WEB 2.0 and cloud computing

A few months ago Hugh MacLeod's blog article, "The Amazing Secrets of cloud computing", caused a discussion. Hugh thinks: Cloud computing can lead to super monopoly. A few weeks ago Larry Ellison the opposite view, saying that Salesforce.com is hard to make a profit and that no one is making too much money on cloud computing.

In this article I will explain why Ellison is right, and why Ellison is dangerous in terms of Oracle Company's future strategy.

First look at Hugh MacLeod's point of view:

...... No one notices the power law. Nobody thought that one day a company would rule the cloud, just like Google's search business, Microsoft's dominant software.





can you imagine such a company except the monopoly problem? We're not talking about a billions of-dollar business like today's Microsoft or Google. These businesses are dwarfed by what we are talking about. This could be trillions of of the company. It could be the biggest business ever.





I think a lot of my friends who work in these companies know this and how much money they have.





who's going to have a relationship between windows and Apple? It's pediatrics. There are bigger things ... Its immense extent seems to be kept secret by everyone, at least for outsiders like me.

The problem with this analysis is that it does not take into account the causes of the online domain power law. The Web 2.0 core I'm talking about is to understand the dynamics of incremental returns on the Web. Ultimately, on the web, if an application is more effective, the better it will be. As I pointed out in 2005, Google, Amazon, ebay, Craigslist, wikipedia, and other Web 2.0 superstar companies have this feature.

Hugh seems to use cloud computing as a synonym for infrastructure like Amazon S3 and EC2, at least in the sense that cloud computing does not. (later on more types of cloud computing.) ）

Of course, large companies have economies of scale in terms of cost, especially in terms of power costs, which cannot be done for small businesses. But now there are a few such big companies Google, Microsoft, Amazon, large enough, some involved in the cloud, some did not. More importantly, economies of scale do not equate to the continuing rewards of user-network effects. Scale economy is the characteristic of mass commodity market, this field has no huge lever effect to the final victor.

Since I'm listening to others, I'm not sure about the accuracy of the message, but it does come from an important source of information: Jeff Bezos said he would welcome Google and Microsoft to compete in the cloud because they would subsidize cloud computing services from other businesses, Amazon always benefits from it. "We are better than the mass goods business," said Jeff, and the facts prove it.

If cloud computing is a mass commodity business, there will be no huge profit Hugh. The business will be big but profitable, just like the virtual host and ISP market. (Be interested in looking at the Racksspace figures.) ）

But because one of radar's goals is to help people think about the future, I'll take a moment to talk about the future and strategy of a super monopoly that might turn cloud computing into Hugh.

The kind of cloud computing

Now that "cloud computing" seems to contain a lot of different content, let me start by looking at the three types of cloud definitions that differ from each other:

Computing tools. Amazon's success in providing virtual machines, storage, and computing with how much they pay for it is a pioneering undertaking, and now everyone wants a piece of the pie. Developers, not end users, are the target group for this cloud.


at this level I have yet to see any strong net effect gains. In addition to Amazon's continued commitment to the business, neither the first crab-eating SmugMug nor any other user has received any benefits, although countless application developers have put their work on AWS. If there are any gains, they are competing for the same resource.


that if developers accept this platform, it is possible to create ecological advantages for developers like Microsoft. The more developers have mastered the AWS application development skills, the more talented people are. But don't forget: Microsoft is charging developers for this, and building tools not only creates revenue streams for themselves but also makes developers rely on the platform. And they build more and more cryptic and complex APIs that bind developers tighter to their platforms.


so far most AWS tools and high-level APIs have been developed by third parties. We've started to see a very important tool chain from companies like Heroku, Rightscale, and Engineyard, not based on AWS, on their own platforms, and then in a ror way to provide the cloud infrastructure. And you can see these companies are fulfilling their promises, independent of any cloud-computing infrastructure provider.


in short, if Amzon wants to do a blockade to achieve a real competitive advantage (rather than the advantage of being a low-cost provider), they will develop their own advanced API and development tools, or acquire a good startup to do it. Alternatively, if current trends continue, I would like to see Amazon as a foundation for Linux-like applications, tools, and services that are not controlled by Amazon, and do not have a Windows-style API and tools like Microsoft. There will be many providers of mass commodity infrastructure, as well as competing and coordinated tool providers. Given the momentum of open source and cloud computing, this is possible in the future. Platform as a service. Further, by sheer computing tools, such as the Google app and Salesforce force.com platforms, they hide machine instances behind higher-level APIs. Porting applications between these platforms is more likely to migrate from Mac to Windows than from different Linux publishing packs.


