The down payment on iCloud

Balance Sheets are not typical targets of focus in software or services or “high tech” companies. Except for Cash, there are usually not much in terms of assets for analysts to contemplate. With outsourced manufacturing, even hardware companies don’t maintain depreciable assets on their books.

Apple, as usual, is different. It turns out there is a lot there and there’s a lot to learn from what’s there. I’ve looked at the Cash and Long Term Securities in the past but in the next few posts I’ll dive into fixed (non-current) assets. Specifically Apple’s Property, Plant and Equipment.

Apple Inc. has selected North Carolina as the location for a new data center and will invest more than $1 billion in the project over nine years, Gov. Bev Perdue announced

via It’s Official: Apple iDataCenter to North Carolina » Data Center Knowledge.

When Apple committed to the North Carolina data center in mid-2009 the iPhone was two years old. Only 26 million iPhones had been sold to date (about as many as Apple now sells every quarter.) Clearly if the commitment was on such a scale there must have been a plan. A big plan.

We don’t know for sure if NC was used last week or whether it was handling only part of the load, but let’s assume it was. What I want to think about is how much iCloud would thus cost and hence what would be the cost for alternative providers of such a service. That analysis might let us determine if “cloud” forms a substantial barrier to entry for competitors.

Beside the public statement above (which may have been designed for political benefit) can we find any evidence of what Apple spent on data centers? That’s where the PP&E line comes in.

The following chart shows the Property, Plant and Equipment asset in Apple’s Balance Sheet.


Each line represents a particular type of asset and its value at a particular point in time. I will now pay attention to the blue line, Land and Buildings. I’ll look at the other lines separately.

Note that each line represent current value so it does not say directly what was happening in the quarter. However, if we look at the change q/q for a particular line we get how much new spending occurred in the quarter for that particular asset.

This change in Land and Building asset value is shown in the following chart:

Going back to 2005, there were minimal changes. Perhaps there is some significance in the $250 million spend in early 2006 or the nearly $78 million spent in Q4 2008, but there is a definite significance to the massive spending that occurred right after the NC commitment was made.[1] Some of that payment also included the old HP campus in Cupertino. “Local real estate experts said HP’s asking price may have been $300 million or more, but they also said the final amount may have been less.”

There is also a clear cut-off on that spending at the end of 2010.

The total spending during this time adds up to nearly exactly $1 billion. If we subtract $250 million for the Cupertino expansion we still have $750 million that must have gone into NC.

Note that the original statement quoted above was $1 billion over nine years. We can see that Apple spent $0.75 billion in one year. On the building alone. This is why I consider this a down payment. It does not include any of the equipment in the plant. It does not include the cost of operation (incl. labor) or any development costs (R&D). In fact, it does not include any of the funds that would be part of the nine year commitment.

What this level of spending implies is that iCloud (and Siri and iTunes) are expensive. They may seem ephemeral and even trivial as services, but they require a staggering commitment few can make. Apple made that commitment and they made it early on, before the first quarter billion users were even on the horizon.

If platforms are moving from local to distributed and if value moves from selling things to “getting to know you” and if that knowledge requires infrastructure control then the number of companies that can participate in the market shrinks dramatically. Not only in terms of who has the capabilities, but who could even afford to acquire them.

The invoices Apple pays are shown above.



  1. Apple’s retail stores are not included in these figures as Apple leases them. The cost of the stores will be discussed in a separate post.
  • Several billion of component repurchases will be in those numbers somewhere. Also, retail capex guidance was for $700m in fiscal 2011, up from $400m in 2010. Plus Apple is ramping up chip development and that can’t be cheap either. So, there’s other stuff in PPE beyond just the data centre. Not that it isn’t expensive.

    • Anonymous

      Components for use in end products aren’t cap-ex.

      • Walt French

        I would assume that “chip development” would include machinery, not the chips per se. Might also include some of those deals we heard of, where Apple would partner in the capital cost of a factory for displays, etc.

      • Anonymous

        As I understand it, the fab-less chip designers don’t require machinery beyond computers and fancy software.

        Plant that is used to build components would only count as cap-ex for Apple if the plant remained owned by Apple, that’s far more likely in the case of Foxconn than it is in the case of firms like Wintek or LG. In most cases the component pre-purchases will be appearing as an asset – I believe specifically under the heading Vendor Non Trade Receivable.

  • Alex Esser

    Hmm, good idea, but I would think that spending for the data center would be mainly in the orange line? And didn’t Apple buy land for their new campus in Cupertino? You might have to go a bit deeper in this analysis. I think your conclusion is valid that Apple created a new barrier of entry.


    • Joseph Moran

      Alex, I have to agree that a significant portion of the data center expenses are tied up in that rapidly ascending orange line. The costs of all the server infrastructure alone must be staggering. I’d guess that number will grow exponentially with the number of iPhone owners (particularly iPhone 4S owners making use of Siri, which Apple said in their presentation uses the data center for the voice-to-text transcription).

      Then there’s the cost of the iCloud system (software, data storage, backups, etc.). The storage requirements are probably going to grow even faster than the server needs.

