IE10 falls below IE9 in market share
thenextweb.com
thenextweb.com
I'm not too familiar with IE (heh), but is the upgrade mechanism better in IE10? That might explain the higher drop.
I'm guessing this has more to do with Chrome being easier to integrate into enterprise/group-policy-based deployment than anything to do with the choices of individual users, though. The only place I see Firefox on a PC in a "computer lab" or "thin-client" setting any more is when those PCs are also running Linux.
...and Adobe Acrobat (Reader) that tries to install Chrome like "adware" - one has to be careful to opt-out on the Adobe website. Also many freeware and some open source like WinSCP come with this kind of "adware" like shady auto-installer and often install Chrome (opt-out is available, bit hidden in advanced options).
For the vast majority of users, none of this stuff matters. What matters is brand recognition and stuff "just working". In-built Flash probably helps Chrome's case.
It doesn't matter which browser is "winning", the vast majority aren't picking which browser to use based on features. If Chrome is the "superior browser", why is IE dominating in market share?
[1] http://www.microsoft.com/en-gb/download/details.aspx?id=4072...
[2] http://gs.statcounter.com/#desktop-browser_version_partially...
Windows 8.0 will automatically update to Windows 8.1 so IE10 on Windows 8 will get automatically updated to IE11.
> They also publish update blockers [1]
You say this like it's a bad thing, some users don't want an evergreen browser. The point of the update blocker is to help those users.
For them, an update just slows them down or leads to confusing changes. What they already have "works fine."
The difference between "auto update" and "an option to update" is huge.
I think that's the point. Users are actually upgrading from 10 to 11.
Its not even a new phenomenon, IE 9 plunged quickly (well, for an IE version) when IE10 started getting uptake, falling below IE 8:
http://gs.statcounter.com/#desktop+tablet-browser_version_pa...
If you want bad news for Microsoft, it's not in the rankings of their previous two versions, but the fact that their decreasing market share is spread across so many versions, each of which is a bit pitiful when compared with a single version of Firefox, with even the last version of Chrome and iPad Safari showing up Microsoft's efforts:
http://gs.statcounter.com/#desktop+tablet-browser_version-ww...
What matters - beyond a certain reasonable minimum size, which is a few hundred - is if the sample is random and unbiased. Both NetApps and StatCounter are far, far above the minimum size, so the only questions are the other factors.
But more important to realize here is that they measure different things. StatCounter measures pages viewed, NetApps unique users. There is no reason to expect those to match up.
My understanding is that they basically look at the total number of unique visitors from each country that they get, then weight that number by the number of reported internet users in that country. So if NetApps sees 10 users from country A and 5 users from country B, but country B has twice as many 'connected' people as country A, then they both get equal share.
What this basically leads to is that China, which has a very large population of people who infrequently use the internet but are still counted as internet connected (and may be using shared computers), gets inflated numbers based on total population. This is also why IE8 numbers are so large- because it's still in wide use in China.
The NetApps number probably makes sense if you were to ask 'what is the browser use of the individual users across the entire planet that could possibly load my website'. But when looking at browser usage by # of page loads or predicted visitors, the numbers could very well be much closer to StatCounter.
Of course it could be wrong in practice, if the data used to re-weight is misleading (while the unweighted raw data was more accurate). That's possible in theory, but seems less likely - there is fairly good information about internet usage in general which is what is used to re-weight, certainly compared to browser share.
http://www.akamai.com/html/io/io_dataset.html#stat=browser&t...
I guess that could support both sides at different times. But NetApps has not always been an outlier, I wouldn't say.
To your other point, election statistics is much more delicate than you make it out to be. They are also not truly random samples, but they are different for each poll--which helps. They also weight the results by location and expected turn out which is a whole different set of assumptions. If surveying the first 1000 people to answer their phone with consent and using the raw numbers was enough, then there wouldn't be any significant differences in the polls.
Then they have no statistical relevance. None. You can't make up for not having a random sample by having more results.
>>They are also not truly random samples, but they are different for each poll--which helps.
They are very close to random samples, which can be proved statistically after an election.
Consider a population where 99% of users use IE, and 1% of users use Chrome.
From that population you can draw a biased sample consisting of the 99% IE users and conclude that there are no Chrome users at all.
Or you could draw a random sample of 1% of the population, and assuming that 1% adds up to enough people, your chances of drawing a random sample that did not include a reasonable number of Chrome users too would be extremely low.
So depending on methods used, even a 99% sample may be entirely meaningless when compared to a 1% sample.
You can, however, still build confidence intervals based on biased samples. In your example, with 100% certainty, between 0-1% of the population uses Chrome. Yet of course, even the 99.999% CI will be wrong due to the severe bias. Now, if in your example only a 1% biased sample were looked at, the 100% CI would be 0-99%. Much less information. Note that you may also still see trends in biased samples if the sample is consistent.