There are also problems at this level: developers on these platforms have an advantage over other developers on the same platform? Force.com has a bit of ecosystem benefits for me, and the more developers there are, the better for Salesforce and other application developers. I didn't see that on the app. and applications deployed on app are trivial compared to Amazon and force.com platforms. Is this because developers have a wide berth at Google, or do Google's APIs not give developers enough control and ownership over their important applications? I agree with you on this issue. End-user applications based on cloud computing. In this sense, any Web application is a cloud computing application. Google, Amazon, Facebook, Twitter, Flickr, and all other Web 2.0 apps are cloud computing. However, it seems to me that people use the term cloud computing to more specifically describe web applications that previously needed to be distributed to local PCs, such as tabular software, text processing, databases, and even mail. While all applications are done on the same server farm, people tend to refer to Gmail, Google Docs and spreadsheets as "cloud apps", but not to Google's search and map business.


This commonality suggests a difference: people like to treat individual user data differently when they are mastered, known as cloud computing applications. The loss or availability of your own data is more frightening than the disappearance of data aggregation services (such as Yahoo! Search or Microsoft Live maps), where the aggregated data can be found elsewhere. This directly brings us back to the center of the Web 2.0 proposition: Users add value by using applications. Without that, you can only return to the old ways of public consumption calculations.


Ideally, user data will become more valuable. That's why Craigslist or ebay are more powerful than individual blogs, why Amazon is more successful than a single bookstore, why Google search results page one page or Google ads are more valuable than Microsoft or Yahoo!. That's why all social networking sites are racing to build their own social graph instead of using someone else's.

This highest level of cloud computing certainly has a network effect. If I had to choose, I think the application level of developers will transition to the underlying infrastructure level, two levels eventually meet in the middle. You can actually think that force.com is already doing this. This platform I feel strongly should give more attention to.

The conservation law of tempting profits

A lot of my thoughts about Web 2.0 come directly from my idea of open source. I spoke in the "Open Source Paradigm Shift" from the history of the IBM Personal computer, a mass commodity platform built on the parts of the market, to see it drain the value of the hardware and turn it into a low-margin industry. But profits have not gone away. As Clayton Christensen's "Seductive Profit Conservation" describes, value migration, from hardware to software, from IBM to Microsoft. Christensen that:

when tempting profits with a product standardization and mass commercialization in one part of the value chain disappear, the opportunity to earn attractive profits often appear in the adjacent parts.

I am very confident that open source and open Internet standards are having the same impact on the traditional software industry. Value is shifting to a new level, which we call Web 2.0, and the application that makes up its entirety is driven not only by software drivers, but also by databases with network effects, which are also driven by explicit or implied user contributions.

So Larry Ellison that cloud computing and open source are unlikely to produce too many lucrative companies, and he's right, but it's limited to software alone. This is like saying that the PC industry is unlikely to have a lucrative business, but is limited to the hardware vendor category. First, Microsoft is now Google's proof of the fallacy of Ellison's analysis. The final winners are those who really grasp the rules of the new platform.

So the real point is that cloud computing is real. Everything is moving towards cloud computing. Cloud computing is just a tool at the level of computing tools, and there is not much room for profit.

But cloud computing platforms have new rules for competitive advantage, just as the software platforms of the year. The most important part of these advantages is what we call "Web 2.0", which is designed to harness the effects of the network, and the more people we use, the better.

If Oracle does not participate in the cloud computing game, it is destined to be marginalized one day, like the former hardware giant--compaq--was eaten by larger companies. Or, like Unisys, may survive in a particular market, too big to die, but nothing to do. Perhaps they understand that the important thing is no longer the database software, but the data, and the services built up by the data.

Companies that can create the right platform for data network effects are likely to achieve the scale described by Hugh MacLeod.

Also: I will participate in two groups at the Web 2.0 Summit in San Francisco in two weeks, one on the cloud computing application level and the other on the infrastructure level. Team members include Paul Maritz (VMware CEO), Russ Daniels (CTO of HP Cloud Computing Services), Padmasree Warrior (Cisco CTO), Salesforce.com Marc Benioff, Kevin Lynch (Adobe CTO) and Dave Girouard, who is responsible for corporate Google Apps. There will be a lot of interesting discussions about the issues raised in this article!