Web 2.0 and Cloud Computing

A couple of months ago, Hugh Macleod created a bit of buzz with his blog post The Cloud’s Best Kept Secret. Hugh’s argument: that cloud computing will lead to a huge monopoly. Of course, a couple of weeks ago, Larry Ellison made the opposite point, arguing that salesforce.com is “barely profitable”, and that no one will make much money in cloud computing.

In this post, I’m going to explain why Ellison is right, and yet, for the strategic future of Oracle, he is dangerously wrong.

First, let’s take a look at Hugh Macleod’s argument:

…nobody seems to be talking about Power Laws. Nobody’s saying that one day a single company may possibly emerge to dominate The Cloud, the way Google came to dominate Search, the way Microsoft came to dominate Software.

Monopoly issues aside, could you imagine such a company? We wouldn’t be talking about a multi-billion dollar business like today’s Microsoft or Google. We’re talking about something that could feasibly dwarf them. We’re potentially talking about a multi-trillion dollar company. Possibly the largest company to have ever existed.

I imagine many of my friends who work for the aforementioned companies know all about this, and know how VAST the stakes are.

Windows vs Apple? Who cares? Kid’s stuff. There’s a much bigger game going on… And for some reason, its utter enormity seems to be a very well-kept secret, at least to non-combatants like myself.

The problem with this analysis is that it doesn’t take into account what causes power laws in online activity. Understanding the dynamics of increasing returns on the web is the essence of what I called Web 2.0. Ultimately, on the network, applications win if they get better the more people use them. As I <a href=http://www.oreillynet.com/go/web2pointed out back in 2005, Google, Amazon, ebay, craigslist, wikipedia, and all other other Web 2.0 superstar applications have this in common.

Cloud computing, at least in the sense that Hugh seems to be using the term, as a synonym for the infrastructure level of the cloud as best exemplified by Amazon S3 and EC2, doesn’t have this kind of dynamic. (More on different types of cloud computing later.)

Of course, it is true that the bigger players will have economies of scale in the cost of equipment, and especially in the cost of power, that are not available to smaller players. But there are quite a few big players — Google, Microsoft, Amazon — to name a few, that are already at that scale, with or without a cloud computing play. What’s more, economies of scale are not the same as increasing returns from user network effects. They may be characteristic of a commoditizing marketplace that does not actually give outsize economic leverage to the winners.

I can’t vouch for the authenticity of the following remark, since I heard it secondhand, but it was from a thoughtful, informed source: Jeff Bezos is reported to have said that he welcomes cloud competition from Google and Microsoft, because they’ll subsidize their cloud services with profits from other part of their business, while Amazon will always have to make it pay. “We’re good at commodity businesses,” Jeff is reported to have said, and the facts bear him out.

If cloud computing is a commodity business, then the outsize profits that Hugh envisioned are not going to be there. This is a business that will be huge, but it may be more similar to the web hosting and ISP markets, which are also huge, but not hugely profitable. (See Rackspace’s numbers for a taste.)

But because one of my goals at Radar is to help people think about the future, I wanted to spend some time on the possible futures and strategies that could turn cloud computing into the kind of massive monopoly that Hugh envisioned.

Types of Cloud Computing

Since “cloud” seems to mean a lot of different things, let me start with some definitions of what I see as three very distinct types of cloud computing:

  1. Utility computing. Amazon’s success in providing virtual machine instances, storage, and computation at pay-as-you-go utility pricing was the breakthrough in this category, and now everyone wants to play. Developers, not end-users, are the target of this kind of cloud computing.

    This is the layer at which I don’t presently see any strong network effect benefits (yet). Other than a rise in Amazon’s commitment to the business, neither early adopter Smugmug nor any of its users get any benefit from the fact that thousands of other application developers have their work now hosted on AWS. If anything, they may be competing for the same resources.

    That being said, to the extent that developers become committed to the platform, there is the possibility of the kind of developer ecosystem advantages that once accrued to Microsoft. More developers have the skills to build AWS applications, so more talent is available. But take note: Microsoft took charge of this developer ecosystem by building tools that both created a revenue stream for Microsoft and made developers more reliant on them. In addition, they built a deep — very deep — well of complex APIs that bound developers ever-tighter to their platform.

    So far, most of the tools and higher level APIs for AWS are being developed by third-parties. In the offerings of companies like Heroku, Rightscale, and EngineYard (not based on AWS, but on their own hosting platform, while sharing the RoR approach to managing cloud infrastructure), we see the beginnings of one significant toolchain. And you can already see that many of these companies are building into their promise the idea of independence from any cloud infrastructure vendor.

    In short, if Amazon intends to gain lock-in and true competitive advantage (other than the aforementioned advantage of being the low-cost provider), expect to see them roll out their own more advanced APIs and developer tools, or acquire promising startups building such tools. Alternatively, if current trends continue, I expect to see Amazon as a kind of foundation for a Linux-like aggregation of applications, tools and services not controlled by Amazon, rather than for a Microsoft Windows-like API and tools play. There will be many providers of commodity infrastructure, and a constellation of competing, but largely compatible, tools vendors. Given the momentum towards open source and cloud computing, this is a likely future.

