Goldman Sachs: Windows' true market share is just 20% | Hacker News Reader

Goldman Sachs: Windows' true market share is just 20%

theregister.co.uk

The graph is designed to make it appear that the total number of devices is constant, which makes it looks like Microsoft is losing ground. What's actually happened is that the market has grown significantly, considering that their definition of 'market' includes a bunch of platforms that didn't exist a few years ago. That growth didn't happen on MS-based devices, hence the smaller market share.

A less biased visualization would have varying heights of each stack to show the absolute number of devices.

The article specifically talks about the "global computing market" ... So it's referring to market share, not number of devices. Nothing misleading here.

Still, the article says "Microsoft is in danger of becoming a small player in today's global computing market". Which is not backed up by the data provided by the article. Unless we have the raw numbers, it could well be that Microsoft has been growing.

The fact that the total share of MS has been shrinking doesn't necessarily mean that MS is a small player in the global market.

"The fact that the total share of MS has been shrinking doesn't necessarily mean that MS is a small player in the global market."

That depends on the context. If you are trying to decide which OS to develop your software for, market share (with some adjustment for the willingness of users of a certain OS to actually pay money for software) is the critical factor. While Microsoft may still be large in terms of number of installations or dollars, losing on market share is significant. In the past, most software was developed for Windows first, and then OSX or Linux if the developer got around to it, which gave Windows a huge advantage in the market regardless of whether or not it was the best from a technological standpoint. Also, market share allowed Microsoft to essentially dictate "standards" (remember all of those websites that required ActiveX?). Microsoft may still be big and wealthy, but losing on market share really does change things -- it makes them "small" in a sense.

If you're a developer trying to decide which OS to develop your software on, and market share is a significant factor in your decision, you're going to choose something cross-platform. That's most likely web, but even if you choose one of the toolkits your choice is going to impact all of the OS producers equally: their marketshare will not be impacted at all by you, unless they actively inhibit cross-platform development on their OS. MS used to do that but not anymore; Apple tries to keep the wall around their garden but they don't block cross-platform completely.

While this is technically correct, what IS misleading is not incorporating significant changes to the market itself. The "market" in that chart starts off before the mobile revolution, and proceeds to include the mobile market as time progresses without properly depicting that significant and extremely relevant factor.

Actually 2013 and onwards, they see a very solid growth in market share and number of devices for MSFT, and AAPL. They see a significant slow down of the Android train. I wish the analyst writes his name next to Goldman Sachs, because if this holds true he ll be a grand visionary.

Au contraire, I would put around 2016, Windows go down below 10% and 10% for OS/X+iOS. Android would go above 60% and 20% would go to other categories of underdog innovations or rebirths. (Glass computing, ChromeOS, Tizen/Bada/Joll/WebOS, FirefoxOS/Ubuntu/KDE/Linuxes, Card-size computing/Rasperry PI etc, and finally YC-accelerated HW innovations.

Their definition of 'market' being 'all consumer computing devices', so obviously the picture changes rapidly after 2005 with the smartphone and tablet explosion. Its not really news that Microsoft are not doing well if you look at the whole universe of computing devices.

The headline is a bit misleading as it implies something more interesting, along the lines of 'microsoft not as dominant in desktop as people assume' ... but thats not what its about.

I wonder if Goldman's count X-Box as a consumer computing device? I bet they don't, not that it would make that much difference to the stats.

Agreed. It's similar to comparing Intel to ARM. Outside of business the OS is becoming less important to the consumer than it used to be. IMHO the key for Microsoft is maintaining its hold on the corporate market. In that regard the key product is not Windows but Office.

That's right, Office is everywhere. And they use that to leverage Sharepoint, Exchange and other systems. There's no shame in being number 1 in the corporate world and number 4 or 5 in the consumer market. :)

Though they do have the xbox.

There's no shame in being number 1 in the corporate world and number 4 or 5 in the consumer market.

That statement is probably not as easy to swallow for a company that was worth $850bil (in today's dollar) just 13 years ago.

Edit: clarify value accounting for inflation

Well, it is a cycle you know? Microsoft in the 90's looked as impossible to beat down as IBM looked back in the 70's and IBM still exists and it is strong, but is not longer that unbeatable beast that used to be.

I wonder if, in a few years, I will be writing this statement again but comparing Apple/Google with Facebook/Twitter/Whatever-is-coming.

"If left without a meaningful competitor in tablets, we believe Apple's dominant share of tablets will act as an anchor that pulls its smartphone share ... steadily upward over time"

Hm, that's not how anchors typically work.