    • The orange line will be discussed separately. This is just a discussion of the cost of buildings and land. I edited the original post to account for the Cupertino campus. Thanks.

      • Anonymous

        Presumably some of that orange line consists of plant machinery that Apple has bought for use by suppliers such as Foxconn? The inflection just seems too early for it to be mostly internal data center resources. There could easily be a billion dollars worth of machinery committed to building iPhone-4s and 4Ss

  • Anonymous

    The moat continues to get wider and deeper. At what point does the competition decide to go elsewhere to look for profitable sales?

    • UpChuck.Liberals NoBO12

      Just after Apple puts Alligators in the moat. I understand that they’ve invested in their own alligator farm in Louisiana. 😉

    • UpChuck.Liberals NoBO12

      Just after Apple puts Alligators in the moat. I understand that they’ve invested in their own alligator farm in Louisiana. 😉

  • ryanpmack

    You don’t think some of this land and building spending in 2009-2010 is also for the design and planning for the new Cupertino campus? Sir Norman Foster doesn’t come cheap. We only first saw the plans in June of this year but surely they’ve been in process for years.

    Also I’m not sure on this, but think they had to buy out other property owners for the new campus outside of HP.

    • I edited the post to account for the purchase of the HP Cupertino campus. I estimated $250 million in Q4 2010.

    • Walt French

      I have it on good authority that Sir Foster is NOT himself a building and so I presume the accountants treat his services as services.

  • Marc Weeber

    What about the reports iCloud leverages both Amazon EC2 and Azure? Is that part of the yellow line, “internal software use”?

    • No. All these are costs related to fixed, depreciable assets. Expenses for Amazon or Azure would be operating expenses, attributed to a particular product.

  • Anonymous

    That analysis might let us determine if “cloud” forms a substantial barrier to entry for competitors.

    It certainly doesn’t form a barrier to the big web firms competing in Mobile. Google, Amazon and even MS have long had these kind of datacentre assets.

    • I was thinking more about the competitors IDC, Gartner and other analysts contemplate. Companies like HP, Dell, Toshiba, Asus, Samsung, Nokia, and HTC.

      • Anonymous

        But they don’t need their own datacenters if they’re using OSes from MS or Google do they? None of them except perhaps Nokia and Samsung seem to be contemplating their own custom online services.

      • kevin

        True they don’t need their own. But without their own, they can’t use this to differentiate their brand vs other Android/WP7 vendors. Which makes it harder to prevent their customer base from switching every two years.

      • Anonymous

        Android users are certainly unlikely to ever develop strong brand loyalties to individual OEMs, any more than Windows users do. But the inability to build such loyalty isn’t a barrier to entry, that’s a barrier to building a barrier perhaps? A meta-barrier?

      • Scott

        The competitor this hurts the most is RIM. Google, Microsoft and Amazon can match iCloud, but RIM probably can’t. It’s also a significant barrier to any potential new platforms, like WebOS.

      • Davel

        Interesting point. HTC and Samsung have brands. Each has an overlay over Android. To keep their customers loyal they need a company specific reason. Google has their own brand which they use, but which makes them replaceable. I wonder what their strategy is other than minor differences in a curved or straight screen, bigger camera etc.

        Google likes them as a commodity. I would think they do not see themselves that way.

      • Walt French

        This is both true and not true. Android is the only high-volume game available to these manufacturers, thanks to the implosion of Palm and Microsoft’s congenital inability to be a disruptive innovator (at least, since their original Windows Mobile foray).

        So yes, they’d like their own brands. Able to really build it? No, they’ll depend on Google or Amazon to provide white-label services that they can re-brand. And when Google tilts too much towards MMI, or has to yank the carpet from underneath them, they’ll do something else.

  • Your numbers don’t need to be exact – you have the message here that Apple has a long term strategy to integrate the Apple Ecosystem into a mammoth IOS delivery machine. The iCloud $ 1BN down payment makes even more sense when you combine this back end investment with another 35 stores Apple has opened in just the last 90 days 400+/- total . The barriers to entry for Apples smart phones competitors are insurmountable. BTW, great write-up on why Apple didn’t release the iPhone 5

  • Pingback: Friday 7atSeven: eroding expectations | Abnormal Returns()

  • Anonymous

    Are you trying to demonstrate that data centers, are expensive?
    Btw. Google spends far more on capex than Apple, despite much lower revenue; presumably most of it on data centers.
    Could part of the spike also be supplier pre-payments? Is it possible those maybe classified as “property plsnt snd equipment”.

    • > Could part of the spike also be supplier pre-payments? Is it possible those maybe classified as “property plsnt snd equipment”.

      Yes, they could be PP&E but they’re not Land and Buildings.

      Data centers are expensive. This data shows how much the land and buildings cost to purchase. It does not cover the cost of the equipment inside.

      • I don’t think Tim Cook has answered this question but I suspect that supplier pre-payments are commercial loans and therefore just another asset in the $76 billion cash portfolio.

      • Anonymous

        They’re an asset but not in the $76BN, that’s just financial assets.

  • Anonymous

    It would be interesting to chart giant new HQ spending vs. turning points in busines fortunes for Silicon Valley companies…

    • Anonymous

      Bingo! I know an architect who worked on many companies “vanity” buildings. He said that he always knew that within 3 years of building their vanity headquarters that they would be in financial ruin.