  2. Platform as a Service. One step up from pure utility computing are platforms like Google AppEngine and Salesforce’s force.com, which hide machine instances behind higher-level APIs. Porting an application from one of these platforms to another is more like porting from Mac to Windows than from one Linux distribution to another.

    The key question at this level remains: are there advantages to developers in one of these platforms from other developers being on the same platform? force.com seems to me to have some ecosystem benefits, which means that the more developers are there, the better it is for both Salesforce and other application developers. I don’t see that with AppEngine. What’s more, many of the applications being deployed there seem trivial compared to the substantial applications being deployed on the Amazon and force.com platforms. One question is whether that’s because developers are afraid of Google, or because the APIs that Google has provided don’t give enough control and ownership for serious applications. I’d love your thoughts on this subject.

  3. Cloud-based end-user applications. Any web application is a cloud application in the sense that it resides in the cloud. Google, Amazon, Facebook, twitter, flickr, and virtually every other Web 2.0 application is a cloud application in this sense. However, it seems to me that people use the term “cloud” more specifically in describing web applications that were formerly delivered locally on a PC, like spreadsheets, word processing, databases, and even email. Thus even though they may reside on the same server farm, people tend to think of gmail or Google docs and spreadsheets as “cloud applications” in a way that they don’t think of Google search or Google maps.

    This common usage points up a meaningful difference: people tend to think differently about cloud applications when they host individual user data. The prospect of “my” data disappearing or being unavailable is far more alarming than, for example, the disappearance of a service that merely hosts an aggregated view of data that is available elsewhere (say Yahoo! search or Microsoft live maps.) And that, of course, points us squarely back into the center of the Web 2.0 proposition: that users add value to the application by their use of it. Take that away, and you’re a step back in the direction of commodity computing.

    Ideally, the user’s data becomes more valuable because it is in the same space as other users’ data. This is why a listing on craigslist or ebay is more powerful than a listing on an individual blog, why a listing on amazon is more powerful than a listing on Joe’s bookstore, why a listing on the first results page of Google’s search engine, or an ad placed into the Google ad auction, is more valuable than similar placement on Microsoft or Yahoo!. This is also why every social network is competing to build its own social graph rather than relying on a shared social graph utility.

    This top level of cloud computing definitely has network effects. If I had to place a bet, it would be that the application-level developer ecosystems eventually work their way back down the stack towards the infrastructure level, and the two meet in the middle. In fact, you can argue that that’s what force.com has already done, and thus represents the shape of things. It’s a platform I have a strong feeling I (and anyone else interested in the evolution of the cloud platform) ought to be paying more attention to.

The Law of Conservation of Attractive Profits

A lot of my thinking about web 2.0 grew directly out of my thinking about open source. My argument in The Open Source Paradigm Shift was that what we learned from the history of the IBM personal computer — a commodity platform built from off-the-shelf parts — was that it drained value out of the hardware ecosystem, turning it into a low-margin business. But profits didn’t go away. Instead, through something that Clayton Christensen calls “the law of conservation of attractive profits,” value migrated elsewhere, from hardware to software, from IBM to Microsoft. Christensen:

When attractive profits disappear at one stage in the value chain because a product becomes modular and commoditized, the opportunity to earn attractive profits with proprietary products will usually emerge at an adjacent stage.

I believe strongly that open source and open internet standards are doing the same to traditional software. And value is migrating to a new kind of layer, which we now call Web 2.0, which consists of applications driven not just by software but by network-effects databases driven by explicit or implicit user contribution.

So when Larry Ellison says that cloud computing and open source won’t produce many hugely profitable companies, he’s right, but only if you look at the pure software layer. This is a lot like saying that the PC wouldn’t produce many hugely profitable companies, and looking only at hardware vendors! First Microsoft, and now Google give the lie to Ellison’s analysis. The big winners are those who best grasp the rules of the new platform.

So here’s the real trick: cloud computing is real. Everything is moving into the cloud, in whole or in part. The utility layer of cloud computing will be just that, a utility, without outsized profits.

But the cloud platform, like the software platform before it, has new rules for competitive advantage. And chief among those advantages are those that we’ve identified as “Web 2.0”, the design of systems that harness network effects to get better the more people use them.

If Oracle isn’t playing that game, they will one day be doomed to irrelevance. Perhaps, like hardware giants of the past – Compaq, say – they will be absorbed by a bigger company. Or perhaps, like Unisys, they will linger on in specialized markets, too big to go away but no longer on the cutting edge of anything. Or they will understand that it’s not the database software that matters, but the data that it holds, and the services that can be built against that data.

The company that creates the right platform for network effects in data may well achieve the scale that Hugh Macleod envisioned.