"The sole aim of a metaphor is to call up a visual image. When these images clash -- as in The Fascist octopus has sung its swan song, the jackboot is thrown into the melting pot -- it can be taken as certain that the writer is not seeing a mental image of the objects he is naming; in other words he is not really thinking." --Orwell, "Politics and the English Language"

The first question to ask about any market-share figure is "Of what market?" Goldman Sachs are defining things using a much broader market than most people would consider useful. Maybe there's something to it, but to call it "true market share" at this point is more than a bit of a reach.

Does anyone understand what's happening in their graph[1] in 2005? I don't understand what would cause Apple to jump from 5-25% in one year, 2 years before the iPhone.

[1] http://regmedia.co.uk/2012/12/13/goldman_platform_chart_lg.p...

The iPod, maybe?

Most probably iPods. They were selling like hot cakes in those years.

They probably changed their methodology but forgot to say it.

I agree. 'Other' category jumped from 0 to 29% as well. Likely at this point they started considering a wider range of devices or something like that.

There's no way they could possibly consider that generation of iPods "computing devices."

Not that I think this is accurate by any stretch, but what would account for the massive jump in "Other" in 2005. Smart phones like Nokia before iPhone + Android?

I wondered the same thing, at first glance I thought that was GOOG's slice of the pie - 4 different shades of blue doesn't make for a good contrast. Then I decided it had to be RIM/Blackberry.

Wow. RIM hadn't even entered my thought process. Funny how quickly things change.

Microsoft needs to get back to the basics. Why must they continue to try and be apple. They are not apple. they are a software company.

Business still use windows for all applications. Excel and Word are the only things keeping them alive. They MUST come up with a new OS that is similar to XP/Vista/7 that is more useable in order for them to stay competitive in the OS market. Imagine when google releases a free OS with free excel/word.

I think Microsoft needs to focus on useability and superior UI. They just keep cluttering things up more and more and making it harder to do simple tasks.

While I do agree that Microsoft probably needs to change their direction a bit (I'm not fan of Windows 8, but then, I'm a Linux user so that speaks for itself), I don't completely agree with your assertion.

First, think about what you said. "Excel and Word are the only things keeping them alive."

That's a generalization. Yes, Office is most certainly their cash cow, and generates tons of revenue from businesses and enterprise in general (to say nothing about higher education and government). But guess what? Office users will (generally) have Windows installed, particularly in larger enterprises. So, using one almost guarantees using the other.

I think it also ignores Microsoft's gaming division (think Xbox et al). So, no, I don't think MS is quite as badly done for as some may think. Focusing on Windows and Windows' success (or failure) exclusively is to ignore the rest of the picture.

I do agree that they may need to reconsider what path they're on (they've been doing some soul-searching to that end) but on the other hand, it's equally important to remember that Microsoft enjoys a lot of momentum in the marketplace, most especially with big business and many governments. Could Windows 8 change that? Maybe. Heck, it might even hurt them down the road. But terms like "vendor lock-in" exist for a reason, and I believe MS has previously done a fantastic job at illustrating why. The death of Windows won't be sudden, and it certainly won't be proclaimed first by Goldman Sachs.

You just agreed with me. A lot of people use windows because of office. This is why people have windows installed, particularly in larger enterprises.

Using one (Excel and word), does guarantee the other (windows). .Doc and .xls are standardized much like .pdf.

Yes, Xbox is a contributing factor, but still...

I like how they imply that the desktop is dead, imply tablets and smartphones are the future cause the new generations buys them and then backs Smart TVs..... which I'm guessing is a generational dead-end.

Does anyone know of any studies that support the article's statement that Apple, when combining OSX and iOS, has more users than Windows now? If you ignore legacy devices still in use I could possibly, maybe, see that. Although, if you count total devices existing in the market, which the graph seems to ignore, I just can't agree with that notion.

Plus, based on the speed of innovation in this market, making predictions up to four years into the future is highly optimistic on their part.

If they are going to compute market share of MSFT by framing it in terms of "market share of consumer compute" and including all sorts of mobile devices, then they should weight each device by the computing power it has accordingly.

Otherwise the number the come up with will underestimate the true market share of "consumer compute" of MSFT.

Or is that what they are aiming for?

Earlier post about same underlying story, take from a different site: http://news.ycombinator.com/item?id=4904151

"if you don't know, now you know"

whatever Goldman says, do the opposite.

Why would anyone trust that statistics released by Goldman Sachs weren't designed to manipulate the market in their favour?

Goldman Sachs: the crooks that funded Spinvox. Absolute frauds and liars.