      • You mean like the Chrysler building in New York. Or how about Pan Am building (now called MetLife building) and the GM towers in Detroit? The Prudential building in Boston? The Sears Building in Chicago? The iconic Oracle HQ in Redwood Shores? How about Toyota City? Or the BMW towers that look like engine cylinders in Stuttgart?

      • Walt French

        Guess your acquaintance didn’t work at Bohlin Cywinski Jackson, who designed the Pixar HQ that opened a decade ago.

        Jobs was said to have had an important hand in the design and functionality of that building, as well; it won a couple of architectural awards; and the subsequent success of the company can hardly be overstated.

  • Rudolf Charel

    My question is simply, can Amazon afford the expense of matching Apple when their margins are below 5% where Apple’s are near to 30%. I think at some time in the not too distant future the market will call Amazon’s bluff.

    See the blue line but don’t forget the yellow line to determine the total expense of Apple’s effort.

    • I’ve written two posts so far covering the blue and the green lines. The yellow line will get a post as well.

  • Pingback: Apple already spent $750 million on the iCloud building alone | Apple World()

  • Matt Martin

    As usual, great post. I think one thing worth considering, however, is the unusually expensive prospect iCloud presented to Apple.

    Most cloud companies (take Dropbox, for example) slowly build their user base as their services become more popular. Thus, they can grow their data centers (which these days can be outsourced as operating costs to Amazon or others) gradually and make adjustments as they go.

    Apple, however, was in a position where it had to take an enormous user base and add on a huge suite of information services after the fact. That’s very ambitious and very expensive.

    While your point still stands that most current competitors (HP, Dell, Samsung, HTC, etc.) don’t have the resources to make this sort of investment (nor, I would argue, the foresight), smaller start-up cloud services have been making significant inroads in the market. Couple these start-up services with widely available resources from current players (i.e. Google), and an Apple competitor can much more cheaply assemble competing offerings from a spectrum of existing services. Of course, this will just further emphasize the existing contrast between Apple’s fully-baked, we control everything approach, and its competitors me too offerings, but it’s not as though the market is only defined by the former.

    That is, I believe your statement that “the number of companies that can participate in the market shrinks dramatically” is only true so much as you define the market as centrally controlled distributed computing platforms. Apple likely is the only company (other than Google) with the resources to pull this off, but as history has shown, centrally controlled isn’t the entire market.

    That all being said, I’ve enjoyed Apple’s approach for 20+ years, so I’m glad that they decided to put a significant investment into this area–it demonstrates that they’re in it for the long haul.

    • Anonymous

      Centrally controlled can also mean more secure. Look at the malwear problem that Android has in comparison to iOS. We’re talking about extremely personal data. If I have my druthers, I’ll trust centrally controlled by Apple every time over a hodge-podge like you describe. But maybe that’s just me….

    • It all depends on the user base. Apple is notorious for its need to control, but when you are serving a quarter of a billion users, you can’t rely on Amazon for your back end. Not because they can’t do the lifting but because if things go down you get the blame and you can’t do anything to maintain your reputation. In other words, you lose control over your brand. This is not an option.

      Can others just outsource to get started? Maybe, but they’ll eventually need to scale and own the “means of production”. But they won’t have the competencies and lessons learned. I’m sure Apple had this debate four years ago and said they need to go this way, and that there was no alternative.

      • Hossein

        Will Apple eventually need to own its “means of” iCloud? I am not sure. So far the company has decided to rely on Microsoft’s Content Delivery Network (which is part of Windows Azure’s storage piece).

      • Walt French

        I know I saw data suggesting as much, but then counter-evidence disputed it. Do we really know about Apple’s technology for iCloud?


      • Hossein

        Neither Microsoft nor Apple will comment on this. So there is no hope of getting an official statement. All we can rely on is the messages that are flying in the wild. They certainly contain Azure signatures.

    • zato

      “That all being said, I’ve enjoyed Apple’s approach for 20+ years, so I’m glad that they decided to put a significant investment into this area–it demonstrates that they’re in it for the long haul.”

      I’m guessing the REAL investment that is difficult for competitors is SIRI. To port, localize, update and test SIRI for every country, city, town, where iPhones are in use is probably the biggest task Apple or any other tech company has ever undertaken. I imagine that it would take a very large number of people to do this.

      I think Apple is intentionally “raising the bar” with SIRI to a height that could be impossible for anyone else to match with an equivalent offering. How would an equivalent Android service work? Would Google absorb the month to month costs of operating the service, and freely give the data to all manufacturing partners?

  • poke

    A company with fewer users could probably rely on 3rd party cloud hosting (such as Amazon Web Services) in the initial part of their growth. This is an important enabler for start-ups at the moment, since they can run much leaner. So I’m not sure this is a barrier to a potential disruptor. However, larger companies with existing user bases would likely need to make similarly large upfront investments or partner with someone who owns existing infrastructure (Nokia could presumably turn to Microsoft for cloud services).