P.S. I will be doing two panels on cloud computing at the Web 2.0 Summit in San Francisco the week after next, one on the application layer, and one on the infrastructure layer. Panelists include Paul Maritz (CEO of VMware, who, by the way totally gets what I’m talking about here), Russ Daniels (CTO for cloud services at HP), Padmasree Warrior (CTO at Cisco), the inimitable Marc Benioff of Salesforce.com, Kevin Lynch, CTO of Adobe, and Dave Girouard, who is in charge of Google Apps for the Enterprise. Should be some interesting conversations on the subjects raised in this post!

tags: , , , ,
  • Jen Anderson

    Interesting post.
    By the way, the web 2.0 summit panel misses the biggest player of today: Amazon. Why don’t you invite Jeff Barr or Werner Vogels?

  • Google’s App Engine and Salesforce’s Force.com provide access to an installed user base for private, hosted software solutions for small teams. App Engine ties into Google’s authentication backend, allowing restriction of each app instance to Google Apps customers of a given verified domain such as oreilly.com. Developers could sell Google-hosted web applications to Google App customers with security restrictions handled by Google. Postini, Salesforce, and other companies have already connected to Google App Premium subscriber data to deliver additional services outside Google’s hosting model.

    App Engine is still a preview release but there have been some developers hitting their daily quotas. Python is a smaller audience than PHP, and BigTable with its Megastore abstraction is completely new to developers.

    Developers also have to worry about cloud distribution. Where in the world are your clouds? Do they replicate? How do you best group data to serve this distributed audience? Behind some clouds is an app delivery network combined with the well-established static content delivery networks for performant worldwide computing.

  • tom saffell

    Nice job.
    A much needed disambiguation of the inconsistently used phrase ‘cloud computing’.

  • RE: Platform as a Service vs. Utility Computing

    I just wrote a little about this at http://www.25hoursaday.com/weblog/2008/10/19/CloudComputingAndVendorLockIn.aspx

    Utility Computing has two things over Platform as a Service. The first is the content of my linked post; there is a lot less vendor lock-in when you go with a utility computing provider than building an app against proprietary cloud APIs.

    The second is that there is far more power that a developer can get from running their own LAMP or WISC stack than any platform as a service provider is providing today.

  • Tim,

    Excellent analysis.

    Having spent the last couple of years talking at various conferences on the same theme (my work having been blatantly stolen from some guy called Tim’s work on Infoware and the open source paradigm shift), I just want to say how much I agree with you.

    For my latest video monstrosity into this field, I’ve provided a link.


    Simon Wardley

  • Whoa, Tim. I attributed you last month for saying that cloud computing is another name for Web 2.0.

    I took that from your own words in this Web 2.0 Expo video around 5:20 inside. Maybe I misheard you.

    So, which is it? Is Web 2.0 analogous to cloud computing or is cloud computing the architecture for Web 2.0?

  • Very interesting post clarifying the question of “cloud computing”

  • Jim Collins

    I like the way you differentiate the ‘clouds’. In addition to the companies you mentioned there are other tool/utility providers that seem to be gaining traction in the AWS ecosystem – Bucket Explorer, Ylastic, Jollat and PersistentFS. They all provide things that you cannot get currently from AWS. The recenty pre announced AWS offerings for next year will compete with Jollat, Ylastic and RightScale. They all offer management consoles of varying levels right now. But there is still opportunity for these vendors to differentiate themselves and continue to flourish. In fact AMZN should just buy one of these instead of building something from scratch. Amazon is the big guerilla right now, but it will still need to rely on these startups to help nourish its ecosystem.

  • Rodrigo Geve

    On the subject of Google AppEngine. I am an independent developer working mostly with Google’s APIs.

    I have been reading a lot of articles on Cloud Computing lately and have noticed that everyone seems to point out ‘Vendor Lock-in’ as a problem with AppEngine.

    If you look closely at Google’s technology stack, you realize that you are not locked in. For example, while Google’s Datastore is proprietary, their Gears technology uses standard SQL(lite). So if you have an ‘online / offline’ strategy, you have stored your user’s data both on their PC (using Gears), and on AppEngine (in the cloud).

    Also, if you are developing an AJAX application using Google’s GWT technology, then most of your business logic is on the client (compiled Javascript). Thus, the back-end code running on Google’s AppEngine is mostly just for storing and retrieving data.

    So ultimately, I could take my existing application and port it to another Database vendor without too much pain. And since my users data is co-located on their PCs, they could easily export the data and re-import after I theoretically switch to another cloud-based datastore.

    It really all depends on how you architect your application.

    All in all, fear not Google. Use them. Leverage them.

  • Are their really economies of scale in IT? As long as certain processes are dominated by manual labor, unless their is automation their won’t be economies of scale.

    Whether its server management, or core network services (like IPAM or DNS) many IT tasks are still arried out manually while systems and endpoints drive user automation. Large enterprises pay more per IP address than smaller ones…

  • Cloud? or prison cell? as some people.

    As Rodrigo pointed out, very depending on the architecture of your applications.

  • Quite an interesting and thought-provoking post. Will need some time to digest it all, but here are my initial thoughts:

    > what we learned from the history of the IBM personal computer — a commodity platform built from off-the-shelf parts — was that it drained value out of the hardware ecosystem, turning it into a low-margin business. But profits didn’t go away. Instead, through something that Clayton Christensen calls “the law of conservation of attractive profits,” value migrated elsewhere, from hardware to software, from IBM to Microsoft.