First, this is counting Android phones many of which are being used as updated feature phones rather than even a browsing device.

Second, this is counting what are being sold the last quarter and ignoring install base and the increasingly longer upgrade cycle of PCs.

For a better picture of install base and actual usage, see http://gs.statcounter.com/#mobile_vs_desktop-ww-monthly-2011...

The entire mobile devices are at ~15%.

Third, this is Goldman Sachs, they might be shorting Microsoft stock now and will buy and upgrade it later just like they bought millions of shares of Nokia when they downgraded it and then the stock more than doubled recently, and GS made tens of millions at the least. Read the very interesting article below showing GS not following their own advice and doing a reverse dump and pump.

http://seekingalpha.com/article/1023231-goldman-sachs-and-mo...

Then we have the Register writing flamebait articles with linkbait headlines, completing the circle of worthless blogspam flamebait junk.

Yeah, there's always an angle with these guys.

Trust what they say publicly as far as you can throw them... And only trust what they say privately if you have a gun to their head.

Is this really what the Goldman brand is worth these days? Not that I disagree with you, but you have to wonder: given that they rely so heavily on their name, this level of distrust must be affecting them. Nobody wants to do business with the mob if you can avoid it.

Who do they rely on their name with? For investors, the image of a company of scheming masterminds who outthink everyone else in the market and make a profit off everything they do is probably actively helpful.

Yes, there are a certain class of people who actually see this shady bullshit in a positive light.

Right up until they get stung too, then they have the nerve to be outraged.

The biggest sucker is the guy who thinks he's in on the con.

Generally, the people who deal with Goldman don't care if Goldman rips off the rubes. As long as they believe they're not being played for a sucker, they see Goldman's otherwise-questionable activities as just part of the game.

The fact that Goldman has such a reputation for playing the people who aren't paying them, plays directly into the very notion that they do have unique insight and talent on offer to those who do pay them.

Haha, it's cute people STILL think that just because they're paying them, they aren't getting shafted too.

Before the mortgage crisis, they were advising clients to buy up subprime debts while in secret shorting them hard, because they knew they were toxic. They bet against their own damned clients... at least actual mobsters have SOME semblance of honour!

If a Goldman banker is patting you on the back it just means he's trying to steal your shirt.

Haven't you kept up with the sub prime crisis? GS rated junk mortgage bonds with the highest rating AAA, and then bet heavily that the same bonds would sink. When they sank, they made windfall profits. This came out in investigations and they were given a slap on the wrist later by the government.

http://www.examiner.com/article/goldman-sachs-sold-sub-prime...

"As a result, Goldman made an approximate $4bn profit from betting on the sub-prime collapse, more than off-setting its own mortgage losses and ensuring that 2007 was a bumper year for the bank while all around are counting their losses."

They basically threw their clients like AIG(which was a massive failure requiring a huge bailout) under the bus for their own profits.

I think you read me wrong. I'm wondering why if they bet against their clients anyone would do business with them at all anymore.

Lack of incentives for long term success? Too big to fail?

A few things : (i) Goldman did not rate any bonds, the rating agencies did. (ii) For every buyer of a security there is a seller. Goldman's trading desk (likely) had a conflict of interest with their bankers. However, just selling something that goes down doesn't make someone evil (it's logically similar to buying a stock that later goes up). (iii) AIG failed due to its own inadequacies : They believed that they were the smartest people in the room, when (in retrospect) they were making a huge blunder.

Oh now, it's like with Bernie Maddof. Many knew or suspected he was crook but thought he was cheating from others so they closed one eye and both ears or whatever, until they got screwed.

Yeah, Goldman Sachs is notorious for doing a ton of unethical stuff. I have no idea why anyone would trust their public statements.

For many people in poor countries, the smartphone is their computer. If nothing else, these people use smartphones to log in to Facebook and use it. That's a computer.

Not just poor countries. It's increasingly common for lower-income individuals in the US use a phone as their only computer.

Phones aren't phones anymore. They're pocketable computers with a phone app.

Poor people will not care about facebook status updates. I have in no way have disrespect for poor people, just saw something wrong with your statement.

Low income individuals also will not care about smartphones. It is not expensive to get a smart phone, I believe most of them come free now as part of 2 year contract. The thing that is expensive is monthly phone + data plan bill.

I traveled across Central and South America for 2 years.

Without exception, every internet cafe I went to (a lot) was overflowing with locals doing one of two things:

1. Chatting on MSN

2. Facebook

Not really different from Germany...

