I’ve found this [survey](http://www.accenture.com/SiteCollectionDocuments/PDF/Accenture-Mobile-Web-Watch-Internet-Usage-Survey-2012.pdf “survey”).
The information of this link seems barely believable : http://www.dailytimes.com.pk/default.asp?page=2013%5C05%5C05%5Cstory_5-5-2013_pg5_9
I believe http://www.w3schools.com/browsers/browsers_mobile.asp is something that you are looking for. There are lots of different statistics there.
> I wouldn’t believe anything that W3Schools puts out.
Oh, that doesn’t seem fair. If you want good, reliable data for the platforms visitors to W3Schools are using, that seems the best place to go. Not sure why you would care, but I suspect it’s accurate.
@rodolpheb: Most of the data you’ll come across isn’t contradictory. The problem is that the question you are asking is rarely what anyone actually measures. Site specific information is exactly that – site specific. Data on sales don’t reflect actual usage and questionnaires or surveys tend to have limited samples.
If you want the information for a site you are involved in, measure the usage of your own users. If there’s some other reason you want more “universal” data, just pick the published data that you think best represents your question. Just bear in mind that even “big” websites like facebook or google may miss whole demographics in terms of worldwide figures (eg. China), so even data published by them is likely to only represent part of the online population.
I don’t see anything in that link about their platform statistics. I understand the point you are making. I just didn’t think it was necessary or relevant in this discussion.
I’m not advocating using w3schools as a developer resource, but I thought the more significant point to make in this thread was not that the information may be inaccurate or false, but that the data itself is of limited value.
Take a look at these.
[A sitepoint Blog](http://www.sitepoint.com/browser-trends-april-2013-is-chrome-unstoppable/) and also [Statcounter](http://gs.statcounter.com/#mobile_os-ww-monthly-201303-201303-bar)
Thanks @BenWalker. I need these information to complete some answers about mobile devices.
Here’s my point: regarding the links that @unasAquila kindly sent the data aren’t exactly identical. For example android’s data are between 37% and 30% even though the comparison with the other is the same.
> Here’s my point: regarding the links that @unasAquila kindly sent the data aren’t exactly identical. For example android’s data are between 37% and 30% even though the comparison with the other is the same.
I suspect that is just an error in Craig Butler’s reporting. The section before looked at trends from March 2012 to March 2013. If you look at the Mobile OS data in statcounter for that period Android is reported as 30.43%, which is very close to the figure quoted. I’m not going to test every combination out to see what period is actually being reported, but I am reasonably confident it is just author error.
Either way, the sitepoint.com figures are simply a summary of statcounter.com’s data, so I would defer to the latter as being the authoritative reference in this case. If you are happy with statcounter’s methodology (with a superficial glance it seems a reasonable approach to me), then I don’t think anyone would criticise you for going with that.
Just try to be aware of the limitations of the measurements and make sure you understand what the figures actually mean. The site has a reasonable FAQ you should look at:
And also a breakdown of their sample population by region:
Again, cursory glances only, but China is (as is often the case) lower than I think it should be.
Depends a lot on where each report gets it’s data from. Often it’ll be from analyzing their own visitors and/or those of partner sites.
Depending on what those sites are about, or the people they attract, it can skew the data. For example sites about Apple are likely to see far more iOS users than android and vice versa.
In short: you can’t get that data.
The only data that companies can report is information they receive from users visiting *their own* website (or websites that they have made a deal with to track stats from). There isn’t going to be one resource that has the real answer.
You must be logged in to reply to this topic.