    Right, it seems. Another example where Clayton’s law may apply is this:

    In the mobile phone business in India (where there has been, for some time now, one of the highest growth rates of mobile phone usage in the world), ARPUs (Average Revenues Per User) have been low and getting lower – because tariffs are getting lower due to competition between mobile service providers, and due to the Indian government’s mostly consumer-friendly policies (in the area of mobile service). (So there is less value in the “hardware” ecosystem.) But as an individual, I see and hear, every day, evidence of the huge value (financial as well as human) that is being created because almost anyone above the poverty level can afford a mobile phone, nowadays. (There are fairly low rates for usage, and low-end phones (*) don’t cost too much either.) So there’s a lot more value created in the “software” ecosystem.

    From sea fishermen to vegetable vendors to rickshaw drivers to junk dealers, many more people in India are now able to afford mobile phones, and are using them to improve their lives in various ways.

    I first noticed this about a couple of years ago, and the trend only seems to be increasing …

    (*) On a side note, Nokia created a low-end phone, the 2110 (specially for India, I heard) that had a flashlight built in. It sold in huge numbers.

  • Cloud? or prison cell? as some people fear.

    As Rodrigo pointed out, very depending on the architecture of your applications.

  • Tim, good to meet you in Berlin. I think the main issue here is that “Web 2.0” in its current form will struggle to capture the value as it has no way to assert control or take risk.

    My thought on your & Nick Carr’s counterpost here:


  • I think cloud computing will be the future. Even Steve Balmer thinks about to change his strategy of Microsoft.

  • The cloud enables horizontal integration, which contrasts with today vertical approach, so there’s no need for a monopolistic entity to keep everything together.

    The ‘trillion dollar company’ will actually be stratified with several interdependent companies providing network, hardware, datacenter, platform, and software (and all the layers in between).

    Stratification was key to the PCs success, with Intel, IBM, and Microsoft emerging as interdependent players. Expect the same rules to apply when “The network is the computer”.

  • A very nice hip-pocket description of the cloud layers.

    Tim O’Reilly: “This common usage points up a meaningful difference: people tend to think differently about cloud applications when they host individual user data. The prospect of “my” data disappearing or being unavailable is far more alarming than, for example, the disappearance of a service…”

    I think this is a temporary, transitional distinction. People have become used to the idea that they don’t “own” their data in other business spheres, so I expect that once there is a sense that data is safe in the cloud, that fear, and hence distinction, will disappear. In some cases it already is – especially when there are benefits to having data reside in the cloud and therefore easily accessible from anywhere. Google docs exemplifies this.

    The question of where profits reside is an interesting one. Traditionally it has been with the players that can extract high profits from a valuable product or service and maintain a high barrier to entry to that activity. Sometimes governments ensure that this cannot happen by forcing competition.

    Utility computing probably cannot extract high profits, if only because Moore’s law means that this resource is becoming so cheap that the datacenter would eventually just be a global collection of computers, mine, yours and eveyone else’s.

    What about the platform. I think that is more likely, but, what is the barrier to entry in this case – the first mover network effects? I’m not sure the “platform” can, or should be anything more that the rules, customs and places that make a market economy work. In this sense, neutral entities that want this mode of computing economy to grow should just supply the platform and let the market for applications use it.

    If the profits lay with the applications, then there is no giant company to capture the profits, just the usual power law distribution of company sizes, with more or less transient positional power.

  • Nice series of articles on cloud computing in the technology survey section of the Economist (subscription required)


    Tim is nicely quoted too.

  • Richard

    Unfortunately I couldn’t get past “Understanding the dynamics of increasing returns on the web is the essence of what I called Web 2.0.” Web 2.0 wasn’t (isn’t?) about incredible innovation, it is the tech media’s love affair w/ social business models. Unfortunately, most of the “businesses” are mere tools or features.

    “Ultimately, on the network, applications win if they get better the more people use them.” Big insight – ultimately, in the real world, businesses win if… more people use them.

  • I don’t think we’ll see someone dominate “the cloud” because of all the different facets of cloud computing that will emerge.

    Rather I think we’ll be able to pick and choose the pieces from each service provider and cobble together a cloud system that works faster or cheaper or more efficiently for a specific application.

  • Tim, no one is using AppEngine because it’s still alpha quality software. The quotas and caps currently in place mean no startup in their right mind would tie themselves to it yet, which is shame. For a more technical discussion on the issues facing AppEngine, see Aral Balkin’s article:


  • nmw

    I dugg this.

    I have heard many people use “dugg” as a verb — and much in the sense, I guess, the way the Beatles used the term.

    Language is the ultimate cloud, language is the quintessential utility, language is the network.

    Like I said about a year ago in the run-up to a Facebook conference in Seattle: Applications don’t matter — and neither does hardware (as, I think someone once said long before me — see http://www.nicholasgcarr.com ;).