  • Pingback: Apple already spent $750 million on the iCloud building alone? | 3dot5 -iPhone and iPad Apps and everything Apple()

  • Anonymous

    Great post Horace.

    If 750 million was spent solely on the NC land and building, what do you think is special about the building(s) there? The land could only account for a small portion of the total.

    Even if 500 million went into the building(s), what kind of building is it if it cost half a billion?!?

    Something is missing here. Either there are some other Land & Building expenditures that we aren’t seeing or those NC buildings are much more than just data centers.

    Any ideas?

    • Anonymous

      The question occurred to me as well.
      Some possible answers (just the ones that popped into my head without much thought):

      Infrastructure build out as part of the agreement with NC. Power, sewer, and bandwidth all needed to be delivered to the site, probably at Apple’s expense.

      The cost of the server farm might have been quite high due to the need for on site generation and HVAC needs.

      Some of the money may have gone into buying land (Brazil?) as part of their partnership with Foxxcon.

    • John

      Maybe, like some of those science fiction movies there are many levels under the visible building plunging deep into the earth. 😉

    • Anonymous

      I know someone who bought 120 acres of farmland along I95 in NC, about 6 years ago, for $6000 an acre, so no, farmland in Western NC can’t be that expensive.

  • Sam M.

    What’s amazing is they spend this much money for datacenters and many users had glitches upgrading their devices to iOS 5 or using iCloud on the first day. People are still having issues.

    • There are a few things between the users and the data center that may also have played a role.

    • Anonymous

      Define “many.” 3 million users with problems would be like 1% of users. That would not necessarily be many. Have they lost all their contacts and calendars like Microsoft’s cloud users? Do they have bricked devices like the first Windows Phone update? Are they unable to access the latest software for the next 18 months like Android users? Did they lose the AppleCare number or their route to the Apple Store where an Apple employee will fix their problems for free?

      • Sam M.

        If you actually read around you’ll find that the users who are having problems, many have had bricked phones, and / or have lost their contact & calendar info. There are even reports of people still not having iCloud mail working properly.

        As far MS Windows Phone is concerned, the upgrade process from WP 7 to WP 7.5 Mango was flawless. Nevermind the fact that WP Mango is first-class mobile OS

        Look, Apple makes great products but they’re not infallible. It’s common knowledge they’re behind the big boys when it comes to consumder cloud services. Does that mean they can never catch up. Of course they can. But they need to bring their A game right away if they want to be taken seriously in the cloud services game

      • Anonymous

        As JohnDoey said define your ‘many’ and just for your info there are 250 million iOS devices out there and more are being added on everyday.

  • Mike

    It might be a moat for smaller competitors but it does seem that against it’s larger competitors, e.g. google and amazon this investment is more of a catch up investment.

    • Anonymous

      You’re saying that Google and Amazon are ahead of Apple in connected devices? They are not. There are very few Kindles, there are fewer Android devices than iOS, and only some of those Android are Google-blessed.

      We’re not talking about who has the most servers, period. We’re talking about seamless cloud for a connected devices platform. Apple not only has way more clients, they have clients across all device categories, more 3rd party apps and better 3rd party apps, AND all the Google and Amazon stuff also runs on Apple devices and not vice versa. You can say iBooks is a catch-up to Kindle — true — but Apple sells iPads to Kindle users. In other words, even though iBooks doesn’t lead in books, iOS does.

  • Anonymous

    Ahem, basically all Apple’s rivals can afford this or already have similar or greater capacity. Do you really think Google or Microsoft don’t have datacenters or that they are worth less? They have more, and not just in the US.

    Also, if you think that this is a huge commitment, think about Microsoft’s plan to integrate SkyDrive into Windows.

    • As mentioned in an earlier comment, I was thinking about the competitors IDC, Gartner and other analysts contemplate. Companies like HP, Dell, Toshiba, Asus, Samsung, Nokia, Sony and HTC. There are also emerging threats like ZTE and Huawei.

      If you consider Google and Microsoft as the real “rivals” then Apple already has achieved what it needs to achieve with this investment. I don’t think it has yet.

      • Anonymous

        Ah, for the likes of HTC it’s obviously too much, I fully agree (and it’s just the beginning, these investments will be ramping up).

        I do think that abandoning its own platform was a necessary decision by Nokia for a number of reasons, this one included.

      • Walt French

        HTC, ZTE, Huawei et al. are only competitors because of Google. There’s no way that 3/4 of these companies could create a competitive OS to Apple’s, and the results from HP and Nokia, who actually did, are not exactly great.

        In the Cold War era, the USSR and USA had proxy armies actually doing the shooting at one another. I take all the Android OEMs as proxies for Google. They have very small objectives that align well with Google’s overarching ones. But the success or failure of any one doesn’t dramatically impact Apple’s competitive position. Likewise all the barely-profitable Windows box manufacturers that are not able to afford competitive designs except at the very lowest cost end.

        So with this iCloud/Siri stuff. Sony’s music store, the closest competition from this list, shows by its irrelevance how this group is not Apple’s competition. Google will keep bolstering its ranks with others, or enhancing the powers of some (MMI) as needed to distribute its services.

