or the Wii because the vendors of those systems bend over backwards to make them accessible. Sony, Microsoft and Nintendo court developers and establish lucrative deals and relationships. In return games developers develop games for these platforms.
Key to getting Ubuntu accepted on the desktops of the third world will be acceptance within the business communities of the first world. Software that is useful to first world business leaders will be equally useful to third world business leaders.
After all, if we’re really trying to develop third world economies and markets. Why are we treating them like dumb apes?
[…] Selling Ubuntu to the “Third World” Ubuntu adoption for communities in the “Third World” seems like it should be a no-brainer: how could a functional, free operating system not prove wildly popular in developing countries? Nonetheless, I believe Ubuntu use outside rich nations remains limited. Here’s a look at some suggested explanations of that reality, and how to change it. […]
Yes, I believe hardware is key. To begin with, many people don’t have the machines to install things to. That map could be a machines/per capita result. And in these countries you need to sell stuff cheap, such as MenQ’s 80$ ARM netbook.
Price is not the issue, what remains is quality.
Pre-installed is important, as many people don’t have connectivity either.
This is for linux in general. For Ubuntu, it helps if the SABDFL please abstain from involving himself in politics of foreign countries. That hurt A LOT when a government is moving towards migration… Now they use a much less user friendly distro and disillusioned people go back to Microsoft.
Its quite expensive to run a full desktop OS on a cheap netbook or old computer. Netbooks are best used a a complement to a full sized pc or notebook. No-name brand or self built computers are still relatively cheap. I beleive that the overriding reason for limited adoption is more structural than anything else: The cost, quality, reliability and availability of power supply and Internet connectivity increases the risk of purchase of electronic equiptment. I worked in a company that had a battery of book-keepers writing out payroll checks by hand because power was rationed in a way that a payroll run could not be completed any other way. Consider going to a govenment office or a bank and not having service because they or someone they need to communicate with have no power or Internet connectivity.
Those people that are likely to have computers in the home are unlikely to adopt linux if they are required to go to the Internet (often by lugging a desktop to an Internet Cafe) to download software or update. They can buy or pirate windows and subsequently obtain software on disks that will not cry out for an Internet connection to download dependencies. How can you get games, and other software, writers to have confidence in Linux if their market suffers this risk.
If the power and Internet connectivity problems are resolved then there will be a tremendous momentum to adoption of the Internet.
This article misses completely the point from my point of view. I am writing you from India, a “developing” country and the main obstacle for the low success of Open Source is a piracy rate in India fro example of 75% !
Most businesses and even the government offices here run unlicensed software without that the international community (including Microsoft, Adobe etc) can do much about it.
Microsoft doesn’t do anything is the first 10 to 20 years in a developing county because they want to make sure that government and industry become dependant on their software. Once that has happened they start to make those countries pay the bill. Even laptop and Desktops sold in India with Linux are sold only on paper with Linux. Lots of dealer deliver them with a unlicensed XP or Vista version.
Who wonders why people here don’t change? They are used to Windows and it’s practically “free” in India together with Adobe and all the other software. And one hardly gets support for Linux. I think that are reasons enough to see why Linux doesn’t make it in developing countries.
The graphic is old…
In 4 years MANY things have changed.
I’m from Brazil and here Ubuntu would be a very useful system… if it have the most common “winmodem” (soft modem for dial-up conections) drivers out of box.
Others issues like… easy to use… simple, programs, etc… are not a problem (obviously… except for those who need professional tools like Auto Cad).
There’s no problem that Canonical could deal with…
Any Apple’s product is very expensive here… but it is growing… and growing fast… first because of iPod… and now it is iPhone. The Mac computers also are getting space in our market… but there’s no special reason for that… too.
Side note: Christopher, let’s be more politically correct and not call developing nations as “Third World” anymore. While many people understand the concept of “Third World”, it’s really outdated and by it’s very name biased for the “First World” and “Second World” countries.
On topic: I think Ubuntu’ (and other Linux distros’) fighting chance is through small and medium sized businesses. Software vendors occasionally clamp down on pirated business users, so using FOSS is a *real* alternative for these SMBs.
MS Office = OOo (unless you are running Access)
Photoshop = The GIMP (unless you’re business is graphic design)
Outlook = Web-based email
Now the main problem with SMBs is that many of them run Windows-based business software, both custom-built databases and off-the-shelf accounting software. WINE would be the obvious solution, but that’s a hit-or-miss experience. So maybe Canonical should find ways to encourage other developers to create GNU/Linux-based business software.
To be totally honest. So long as Windows is pre-loaded on desktops and Linux keeps trying to accommodate the Windows orientated hardware Linux will always be in second or third place.
If you take it a little into consideration what hardware you buy, Ubuntu runs smoothly. It’s just if you buy anything and then want to run Ubuntu on it, you might experience problems. But this is also getting better and better. Ubuntu 10.04 comes up immediately after install telling you (e.g. for NVidia graphic cards): “Hey, there is a probably better fitting, but non-open-source driver available.” – It is a one-click to download and install the other driver.
Of course, some vendors simply don’t get it, that they will loose market share in the long run if they don’t offer support for Linux…
It is quite sad. I can give an example of an NGO where I work. The hardware is about ten years old. I managed to get Xubuntu onto my own workstation, but getting my colleagues to change from Windows is very difficult. They are all afraid of change. We need a different distribution model that would encourage people to shift. I keep thinking of the way tupperware and herbalife products are distributed.