    The thing about language is: Some words are “more equal” then others (isn’t that kind of what you mean by power laws?). To say that there are 1×10^100 possible 63-char strings would only show naiveté. There are, in fact, only a couple thousand strings that, umm, matter.

    Google thinks only Google matters. I don’t care about it much — let them believe that: I know better!

    ;D nmw

  • dormor

    “So when Larry Ellison says that cloud computing and open source won’t produce many hugely profitable companies, he’s right, but only if you look at the pure software layer.” – true. Think of company valuation of http://folding.stanford.edu or http://boinc.bakerlab.org/rosetta/ if they find a cure to cancer or http://www.gstock.com/ and http://www.moneybee.net/ if they manage to find the formula that beats the stock market.

  • “that it doesn’t take into account what causes power laws in online activity”

    This is a dangerous phrase, since it assumes the following: that in the absence of the reasons you give – the absence of a network effect – that no power law would emerge.

    That’s a dangerous assertion since power laws occur everywhere that a human has a choice. You find it in city population sizes, usage of certain letters, aside from the more mundane examples related by “long tail” popularism.

    I have never yet seen a convincing reason as to why you get power laws in things like city sizes, letter frequency usage, or indeed a dozen other things, so assuming that a network won’t exhibit itself simply because there isn’t (yet!) the factors you expect to see for a network effect doesn’t imply there won’t be one. After all, power laws in the web predate the creation of google and friends (I was at web cache workshops discussing them pre-google, though they tended to be referred to as zipf distributions).

    What would be interesting is if you get a qualitative difference in how the power laws form when the choice is assisted in some fashion, leveraging human behaviour – ie if you get some sort of power-squared effect. This would show up as a graph looking like a long tail graph, but on a log-log scale (whereas of course a power-law is a straight line on log-log).

    However, I would fully expect power laws to exhibit themselves in any situation where people are given choice. In cloud computing, the choice is the website you go to to choose to use the system, so you would expect a power law there.

    Indeed, if there wasn’t a power law there, that would be extremely interesting since it would be very unusual, and worthy of research.

  • Michael –

    You make a good point. Network effects are not the only thing that lead to power laws.

    But I think my point stands: assumptions that cloud computing will automatically be a very profitable business, with increasing returns to the winners, are likely wrong. It is more likely to be a commodity business – at least at the infrastructure layer – with multiple competing players.

  • Niall – Really good points. Interesting to hear what you say about AppEngine. I’m clearly not as close to that as I need to be. Will quiz Dave Girouard about that at Summit. And you’re totally right that we could be surprised by some unexpected power law turning up in cloud economics.

    Alan P – Great comments, and great post. I recommend that everyone in this thread read it.

    Baynardo – Of course cloud computing is the future. In fact, if you look around, you can see that it’s the present. Every app that matters is connected. But the question is what are the levers of competitive advantage in the cloud, and how do you use them to increase the success of your application or business. For example, cell phones are cloud devices, and yet they make miserable use of the social network possibilities implicit in their use.

    Richard – You miss the point. *Why* do some people use the services of some businesses more than others?

    Mike Leach: I like your point “Stratification was key to the PCs success”

    Rodrigo – thanks for the thoughts about portability and Google AppEngine. But keep in mind that portability doesn’t mean you can opt out of the system if it truly does have network effect characteristics – i.e. your app or data becomes more valuable if it’s in the same space as others.

  • Tim – very interesting article. We are certainly witnessing the start of a way of computing that will change the game forever. I see that you have several Utility Computing companies listed above, but I did not see Terremark. Terremark’s Enterprise Cloud product has been in production for a couple of months with an SLA and support for Windows, Linux, and Solaris. If you have a chance, an overview of The Enterprise Cloud by Terremark can be seen at http://www.theenterprisecloud.com.

    -Ryan McDermott
    Technology Evangelist

  • Great article because it clarifies the terms and demystifies the subject somewhat.
    My take on this is that for most businesses, cloud computing has major benefits, particularly from a capital point of view. Unfortunately, most businesses will defer the research and decision making to IT departments who will see this as a threat. A major dependency on cloud is the strength of the security measures and the adoption of a strong security policy – and I haven’t seen yet one company who has ticked every box. Again, down to the IT department!

  • Hi Tim,

    You forgot a last category of cloud application: cloud infrastructure application.

    For instance, MxM (my company) is building an mail delivery system in the cloud aimed at cloud computing application (in others). You can find the same infrastructure building blocks with DNS or even Google Maps (a lot of applications use G Maps as an infrastructure layer).

    IMHO, this is a very important development. We can see it with our customers: it reduces cost and is easier to operate.

  • Jen –

    Re. Why no Werner Vogels. Two answers, both of them true.

    1. I think everyone knows Amazon is the leader in this space. We wanted to get some visibility into the thinking of all the pretenders to the throne, as well as to clarify some of the different levels at which you can think about cloud computing (application and infrastructure)

    2. @Werner had a schedule conflict :-)

  • Hi Tim,

    But I think my point stands: assumptions that cloud computing will automatically be a very profitable business, with increasing returns to the winners, are likely wrong. It is more likely to be a commodity business – at least at the infrastructure layer – with multiple competing players.