        So I respectfully disagree with your last para. The Greater Google Complex, Microsoft’s Windows ecosystem and Amazon (hooda thunk it a year ago!) are Apple’s competitors today. I think that Microsoft has tragically failed to recognize the threat of today’s smartphones, and is thinking in terms of protecting its franchise rather than cannibalizing it before others do; this reduces them to a dark horse. Google, in moving at light-speed, has left its IP flanks exposed and Apple may be able to slow them down a la Samsung. But the real competition will be fought over building a robust enhancement of internet services integrated into multiple user devices, among just these 4 competitors. Any time there is robust competition, all players have to race hard: Google on integration; Amazon on breadth of services; Apple on useful, value-for-the-buck cloud.

  • Hossein

    Apple is far behind in this game. I would say their investment was late. Their reliance on Windows Azure is a testament. I predict consumers will suffer from interrupted iCloud service. I hope you will do a good financial analysis of the damage it’s damage to Apple’s brand.

    • Can you provide any verifiable evidence that Apple is using Windows Azure? Can you point to anyone who is ahead of Apple “in this game”? (and if you say android has had voice control already, you lose credibility completely.)

      • Hossein

        If “this game” is iCloud services, then there are plenty of examples:

        – Facebook, Yahoo, and Google handle millions of pictures and stream them to many devices with a single login every day (what Apple calls photo stream)
        – Google and Microsoft have been offering online documents (chances are the total number of documents that are handled this way are many more than total documents generated and consumed on iOS and Mac OS)
        – Yahoo and Google have been offering cloud supported email and calendar services for a long time
        – Amazon and Google offer more capacity in their cloud to store music

        Many developers have reported Azure signature in iCloud communications:

        PUT HTTP/1.1

        HTTP/1.1 201 Created
        Content-MD5: [redacted]
        Last-Modified: [redacted]
        ETag: [redacted]
        Server: Windows-Azure-Blob/1.0 Microsoft-HTTPAPI/2.0
        x-ms-request-id: [redacted]
        x-ms-version: 2009-09-19
        Date: [redacted]
        Connection: close
        Content-Length: 0

      • Anonymous


        AGAINST what you are saying:
        The POINT of iCloud is not to replicate what others have down. iCloud is not Dropbox done by Apple, it is not calendar and mail, it is not Google Docs. The essence of iCloud is sync and replication of data across multiple devices.
        If you want to compare it to predecessors, the predecessors of things like what Ray Ozzie has spent his life working on — things like Groove. But iCloud is trying to do this sort of sync for very different devices, and in a way that is available to third parties. It IS something new under the sun.

        If you want to criticize Apple, the ridiculous thing is that there ARE plenty of things to criticize — and the fact that you have chosen to criticize chimera and phantoms rather than real problems makes one wonder…
        What could be criticized? Oh god, so much.

        – the fact that iOS5 caused so much server meltdown in unforgivable. Apple has done plenty of these now, and should have EVERYTHING in place to handle these huge load spikes.

        – the chaos was made so much worse because the upgrade UI was so atrocious — you enter into this convoluted many many stage process, with no indication at each point of how many stages are left and how long the process will take. There are many points along the way where your mac simply seems to be nothing for thirty minutes while it puts up some essentially meaningless message about “preparing something or other”.
        It is quite clear that iTunes has hit the absolute limits of its current software architecture, unable to give decent UI feedback in these sorts of multistage processes, unable to ensure that ALL its ducks are lined up before starting the non-reversible parts of an upgrade, unable to make decent use of multicore CPUs

        – the bug testing for the upgrade process was clearly unacceptably flawed. One obvious problem, for example, was the upgrade’s completely broken handling of iOS devices that make use of iTunes’ “convert audio to 128kbps” option. Another problem seems to be broken handling of album artwork, and some audio files being copied over are corrupt. Meanwhile iTunes continues to be perhaps the buggiest Apple app, and widely hated on Windows.
        Generally we see with iTunes a 1990s level of software architecture and design — a bunch of concepts invented by amateurs, and glued together with spaghetti code — it is SIMPLY NOT GOOD ENOUGH.
        Apple needs to accept the cost, throw away the existing product and start again — come up with an adult software architecture, then implement it TWICE — once as a proper Windows app, and once as a proper Mac app, rather than the current grotesque hybrid that tries to share code between two OSs and ends up doing EVERYTHING badly.

        – the thinking around iCloud appears to have been more than a little brain-damaged. The idea of large-scale sync is good, yes, but there appears to have been NO thought given to iOS apps that want to maintain a large data store that is NOT synched. Marco Arment has already discussed this in connection with Instapaper, but it’s going to affect many many serious applications — eg any reader/PDF viewer application, any alternative to the Apple Music or Video applications etc.

        My point is
        – this transition has been very rough on Apple. Heads deserve to roll for the screwup that have been made so far; and I suspect that Apple have less than a year to fix up the most obvious problems before they start being treated in the press the same way as RIM, as a tech has-been that can’t play in the big leagues.
        – BUT I see no evidence that any of these problems have anything to do with what you (Hossein) are complaining about. The real problems are the results of
        (a) unwillingness to accept that iTunes has degraded to a POS that is ruining the Apple is experience and
        (b) poor capacity planning for huge demand surges
        (c) the (unfortunately standard) Apple belief that they know and understand exactly what all customers and developers need, and so don’t need to consult with them on possible holes and flaws in their APIs and vision.