If biased samples were meaningless, then how are Stat Counter or Net Applications results valuable at all since they are not random samples?
As mentioned above, they measure different things - page views vs. unique users. That is, usage vs users. There is no reason to expect them to be identical.
40,000 sites / 200 countries = 200 sites per country on average. How many participating sites do they have in smaller countries?
But, even 200 sites per country is enough for a proper random sample (that's about the same size as in election surveys). Even more so when you consider that a site is not a single observation, but reports data about all the users visiting it.
Yes, but my concern is that they may only have a handful of sites (or none at all) for a long tail of smaller countries.
>Even more so when you consider that a site is not a single observation, but reports data about all the users visiting it.
True, so it depends on whether or not users of popular international sites from a particular country are representative of all users from that country.
It might not be the only relevant thing, but it sure is relevant.
I wouldn't trust any blogspam that references the data, it's so out of align with other sources that reference billions of pageviews.
In some ways 'what browser do people use' is a bit like 'what language do people speak', except it should be easier to measure the statistics. Yet, unlike the situation with languages, we have a really hazy idea of what browsers are used. As a developer it is easy to get things working in developer preferred browsers to then be clueless about what percentage of your audience use IE and what versions. Does IE7 matter any more, or IE9? Do people with those browsers almost expect sites to look wrong?
I use FF27 on Ubuntu, no wait I use FF28 on XP, no IE9 on Win7, no it's Safari, no Opera on Wii, no Opera Mobile, no ... it's all of them. But then I'm a special case, however I imagine that a lot of people use various browsers - home, work, mobile, etc..
As with language just having general stats is no use - what's it matter that Chinese is so popular across the world if I'm selling canoes to Amazonians?
IE7 matters if your users use it and you want to make sure they have a good experience. It's not so relevant that most sites don't have high IE7 numbers, what does your site have; what will that support costs, what revenue will that support gain.
It is hard to get a random sample of internet usage. There is just no easy way - how do you do it?
In theory you would pick a random router around the world, and pick a packet going through it, and if it comes from a web browser, count that. Then normalize by traffic amount in that router. That would, in theory, give the right result with a large enough sample size. But no one (except perhaps the NSA) can do it.
On the other hand, there are at least two companies that pretty much know the truth - Google and Facebook. A huge percent of people visit those sites daily, so much that they don't need to randomly sample, worry about bias factors, etc. - they can just look at the data they have. So internally, they have the numbers. But apparently they consider that information valuable, and do not share it.
I have never worked on a site where the breakdown is close to NetMarketshare's figures. The ones from Akamai have been much closer, particularly internationally:
It depends what sort of sites they use the most, and just the sheer number of sites on the Internet, most of which have been around for a long time, means that only the newest ones tend to be incompatible (and as a data point, HN is usable with IE6.)
StatCounter measures page views.
Those can be very very different, if some users browser more or if some users use more than one browser, both of which are probably true.
So it is very possible both are accurate. Or that both are wrong.
But they also browse (far) less compared to people who use alternate browsers. So statistically, your sites have a much higher chance of registering Chrome and Firefox users than IE users.
Add: OTOH, I think StatsCounter, Clicky and Wikimedia stats are more useful. They represent the "active" Internet population better. The ones who are more likely to come to your/any website.
It's also so important to filter down to geographic regions you care most about when looking at these stats. It's common to see developers/companies making browser support decisions based on global stats, even when they often derive most of their profits from a much smaller region (with different browser usage stats).
Other commenters have mentioned differences between NetApplications and StatCounter, but I suspect there are other factors that can heavily skew the apparent market share of different browsers.
For example, it is already well known that programming-related websites get a higher than average share of non-IE visitors. I don't think this would affect the likes of NetApplications and StatCounter, because they're used by all sorts of websites. But ...
... it is also much easier to find and install an ad blocker in Chrome and Firefox than it is to do so in IE. Many of those ad blockers and similar extensions will also block the tracking beacons used by NetApplications and StatCounter.
This probably has a negative effect on the apparent market shares of Chrome and Firefox. Especially the latter, I suspect, because lots of people who consciously choose Firefox nowdays do so for privacy reasons. (Can't trust Google anymore, etc.)
And IE11's dev tools, while still a little clunky to get around, are very comparable to Firefox/Chrome for most of the features.
We also still have some users that haven't upgraded (also smaller resolution monitors) but all less than 10%.
http://www.mobilexweb.com/blog/android-browser-eternal-dying
There are still some legacy applications that require a specific version of IE, or specific IE functionality, in order to work correctly. The cost of "fixing" the application to work correctly may cost too much or be too much effort.
Of course, there may be others who may have no idea how to update, no doubt there is a whole group of (older) users where auto-update may be turned off and no idea the update process exists.
It's a pain in the ass to support this.
Think of % of 2.x.