There is a new debate brewing in the tech blogosphere about whether network
effects play a role in the success of a cloud computing vendor or not. It all started
with Hugh Macleod making an argument in his post that there is a potential for a
single company to achieve monopoly in the cloud infrastructure segment due to power laws.
But 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.
Web thought leader, Tim O’ Reilly, countered Hugh’s argument saying that it is the network
effects rather than power law that will ultimately determine the winner in any web marketplace and he, further, argued that since there is no possibility for network
effects in the cloud infrastructure space, we may not see the kind of monopoly Hugh is talking
about.
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.
Immediately, the czar of Cloud Computing Nick Carr jumped into the debate and argued that Tim O’ Reilly’s
arguments are wrong because the very basis of his thesis, Google’s success based
on network effects, is not true and hence Tim’s argument against Hugh’s power law
description is invalid.
Let’s stop here, and take a look at the big kahuna on the Net, Google, which
O’Reilly lists as the first example of a business that has grown to dominance
thanks to the network effect. Is the network effect really the main engine
fueling Google’s dominance of the search market? I would argue that it certainly
is not.
He then argues that the Cloud Computing infrastructure industry is shaped in
such a way that the power law distributions will emerge and he also lists several
reasons to support his claim.
I still haven’t found time to think deeply about the debate and come up with
my own conclusions. I might do it sometime in the future. However, I would like
to list the thoughts that popped up in my mind as I was reading through Tim O’ Relilly’s
and Nick Carr’s articles. I want to emphasize once again that these are not well
thought out ideas but rather some quick thoughts that came up when I was reading
these articles. But, I am writing this post because I want to throw them into the
debate, when it is hot, and see if it makes any sense in the first place.
First, I agree with Nick Carr that Tim O’ Reilly is sort of obsessed with the ideas
he puts forward and tries to see everything else based on that framework. Whether it is
his assertion that open source has no place in the SaaS world or the current arguments
based on the networks effects in Cloud Computing space, he tries to see the whole
world using his lens. I disagree with Tim on this approach. However, I want to highlight that his recent post is a well thought out piece with some
great insights on the issue.
I also completely agree with his conclusion that no one player will end up
monopolizing the cloud infrastructure space but I disagree with his arguments to
reach this conclusion. The cloud infrastructure marketplace is similar to the
web itself. We all know that there is no single monopoly holding the web infrastructure hostage. If the web can
survive with an “open source” philosophy, I don’t understand why an open
standards based diverse cloud infrastructure marketplace cannot exist. I somehow feel that the cloud marketplace will have a diverse group of vendors playing the game
defined by the rules of open standards.
In order to disagree with Nick Carr’s arguments on the possibility of
a single player monopolizing the cloud infrastructure marketplace, I want to use
the same example which he had used in the book The Big Switch.
Even though I would like to argue against Nick Carr’s dismissal of the role of
open standards, I will just accept it for the sake of arguments and assume
that the cloud infrastructure space takes a proprietary route. I can then use
the electricity analogy used by Nick Carr in his book. In a way, electricity
production and distribution can be considered as a proprietary system. Nick Carr
compares Cloud Computing with electricity in his book. If Nick Carr’s
arguments in his response to Tim O’ Reilly is valid, why are we not seeing a
single electricity provider monopolizing the entire world market in this global
economy? I would argue that Nick Carr’s and Hugh Macleod’s argument about
a single cloud infrastructure player is very simplistic and it fails to take
into account diversity in the marketplace both from the vendors side and the
consumers side.
Even though I don’t agree with Tim O’ Reilly’s attempts to bring in network effects to
study the cloud infrastructure marketplace, I do want to point out about the
possible role of network effects and what it means from a cloud computing
perspective. Tim argues that there is absolutely no role for network effects in
the infrastructure marketplace. I wouldn’t say there is no role at all. I would
only argue that the role of network effects are not big enough to create
monopolies. For example, if we consider the Amazon ecosystem, having more and
more users (developers or companies) will be useful in supporting an environment
where all these services share data with one another using open standards based
API. If all these services are within the Amazon infrastructure itself, the
developers and companies will not incur charges for the data transfer through
the APIs as Amazon doesn’t charge for any bandwidth used within its own network. This cost savings will motivate more and more developers and companies to share
data more freely without worrying about the cost overhead to their business.
There is some role for network effects in the cloud infrastructure
marketplace but it is not substantial.
As Tim himself pointed out in his post, the network effects plays a significant role in
the case of SaaS. If more and more people use Google Docs or Zoho, it becomes
easier and easier for the users to collaborate with other users. Here I could visualize a
role for network effects and how it will help a vendor capture a huge
marketshare.
Anyhow, these were my initial observations while reading both the articles.
Both Tim O’ Reilly and Nick Carr raise valid arguments justifying their points
of view. I think I disagree with both of them on their line of thinking but I do
agree with Tim that a single huge provider theory is just a myth. What do you
think? Who do you think is right and why?
I was curious about Tim’s assertion that Oracle is missing the boat on the network effects benefit of cloud computing:
“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.”
I disagree that Oracle has the wrong approach. In the enterprise market, I don’t think companies are ready to put their databases in the cloud…and let third parties access the data or metadata about its usage to achieve network effect benefits. What benefit is there to the enterprise?
Large companies’ size is one of their advantages – they see more about a market than do smaller players. If they participate in a platform where everyone benefits from their participation, arguably they’ve just given up on a competitive advantage.
I don’t disagree about network effects with office apps (docs, spreadsheets, etc). But I don’t see any advantage to enterprises in letting third parties access their data to achieve network effects.
@Hutch, true. Thatz my point too. Network effects don’t play a predominant role in IaaS and PaaS.