      • Hossein

        Implementing a generic sync service is not feasible (and more important) not useful. Even Apple will sell their generic sync service (a.k.a. iCloud) through specific use cases (a.k.a. applications). The reason why Google and Microsoft do not offer generic sync service is not that they never thought of it. It is because they are not ready for it (although they have been in the cloud game much longer than Apple).

        On other points, I cannot agree more with you; I have repeated several times in comments on this blog and elsewhere that Apple’s engineering decisions are often short sighted. They seem to have great aesthetic design sense, but poor software engineering judgment. This will haunt them soon.

        This last move (iCloud) was far beyond Apple’s resources and expertise. WheniCloud launched I had two predictions:
        1. There will be numerous/frequent back-end problems
        2. Days of “good enough” battery lifetime of iPhone are gone

        The first one we is already happening (even though iCloud has not fully rolled out yet). The second, we will have to wait and ask iPhone 4S users who have serious workload (to be synced to their iPhones).

      • Anonymous

        Now no wonder one of the wonder boys at google was telling developers that they write rotten softwares that suck battery life and not because of the through the air stuff all android products have to go through.

      • Anonymous

        “(and if you say android has had voice control already, you lose credibility completely.)”

        The major feature of Siri and all other voice recognition is dictation for messages, search boxes, essentially “voice input.” The rest is nice window dressing. Sleight of hand to make everyone ignore that the core features have been lifted wholesale.

        Who is going to talk to their phone and ask it “What time is it in London?” Some small minority of the already very small group that is currently touching the clock widget and looking at the clock that says “London” next to it. Or already looking at their weather widget on their home screen rather than asking Siri what the weather will be like. Siri’s most compelling feature is finding a cab for you if you tell it “I’m drunk.” Maybe you can get it to play beer pong with you too…

        It will change nothing.

        It’s a novelty. And again, it’s brilliant marketing to make everyone look at these lofty ideas rather than the meat and potatoes, the voice input core feature, which was slavishly copied. Unfortunately, Samsung copies hardware and Apple just copies software. The notifications system they added is exactly like Android as well.

  • Brock

    What about the Foxconn factory in Brazil that Apple purportedly had a large hand in developing? And the various reports that Apple signs agreements with Foxconn & other manufacturing partners to build and outfit factories with expensive assembly lines in exchange for time-limited exclusivity? Wouldn’t both of these potentially be accounted under L&B and thus throw off the $750m number?

    • It’s unlikely Apple owns the land or buildings that are used by Foxconn. The equipment in the buildings would not be under L&B. Perhaps Machinery and Equipment.

  • I see this long term strategic planning in other things as well. Was it in 2006 that Apple first put accelerometers into the MBPs to park the disk’s head (hopefully) before it hit the ground? Were they thinking way back then: “Gee, I wonder what we could do with these puppies. Let’s check them out.”?

    To me, Snow Leopard — not a lot of new features, but increased efficiency and significant decrease in size — was a preparation for iOS.

    This long term strategic planning is one of the things that sets Apple apart from most companies.

  • Walt French

    “Not only in terms of who has the capabilities, but who could even afford to acquire them.”

    Horace, I can’t find the right value for iCloud data storage costs. Swagging 10GB per 200 million users (remembering that many users have multiple devices), I get 200K Terabyte drives — a mere $60 million based on my laptop’s terabyte disk, or $2 billion per year at the fully loaded charge-back rate of $10K/Terabyte/year that I hear in EnterpriseLand.

    If I split the difference to $400mm/year, and recognize the points above that iCloud competitors start up on shoestrings, this is not a moat. Likewise, iTunes, Netflix and others actually don’t need hugely expensive storage as much as contracts, software and connectivity.

    OTOH, Siri will have 20 million requests per day by the end of this weekend (4MM phones X 5 “Siri, open the pod bay door” show-offs). My own take is that Siri’s IP and software is itself the moat, and the billions of hardware merely a fraction of the chicken-and-egg problem for would-be competition.

    The latest Gartner (?) PC share data suggest that Apple is capturing something like 30% or more (!) of worldwide PC/smartphone/tablet REVENUES, the rest being split across uncoordinated firms. So there simply CAN’T be competition from them, as non-Apple hardware has been almost completely commoditized. Consumer competition will come from Amazon, Google and Microsoft services. (You wouldn’t expect Apple to fight giants like Oracle, SalesForce, SAP and the like, preferring to sell consumer-type hardware into the Enterprise.)

    So the real race is which service providers can strike deals to intermediate various providers of data. Google guaranteed that its IATA (flight) data would be open, as part of its DoJ agreement, so methinks there are a few dozen Apple engineers preparing to answer, “Siri, what would it cost me to fly to Chicago next Monday, returning Wednesday?” as well as Facebook wondering whether they want to work with, or against, Apple’s being able to answer, “any news from my high school friends?” Open Table seems like a lock; I wonder why it’s not ready to go, especially as OT would immediately understand the risk of Google extending Zagat into reservations. If *I* were Mr. Ballmer, I’d be looking at how Microsoft could be the behind-the-scenes search engine for Siri in several knowledge domains, knowing that unless Apple offered to license Siri, the move would guarantee that Microsoft phones NEVER gain double-digit shares.