    Oh, I wasn’t disagreeing with that point. :-)

    Incidentally a power-squared law is something I would look out for, it’s something I suspect will come on the horizon at some point in the future – which will change the economic model again. The scenario I can see causing that is where we continue to rely more and more on systems that model our choices to guide our choices. The facebook apps usage data strikes me as interesting for that reason. (the graphs I’ve seen look closer to power-square law rather than power-law)
    (Such a system would make it worthwhile to be a monopoly, and potentially naturally give rise to one…)

    Whether the cloud will accelerate that possibility is unclear to me.

  • Mashable’s post about Azure and distributed computation provides some really important additional insights into layers of the cloud.

    It reminded me how, back in 2001, when I did my first “internet operating system” conference (long before we called it Web 2.0), I was trying to show how P2P, web services and distributed computation all together showed us something about the shape of the future, what the network could become.

    Since then, so much of the focus has been on the web end of the cloud. Mashable does a good job of reminding us that that’s not all there is.

  • SolarCatcher

    Hi Tim,

    very nice analysis – I think the distinction between utility computing vs. platform as a service vs. cloud-based end-user applications is very useful to understand the different types of cloud computing.

    But it seems a bit as if you were suggesting that every user and every use case was equally suited to the different types. I don’t think that this is the case. I believe that every company or user may have different use cases in which he needs a specific type. Not all data are meant to be aggregated, shared or analysed à la Web 2.0. Payroll data would be one of these – because I would not want any company, including Google or Microsoft, to meddle with them. I could probably get more insight into the overall payroll structure in my country or field of business. But would I really, really trust them? For this type of use cases, I would prefer either my old-fashioned Oracle database on a server running on my premises or a reliable utility computing provider, which offers me to host the database and possibly the whole payroll system for me. But which promises me by contract to not look into them in any meaningful way.

    Another reason for choosing one type over the other could be legislation. We are already seeing more and more regulatory questions popping up in the use of cloud computing. E.g. I recently read that Canadian universities had legal problems outsourcing their campus’ email operations to Google, because Google operates in the US where the government has wide-ranging powers to get hold of the data – which may not be in line with Canadian legislation. The question of where the data are stored may therefore become a very important decision criterion for those seeking cloud computing services. Would I entrust my data US jurisdiction? Or Chinese jurisdiction? Am I allowed to do this at all?

    Questions like these add another layer to the complexity of the cloud market, and I believe that different offerings will stay with us. This would make a single big monopoly rather unlikely.

    Best regards

  • Tim,
    Google just announced its first trial with App Engine powered applications available to Google App users. The first three experimental shared instances are Moderator for conference Q&A, Code Reviews for developer workflows, and Short Links for URL shortcuts.

    Google App administrators configure these new App Engine services inside Google Solutions Marketplace, the premium add-on store for Google App accounts. Solutions Marketplace offers paid and free add-ons to the growing Google App audience.

  • Karsten

    The cloud is basically the web as the new applications space. This means that the Desktop will be thinner and faster but not in terms of thin client. Google Chrome pointed into that direction, a browser that is faster and more lightweight in its user interface. Of course it has nothing to do with cloud computing as such but it demonstrates that the user interface of the client will only get easier and you won’t install complex local software anymore except the elementary stuff. Its all out on the web. Mandriva and gOS show the way with their endorsement of LXDE (http://www.lxde.org). Windows-me too is not the target anymore. This also explains why Microsoft Vista was such a failure. Users don’t want fatter clients anymore that make it complicated for them. They become operating system indifferent and want fast and lightweight operating systems that make more efficient use of the hardware. The Asus eeePC is only a first step into that direction. When I buy a brandnew PC I want to have a faster one. I don’t want to wait for 1 minute to have the desktop booted.

  • Tim:

    I have been following the discussion between you and Nick Carr about Web 2.0 and Cloud Computing. I believe that you are right that network effects have a huge impact on the Cloud Computing marketplace. However, I was thinking that both Nick and you are missing the main point here in assuming that end-users are Cloud-users. But Cloud Computing is a developer-facing business. So developers are the Cloud-users. (http://markusklems.wordpress.com/2008/10/30/tim-oreilly-vs-nick-car/)



  • Tom

    Is Cloud Computing going to be The “Web 3.0”? It certainly seems going that way, if not already.

    Cloud computing seems to be a natural progression for people to access computing powers — from Mainframe to minicomputers, to personal computers (PC), to Cloud Computing.

    What do you think?

  • Hi Tim!

    Great post!
    I’ve been reading about Cloud Computing a lot more lately, since I’ve been more involved with it professionally. What I’ve noticed around is a big resistance to it: the main concern is exactly data loss, therefore, security.
    Also, I believe there’s a big myth by using the “Cloud” term and many people truly get lost with it, so any way to create awareness and identify the real meaning of it can just bring development in general.