How are you defining 'poor' in this context? Because I'd expect smartphone ownership to be more correlated with age than wealth, and at least in the UK it's quite easy to get a cheap Android phone for £99 on PAYG, or free on a £12.50/month contract. If you're willing to shop around, you can probably get a cheap phone on a £7.50/month contract (including data).

To me a poor is who can't get food and shelter for living.

Expensive in the US. Because we accept the ridiculous pricing and behavior of the telcos and convince ourselves we have it the best it can be. In Europe things are far more competitively priced. I have traveled a lot in the third world and mobile devices are everywhere with smartphones increasingly prevalent. Stopping on a third world convenience store this summer, lost and looking for directions, the scruffy kid behind the counter broke off from playing with apps to show me on the map... On his iPhone.

It's possible that their analysts division is separate from the investments division. Anyone from investment banking care to comment?

Yes, there is always a chinese wall which separates the public and private sides of an investment bank.

The wall you are talking about is between investment banking (private side) and the rest of the bank (public). The two sides don't see each other on the internal IMs systems, if they try emailing each other, their emails get auto routed to compliance, etc. This ensures that private insider information from M&A advising to the companies doesn't leak and get traded on.

In this case Goldman analysts and traders are both sitting on the public side of the wall. Neither analysts or traders know anything private about the companies that can't be found out or calculated without some sleuthing. Additionally under current regulations in the US, the analysts are required to release their research reports to their clients simultaneously( no client advantaged over another ).

I worked at an Investment Bank 15 years ago. They might be separate, but they have lunch together, go to the same holiday parties, stand in the same elevators and all get a bonus based in part on overall bank profits... It's frowned upon to take a physical baseball bat and use it to on analysts because that would leave a mark.

Since they are throwing desktop operating systems in the same cluster as tablets and phones, why did they omit all the different OSs for cars, manufacturing and power plants, heavy machinery, TV sets, satellite boxes, and so on? At least that would be a true "OS market overview". They are cherry-picking a slice of the OS market by the common denominator "consumer products" and lumping a select few totally distinct segments into a single worthless graph.

Not that I would have expected anything more sophisticated from Goldman Sachs, though.

As someone above said, they limited the scope to all consumer computing devices. A car is not a computing device per say, even if it has computer parts. Same goes for power plants, heavy machinery, TV sets and satellite boxes.

One could argue that electronic calculator should be included, but I am not sure it would actually change anything.

Can you define what a computing device is, per say?

per SE

A general general purpose device that can be programmed by the user to perform tasks beyond what was initially designed into the device.

This is my view, and not the only definition that exists. But if we want to avoid having to include every item ever produced in the last 10 years or so, the definition of a computer device needs to be slightly limited. Some Teddy bears nowadays has FPGAs in them. If everything from a stuffed toy bear to a 600 ton airplane can be called "a computer", "a computer" loses its meaning.

If that definition includes iphone, it also includes xbox and the ds lite.

(You could perhaps limit it to devices on which you could run a web browser. I remember there was a version of Opera that ran on the ds, though I didn't get it myself.)

I think your definition is spot on. The reason you can define a computing device that way and exclude 600 ton airplanes and stuffed toy bears from this discussion is that the devices are embedded in the actual item. An airplane can't (shouldn't) be programmed by the user to perform tasks beyond what was initially designed into the device, but the FPGA's in the teddy bear can be. It's not reasonable to look at how many of today's planes are made by Microsoft, but it is reasonable to look at the embedded systems market as a whole and see how influential Microsoft is.

At a minimum, then, that would include graphing calculators.

Also, while most routers aren't exactly intended to be programmed, it's pretty easy. Easier than reflashing an iPhone, which is considered a general-purpose device, even though it's not really user reprogrammable.

Heck, the Canon Hack Development Kit makes a lot of 'dumb' cameras as easily reprogrammable as an iPhone.

And as others have noted, this study apparently excludes gaming consoles, which seems to me an unreasonable omission by any definition. ( I think the most reasonable definition that captures what you're trying to say is "something designed to run third-party applications." )

Agree.

>A general general purpose device that can be programmed by the user to perform tasks beyond what was initial designed into the device.

Does that mean the iPad and iPhone are not computing devices?

Also doesn't your definition encompass routers that you can install DD-WRT/Tomato on which will enable you to run things like Bittorrent on them?

> A general general purpose device that can be programmed by the user to perform tasks beyond what was initially designed into the device.

The user bit is sticky. What about green screen mainframe apps? Those weren't programmed by their users, but they were certainly computers. Chromebooks come to mind too.

If that was the case than the world's #1 OS is probably something you never heard of like VxWorks

Reply on news.ycombinator.com