    I’m going to *speculate* that Siri becomes a dominant way of interacting with iPhones in about the same time frame that Apple’s patent crusade eviscerates Android’s touch interface, with the obvious rejoinder that “Touch obviously isn’t necessary — you could’ve bought/licensed/invented Siri.”

    • Much of what you say sounds plausible though still speculative. This is however what needs to be watched very closely. The data only points to there being much more going on behind the scenes.

      • Walt French

        “…plausible though still speculative…”

        Yup, that’s my stock in trade. I put it as a complement to your very fine, more specific and fact-based work, in that I think it’ll open up ideas as to what Siri means in the bigger picture.

      • Anonymous

        I’m not sure Ballmer is smart enough to play in this game, but I figure Kleiner Perkins is trying to get as many of their portfolio companies as possible involved in supplying information to Siri.

    • Anonymous

      I doubt Apple will waste any time on Bing integration. Microsoft is a competitor with Apple’s core products, and Bing loses so much money it could be shut down any time. Not to mention that Apple is drinking their milkshake right now, who even knows how much longer Microsoft exists in its current form. It’s a little like investing in Adobe Flash. Also, Microsoft is not the best at anything.

      I think they will add 1000 more Yelps and Wolfram Alphas, so they can lose a partner at any time and sub in another and not notice it.

      • Walt French

        And I doubt Microsoft will do it, either. I certainly wouldn’t tell Apple to invest much in it, but if MS would open up some programmatic interface, rather than just accept a handed-off search, some fun could be had!

        Mostly, the Microsoft idea was a bit of a gratuitous reference to Horace’s earlier observations about Android putting Google in competition with its customers: I think that Microsoft is trying to leverage Windows and Office to get into the mobile space, while what they WILL do is to hasten their irrelevance by not providing software to Android and iOS. I think MS ought to get out of the phone business: a money pit that is also speeding the demise of their desktop franchise.

        I don’t think that Apple can handle 1000 Yelps right now. Unlike Google’s genius algorithmic relationship engine, Siri is built upon hand-engineered knowledge bases, language relationships, “ontology” — an approach that Google consciously avoided as too limiting.

        And there ain’t a lot of Wolframs.

        So yes, 1000 Yelps describes a great end-state for them, but I would imagine they’re looking for a more manageable number, perhaps dozens, that they can plug in. I imagine there are a bunch of Amazon wannabes knocking on Apple’s door, but that Apple will do its own general purpose store. Besides that, obvious suspects such as Open Table, etc.

    • BL

      Basing your “analysis” on your laptop hard drive is completely bogus. Enterprise storage is a very different animal than local consumer-grade disk drives, and considerably more expensive. You don’t just have the cost of a drive (which is much higher for enterprise quality disk), you have controllers, cache, high-speed network components (fiber channel or 10GbE, generally), redundant infrastructure, power, etc.

      • Walt French

        I welcome more realistic data; I thought using the word “swag” would convey that I was making a very thinly supported, “Scientific Wild-Assed Guess.”

        So how much storage and traffic is NC built for? How much supports which functions?

        “Inquiring Minds Want to Know!” ®

      • BL

        Fair enough, it was stated as s SWAG.

        I would love to provide more information, if only I were in the inner-circle at Apple! 🙂

        Like you, I’d love more data though!

      • Anonymous

        On the other end, you are making the mistake that large scale computing relies on “enterprise” grade hardware and architectures.

      • BL

        Actually, it’s not true that I make an assumption. I earn a living as an enterprise Storage Architect, so I’m quite familiar with not only the architecture of modern storage arrays, but also the mix of hardware that’s been deployed into Apple’s datacenter(s). That doesn’t mean that I know the amounts or ratios of specific platforms they have deployed – I don’t.

        The fact is that most, if not all storage products used in large scale IT infrastructures today use commodity components. But it does not follow that those are identical to, or at the same price point as, consumer-grade devices. The point I was making is that it’s not a good expense comparison – this stuff is far more expensive than laptop hard drives. Believe me, I wish it wasn’t true – I spend $Millions on storage every year.

    • Hossein

      The last paragraph could not be more to the point.

      • Captainwhizz

        Agreed. But what happens to Google? As Steve Jobs said people now used specialised Apps for search even more – Google doesn’t know who you are, what sex or age you are, where you are or why you’re asking this question in the first place. Siri not only blows this away it also locks out Google – pay back time for stealing code from Sun, Unix, Microsoft and Apple.

        People old enough to remember the birth of the Internet will remember AltaVista and Yahoo and that Google just invented a better mouse trap. However the list is the same – an ill-informed guess as to what you need. We are entering ya whole new paradigm here that only science fiction writers dreamt about

        Google will rapidly join Nokia, Rimm, Dell, HP and Microsoft as the 21st centuries wall flowers.

        An exciting and different future awaits us all as finally computers work for us not vice versa.