    I used this post as a reference on my personal blog. I hope that, by it, you don’t feel offended in anyway, but if you do, please contact me and I’ll make sure to remove it.

    Maisa. (twitter/maydbs)

  • Dave

    I am a developer who uses Amazon’s EC2.
    Amazon EC2 is definitely a commodity business model at this point. There 2 reasons I think Amazon got it this business, 1) practically speaking they have a lot of excessive computing resources that they could resell ( ala when the airlines cut prices to fill seats). This way than can recoup some of their infrastructure costs.
    2) Once customers sign up with EC2, upsell them with the really profitable future software services like consulting services, monitoring software, configuration software, and load balancing software in the cloud. Configuration software/autoscaling provider Rightscale charges a pretty penny for their service.

  • I did a presentation on Cloud Computing two weeks ago at PubCon in Las Vegas on this topic. I did a 15 minute session to help identify truth from fiction in the “sale pitch promises” made by Cloud Computing advocates. Boy did I ear back. So much so that I felt compelled to follow up with an article that addressed on point in finer detail:

    As Cloud Computing gains steam, it is vital that we remember what Cloud Computing is and–perhaps more importantly–what it is not.

    Be well,
    Jeffrey J. Hardy

  • Cloud computing has so many definitions and applications that generalizing on the viability, or trying to pin down exactly what Cloud Computing is and is not is a daunting task for many professionals.

    Some say it saves CIO’s from having to solve problems internally that would be or could be un-solveable due to latency within their own corporate infrastructures.

    In past scenarios, CIOs, CEOs and CFOs have opted to simply buy companies or acquire them as a subsidary company to leverage technology as an edge without having to actually build the technology or service in house.

    With the emergence of web services, common scripting and data exchange languages like Javascript and XML and Ajax and many other emerging technologies, the distributed Cloud Computing model becomes a reality.

    Leveraging systems like Salesforce, Google, Windows Azure, Intuit Online, Adobe Online, and many others to achieve Enterprise Level Mashups is a viable alternative to the Buy verses Build and with the right consultants and teams in place, aggregated tools, data feeds, web services, and advanced-ly powerful applications systems can be defined, created and deployed to the corporate and web facing infrastructures in new record times.

    Cloud Computing is here and has been touted as being as profound an addition to the web as E-Commerce was to the Net-Telligence only a mere decated ago.

    At Nubifer.com networks, we leverage state of the art Neural networks which study real time usage, cache latency, feed uptime, and fastest path switching technologies for companies on the leading edge of integrating more than one Cloud Based system.

    It is clear, that Cloud Computing as a term of service and platform offering is ever-changing and growing every multi-milli-internet second that passes. The question now is, how will you contribute and or capitalize on this paradigm shift enhancement to distributed application development, usage and functionality?

  • Reginald Pugh

    Moving Beyond the Cloud

    The final frontier in computing may move beyond the cloud and reach into a realm that I coin as “galactic-computing”.

    This is where all cloud meet and the neural-centricity of the web is harnessed in a fashion that is far unlike any platform we have seen.

    This “galactic computing” is a be all, see all network of networks.

    There will be 4 levels of access defined for users in the clouds desiring to access it.

    1- stealth access- invisible to others a very high trust level(government)
    2- peer to peer access- IPV6 adhoc connections
    3-multicast access – social interaction with users and networks you trust
    4-broadcast access- open source and access by anyone at anytime, anyplace.

    Each level provides a classification that allows for security, confidentially and integrity.

    Galatic computing will be much like a universal human brain (Wizard of Ozzie-Microsoft ;-) except the level of access is granted — provide the user “mindshare to other brains that are in this neural network. Each user or cloud that is formed becomes a part of the “Master Brain” that has built in “virtual redundancy to avoid becoming part of the blackhole viral disaster that places those joining this Network of Networks.

    Again this is a conceptual description , I am in no way a computer scientist, just a everyday lover of future technology.

  • Thanks for the posts and comments.

    For organizations building commercial Cloud Computing products, my whitepaper “On-Premise to On-Demand: Product Migration or Business Transformation” provides the Dos and Don’ts of making the transition to cloud computing.

    The whitepaper talks about why established ISVs with On-Premise software products need to approach their move to the On-Demand model as a transformation of their business, as opposed to a technical product migration exercise. It highlights the challenges ISVs face and examines the most common missteps during various stages and areas of this transformation. It also provides strategies for success, including advice from companies such as Thomson-Reuters, CODA and Intralinks that have already been down that road.

    You can download the whitepaper from:




  • Hi Tim,

    Great post. Thanks.

    A note on AppEngine. I don’t think its fair to be critical of Appengine especially with comparison to force.com or Amazon because its fairly decommercialized and I presume its meant to be. Google APIs are more like hook up and play stuff than anything else.

    Looking forward to your next post
    Sumit Roy

  • scott silverman

    Cam across this from a colleague. Beta w/ no strings attached. Online workspace. http://www.goeverywhere.com/web_desktop/

  • I wonder how useful it would be for big business units like well established players Exxon, BP who have spent billions in their IT infrastructure.

