Now this is hilarious - Net Applications market share reports

[Queso]

Those of you into Web development will know about Net Applications, who track the usage of browsers and/or Operating Systems and use it to provide monthly reports on market share for various pieces of software.

Well, over the past few years both the Mac OS share and the Safari browser share have been showing significant gains. OSX in particular has been climbing in usage and for the past three to four months has hovered just below Apple's 10% "glass ceiling" for Web use market share. Anyone following the reports would have expected to see this threshold crossed in the June 2009 report.

However....

June's report didn't appear. At all. Net Applications instead issued a statement about statistical irregularities...

And now, it appears Net Applications have decided, completely co-incidentally of course, to change the metrics by which they calculate their figures. To quote:-

In the past, we reported only on our raw numbers. As of August 1st, we have implemented retroactive country-level weighting in our reports. This means that we adjust our reports proportionally based on how much traffic we record from a country vs. how many internet users that country has. For example, although we have significant data from China, it is relatively small compared to the number of internet users in China. Therefore, we now weight Chinese traffic proportionally higher in our global reports. This change produces a much more accurate view of worldwide usage share statistics.

After consulting with many of the organizations we report data on, we decided to use C.I.A. data as the source of the number of internet users per country.

In addition to providing better share numbers, the reason we made this change was due to growing traffic imbalances in certain countries. Some countries were growing traffic at a much higher pace than the rest of the world and it was creating unacceptable variances in the share numbers. The reason we delayed June numbers was due to these imbalances. From now on, a single high growth country will not be able to affect the global share numbers.

Now why would anybody have a problem with this, do you think? Because it smacks of external pressure being applied. The Web shouldn't care about national boundaries, so the amount of Net users in a particular territory is immaterial. The new metric also fails to take into account that the total of "internet users per country" fails to differentiate between someone who checks their e-mail once a month and someone who spends two to three hours a day on the Web. The ONLY data that matters to a Web developer ought to be the amount of hits from different browsers, and the underlying trends this implies. Where the generator of these hits happens to live simply isn't important.

To me, it seems rather co-incidental that this new method is announced in the same week that Steve Ballmer, CEO of Microsoft, dismisses Mac market share growth as a "rounding error". The fact that the raw data reports are also not being published beside the new reports smacks to me of serious statistical manipulation in order to avoid a set of pro-Apple headlines. By my reckoning this also makes Net Applications reports themselves utterly useless.

BTW, under the new system OSX market share shows as 4.86% in July compared to 4.73% in May (9.93% under the old calculation method). Although there's no way of saying for certain what OSX market share in raw hits now lies at it's a safe bet to say it is far in excess of the magic 10% barrier.

Net Applications explain their new weighting system

 

[clevin]

The new metric also fails to take into account that the total of "internet users per country" fails to differentiate between someone who checks their e-mail once a month and someone who spends two to three hours a day on the Web. The ONLY data that matters to a Web developer ought to be the amount of hits from different browsers, and the underlying trends this implies.

its only a problem here at MR, where people are unhappy with apple's number.

You are mis-defining the term, marketshare is just that, matketshare, it means how many computer users use this or that piece of software or, mac or pc. One user only contribute one count, a mac users going to a web 5 times more frequently doesn't give mac 5 times bigger market share than it should have. and a american user going to web 5 times more frequent doesn't mean his count has 5 times more weight than a Chinese one.

Whats really hilarious is that Macrumors has been enthusiastically using NA numbers for mac's success for so many years. and numerous people predict apple reach 10% by 2011, lol, I guess its a rounding error indeed.

Netapplication's figures have always been further away from any other global counters, while more closer to the US figures from those vendors. Its absurd so many famous online communities, such as ars, zdnet, cnet, etc, all buy into netapplication's obviously strange numbers.

 

[Queso]

its only a problem here at MR, where people are unhappy with apple's number.

You are mis-defining the term, marketshare is just that, matketshare, it means how many computer users use this or that piece of software or, mac or pc. One user only contribute one count, a mac users going to a web 5 times more frequently doesn't give mac 5 times bigger market share than it should have. and a american user going to web 5 times more frequent doesn't mean his count has 5 times more weight than a Chinese one.

Whats really hilarious is that Macrumors has been enthusiastically using NA numbers for mac's success for so many years. and numerous people predict apple reach 10% by 2011, lol, I guess its a rounding error indeed.

Netapplication's figures have always been further away from any other global counters, while more closer to the US figures from those vendors. Its absurd so many famous online communities, such as ars, zdnet, cnet, etc, all buy into netapplication's obviously strange numbers.

But that's sort of my point. Why, when a milestone level is about to be breached, do they suddenly change their methods? Is it simply to ensure that the various online communities, and the journalists that frequent them, don't report anything significant?

I also disagree with your definition of the purpose of the Net Applications reports. They were never about market share in terms of users or sales, they were about market share in terms of web hits. They are supposed to be a counter as to who is actually more active on the Net, not how many people have access to it. We already get numbers regarding sales, installed base, etc. from other sources. With this change NA have effectively destroyed the only reason for anyone to view their reports. What is the point of that unless external pressure or incentive is being applied?

 

[clevin]

But that's sort of my point. Why, when a milestone level is about to be breached, do they suddenly change their methods?

I also disagree with your definition of the purpose of the Net Applications reports.

no milestones can be set without facts behind it. if NA reports mac having 10% market share, while in reality only 4 in 100 of computers are macs, it would be stupid for them.

and You are entitled to your opinion, Im just saying you are probably disagreeing with 99% of people in this planet.

Think differently shouldn't be an excuse to twisting the reality. Something is up for debate, some aren't. No discussion or analysis can proceed if everybody were to think "differently" in their own minds and disagree over every commonly defined terms.

 

[MisterMe]

its only a problem here at MR, where people are unhappy with apple's number.

...

It appears that without this change, only a few MacRumors members would be unhappy. However, one of these unhappy souls would be clevin because Apple's numbers would be too high. All this does is to give these people longer to prepare themselves. Apple's numbers will burst through the 10% threshold and keep on rising. Don't fret. You will survive.

 

[xUKHCx]

no milestones can be set without facts behind it. if NA reports mac having 10% market share, while in reality only 4 in 100 of computers are macs, it would be stupid for them.

Net Applications purpose is to show the Net Marketshare which aids those who used/develop/report on the state of the internet.

So in terms of their market it doesn't matter about installed based marketshare it matters about the installed base that uses the internet.

 

[clevin]

Net Applications purpose is to show the Net Marketshare which aids those who used/develop/report on the state of the internet.

So in terms of their market it doesn't matter about installed based marketshare it matters about the installed base that uses the internet.

then that should be just called "internet usage share", rather than "market share"
http://en.wikipedia.org/wiki/Market_share, if netapplications reports a global market share of mac as 9%, safari as 9%, what do those data mean for norway, australia, china, Japan, Britain, France? Nothing.

now what do you think is the reason behind the metric change of net applications? if they dont feel previous metric was misused or improper?

 

[windywoo]

Paranoia.