      • Hossein

        I wish it were true and the few guys at Siri that Apple acquired had such great AI technology. But it is not true (yet). We will have to wait a year and find how often iPhone users actually use Siri in the long term. I hope it will be a great product so that all others will do the same and all consumers will benefit.

        On the search vs. app thing: Steve Jobs is (was) a marketing dude. He also claimed Apple has the best hand writing recognition algorithm on earth. He even swore in a court that he was “sterile and infertile.”

        Find one objective research study (the kind done by Nielsen or others) that supports his claim. That was most likely his wishful thinking or to send warning messages to Google. If you are interested in solid facts about iPhone app vs. web usage here is an academic tech report:

        “Tales of 34 iPhone Users: How they change and why they are different” by Ahmad Rahmati, Clayton Shepard, Chad Tossell, Mian Dong, Zhen Wang, Lin Zhong, Philip Kortum from Rice University.

      • Anonymous

        Siri will be important to Apple and I can safely say the next iPhone iteration will do away with google map and Siri will do whatever search we need.

        Don’t tell me you have never lied, oops I forgot certain people will do no evil.

    • poke

      It’d be interesting to look into Siri’s IP position. Siri was spun-off from a DARPA project at SRI International that started in 2003. If the relevant IP was transferred to Siri then Apple’s patents here could be very good.

  • deV

    iCloud’s current…issues…are all over the web. Search Google News or…here:

    Biggest issue seems to be upgrading to iCloud makes your email stop working. Not sure how they accomplished that feat, but seems to have been mentioned from several news sites. Other problems include not accepting correct passwords, not syncing properly, not backing up devices, site unavailability…

    Oh and that small issue that any er…indecent…pictures that end up in your PhotoStream are stuck there until you disable syncing on all your devices. Bizarre that you can’t just…delete. I’m sure it will all “just work” someday…

  • now maybe those investors that keep insisting on a dividend will begin to understand how important have a large cache of money available to create what is needed to maintain distance from its competitors.

    • In a general sense yes, but does Apple really need $80 billion in cash sitting around to keep investor confidence? I don’t think so.

      • Anonymous

        You reckon they should pay a dividend to keep investor confidence.

        History has proven that these people who hold large quantities love to drive down Apple shares and then buy them back again. They don’t care about confidence only how much they can milk each time they short the shares.

        You can always sell your Apple shares if you don’t gave any confidence in Apple provided you have them.

    • $390AShareIsTrulyExciting!}:-D

      I don’t think these so-called investors care that much about long-term or they’re worried about Apple’s future. They want their pocket money guaranteed every few months. I’m an individual Apple investor since 2004 and I’ve made plenty of money. I want Apple to keep that cash to make some huge acquisition or buying components up front in quantity at reduced cost to grow their economies of scale beyond what most any other tech company can afford to do. If I need pocket money, I’ll just sell some shares. If the stock keeps climbing or splits, that’s good enough.

      I like to know exactly who these investors are that are constantly crying for dividends or buybacks. 10 to 1 they’re recent institution or hedge fund managers. What do you most of you individual owners think? My only concern is that if Apple doesn’t run the company the way these “investors” want, they’re going to continue trying to hold down Apple’s share price. I’m guessing that’s the only reason why Apple’s share price is lagging behind target prices the way it is.

  • Anonymous

    Horace, for a good rundown on “cloud” economics and technology, you should have a look at this paper: It is a couple of years old, but the analysis is still useful.

  • Anonymous

    We are only blowing hot air here or trying o show how smart we are, we side with our heroes and condemn the ones we don’t like.

    Well talk is cheap we have nothing to lose except time and not money which the many enterprises are betting on that they are backing the right technology.

    • chandra2


    • Walt French

      Well, today I’m at a research conference where, yet again, a researcher highlights how investors under-appreciate how statistical info matters. Versus over-score simple, 3rd grade stories.

      So kudos to Horace for doing the heavy lifting and a hearty FU to somebody who thinks he knows it all despite what the data show.

      Have a nice day!

  • Just a note: Change in assets does not reflect the full picture: change in assets + depreciation = investments. I am assuming depreciation is very small on land and buildings (20 years depreciation time), so change in asset base is a good proxy.

    • Exactly. The trouble is that Apple does not break down depreciation by asset and only offers accumulated depreciation for all assets. I should have noted that these are approximate investments. The more important take-away is that these investments correlate to particular strategies which are what I’m trying to focus on.

  • Pingback: How much do Apple’s factories cost? | asymco()

  • Pingback: THE APPLE INVESTOR: What To Expect Tomorrow From Apple As Analysts Raise Estimates (AAPL) « Tax Australia()

  • Pingback: THE APPLE INVESTOR: What To Expect Tomorrow From Apple As Analysts Raise Estimates | | LATEST TECHNOLOGYLATEST TECHNOLOGY()

  • Pingback: The welcome shift to context-based e-commerce()

  • Pingback: The tipping hand of production: How Apple foreshadows iOS volumes | asymco()

  • Pingback: THE APPLE INVESTOR: What To Expect Tomorrow From Apple As Analysts Raise Estimates (AAPL) | Cheapest Motorola Xoom()