  • Werner

    From an end-user perspective, especially (large?) business users, the issue will be whether they will have the same or better product/service at all the levels you describe, at a lower cost, and, more importantly, at a variable cost as opposed to a fixed cost.
    Management is notoriously bad at judging whether they are gettint good value for money out of their (fixed cost) IT department and services. If cloud computing delivers similar or better benefits as the IT department at variable prices, and the opportunity to switch willy nilly elsewhere, then management will go for it.
    What remains interesting is who will actually make money from the cloud. I’m reminded of the gold rush, where plenty of folks made money, just not many gold diggers.

  • Cloudguy

    Few short years back we had a similar concept called as Grid Computing. Everyone said that it is is the next killer app. Some even dared to say that we will someday get computing as we get electricity from outlets. Everyone rode that bandwagon.

    Grid computing never really took off. Ian Foster remains known to a few people outside the Grid computing community. I would say that there was enough momentum in Grid Computing to hit a tipping point. Cloud computing is a remnant of that. Of course, cloud is very different in so many ways.

    I personally feel that Cloud computing will not fizzle out as Grid computing did. The key factor to adoption is the availability of tools that can “Cloud-enable” an application. As long as there are open source tools (like Hadoop) that are available and easily implementable cloud will take off.

    Another factor, is even though many companies want to run away from suggesting implementation strategies this will turn of a lot of potential customers. Customers will never adopt cloud computing unless they know how exactly they will implement their code on “Cloud platforms”. This is where Grid computing failed and if such a mistake is repeated Cloud will be just another paper killer app

  • Cloud computing is not just old wine in new bottle. I have a detailed post that describes how it is different than virtualization, P2P, and grid computing. It is not just the technology but the access to the computing platform and the network effects that cloud allows to create without any barriers is a differentiating factor.


  • I enjoyed reading this post – interesting and thoughtful. Only time will tell what cloud computing can be. One thing that both sides of the argument have in common is the interest in the cloud of course, but more importantly, they all have a stake in it.

    I am from the school of thought that says Google will trump us all with their apps one day! Who knows.

    Thanks for the posts and keep it up. I especially liked your posting about Aneesh Chopra, America’s new CTO.

  • lily

    I’ve found an organisation called Servage who apparently have a groundbreaking service. It’s always hard to tell whether something is going to take off, especially with so much hype but the landing page looks cool! http://www.servage.com/. Have checked out their Twitter page (twitter.com/totalcloud)but seems they are waiting for the launch before tweeting their own news…..

  • Jacky

    A lot of service providers thought cloud computing means moving away from PCs. They create applications on the web (cloud) so users can use the applications online, store files online, and never need to use any locally installed applications. The PCs are relegated to dumb terminals. However, this approach will not work for the mass market. Why?

    (1) PCs will not disappear; they become more and more powerful yet the price keeps falling. It is far more competitive than dumb terminals. Even the Net PCs and smart phones also become more and more powerful with every new model. Simply put, vendors will not just sell dumb terminals; and few users will buy dumb terminals.

    (2) With the tremendous horse power available at the local PC, treating the PCs as dumb terminals is a dumb idea. Users are used to desktop software, such as MS office, photo editing software, video editing software, HTML authoring software, etc. they will continue using the desktop software for its more advanced features and better performance than web-based applications. Moreover, be able to work on something offline is a huge advantage over web-based applications.

    (3) Many applications that require a lot of computing power and a lot of network bandwidth are not suitable for the cloud computing. (e.g. video editing, photo editing, etc.)

    (4) Moving everything to the cloud means you are completely locked in by your vendor. It is extremely hard to switch vendors, not to mention that you may not be able to find another compatible vendor.

    (5) Lack of control and flexibility: when you move everything to the cloud, you no longer have full control of your system and data. If the service is down, it will shut down your system; also, you cannot easily add any new services / features to your system; it is also hard to change the current services / features. Any such changes will have to go through your service vendor and at its mercy.

    There are better cloud computing solutions than those of Google, Amazon, and other cloud computing vendors, whose goal might be more focused on destroying Microsoft and PC than offering real better solutions. http://www.DriveHQ.com is one of the first few companies offering cloud storage and cloud IT Solutions. Our cloud solution combines the benefits of local computing and cloud computing. Users can continue using their local desktop software, while at the same time, they can store the data in the cloud. They can store, backup, share and synchronize local and remote data; they can collaborate with a group of people; sharing different folders to different users (or groups) with different permissions. Small businesses can host file server, FTP server, email server, web server, offsite backup system on the cloud. It is much easier to setup and manage group accounts and security settings than local solutions. Not only it saves on hardware, software and system administration cost, it is also more powerful than local solutions as our solution works anywhere. Businesses can easily switch to a central cloud IT system on DriveHQ.com to serve multiple offices.

  • Terry Lawton

    Nice post. I found this blog which describes the top benefits of utilizing cloud computing in general, and a cloud database in particularhttp://blog.caspio.com/web-database/top-benefits-of-database-cloud-computing/