Is there a cloud market, or is there an Amazon market? Even as the cloud market booms, it's an open question whether there is room for anyone besides Amazon to benefit. Even as Microsoft dominated desktop computing over the past two decades, Amazon seems set to own the public cloud for years to come, notwithstanding attempts to …
Iaas = obsolete
Startups are buit on AppEngine (or other PaaS). They don't have money to spend on installing and configuring servers, let alone networking. Any startup, that goes Iaas, it's competitor based on a PaaS will outdistance them.
True story from 2011: two startups, same product, one on AppEngine, the other one using node.js + mongoDB (in order to "be cutting edge and cool") on a IasS. The first one launched. The second one launched, crashed under load and is now spending valuable engineering resources on scaling their servers instead of working on features.
Early Age Of the Market
Enterprise it is often moving at snail's pace. We still have millions of corporate XP and mainframe users. There will probably be much more competition in the "Cloud Business" in the years to come.
All the big and financially strong vendors fear that their traditional business will be lost to Amazon, so they develop their own cloud services/products. Google, MS, HP, IBM. There is also scope for differentiation in the dimensions of security (better firewalling etc), hardware platforms, national location of data centers (US data protection is not up to EU standards for example) and also different quality of serivce assurances (provisioning of dedicated CPUs, fiber links etc).
Let's see what happens - Amazon's dominance only exists in a quite young market.
Not like a Microsoft, more like an IBM?
I would think the comparison would be better if it were made against a hardware manufacturer. AWS is IaaS - the key being Infrastructure, not SaaS or PaaS.
As such, the comparisons aren't the same any more. The accusations against Microsoft are that they force you to do things their way, use their software, and tie you into all things Windows. I would argue that AWS is more like there being one supplier of 90% of the PCs in the world - you can put all sorts of software on them, use them in a much more varied way.
Granted there is still a sort of lock-in, but the lock-in is only with the hardware that they are virtualising, and we are almost all using that in the non-cloud world.
The only companies they might put out of business are those that offer infrastructure related products. Amazon won't be writing Office type applications, but they will be offering more things that make it easier for people to build SaaS products on their service.
Here's a thought.
Amazon started out with this cloudy thing because it was a nice way to sell off the needed-for-peak-but-otherwise-excessive computing power and/or perhaps monetize written-off kit for a mite longer. That means their hardware costs are effectively marginal, making it much harder for also-rans to beat them with a cloud-only startup. You win that by scale, something amazon gets for free, as a side effect to their business needs.
This also has its own weaknesses, of course.Seeing, for example, how their infrastructure can crap out in mysterious ways with no redress, well, even just that could provide a much-needed opening for competition. Do you really want to bet the company on a giant that's essentially providing your critical infrastructure as a hobby?
As so often, so here too: The fact that we gained choice means we also gain the need to conciously choose.
I agree - the main reason Amazon doesn't have any major AWS competitors is that they have a huge advantage in marginal costs, from scale and from recovering efficiencies in their core business.
It's the IT equivalent of one of their other sidelines - doing order fulfillment for smaller online vendors. I often see orders from some-small-outfit.com arrive in an Amazon box. Amazon has a big online-order-fulfillment infrastructure in place, with warehousing and shipping and the rest. It's a big, robust operation with excellent economies of scale. So they sell their excess capacity; that's all gravy.
AWS is doing the same thing with their computing infrastructure. It's successful in large part because it's cheap, and it's cheap because the real costs are already absorbed by the primary line of business. Google is doing the same thing, except with PaaS frosting on their cake.
Back in the day, it was pretty common for smaller mainframe shops to run a service bureau business on the side, renting out spare mainframe capacity. That's how startups writing mainframe software (I worked for one for several years) did their development. None of this is new.
Amazon going the way of the mainframe?
This got me mulling over the time they were racking up the debts to build out their infrastructure. And it's one big smooth operation now, very pretty, very profitable, and built on books. Book shipping that they're now single-handedly killing off with the kindle.
I'm sure they'll still have a use for all that infrastructure they've built even after most books cease to be shipped in paper form. Mayhaps even moreso. It just, well, makes you think. They're actually reinventing themselves. That may or may not have an impact on just what is marginal for them, so maybe their billing model will have to follow, too.
Anyhow. Don't focus on the hype. Look at what amazon is really up to. And, of course, while software is far easier to scale out (with or without the cloud) for a start-up, amazon has also made scaling out easier and cheaper for certain classes of "real world apps". In fact, with the advent of additive manufacturing and related tricks (like easy pcb-to-order manufacturing, fabless fabbing, drop shipping, fulfillment, what have you), the next big thing might not even be software/internet/cloud centered. Don't get too stuck in the past (bubble), eh.
There's certainly room for competitors. The only problem is though that so far they all seem a bit naff. The big names anyway.
Take Azure for example. I like MS stuff generally and Azure looked great, but it's a joke. Everything about it seems unfinished, hard to use and pointlessly limited. SQL server for example - one of MS' best products, is nobbled so much on Azure that it's like using SQL Server 2000. Not only that but it is EXPENSIVE.
Maybe if there was some decent competition Amazon wouldn't be having as much of a free ride to monopoly but as it is, they really are the only game in town.
On why no one is catching Amazon…
I wrote a post with my detailed thoughts a couple of months ago on The Wisdom of Clouds: http://news.cnet.com/8301-19413_3-20100535-240/can-any-cloud-catch-amazon-web-services-part-2/
The gist is that only Amazon is doing a great job of focusing on individual developers, rather than "enterprise operations", as most hosting companies are doing. The former results in applications moving to AWS services. The latter results in IT trying to figure out which applications to move to the cloud, which is a much riskier model.
The assertion that "Amazon isn't doing anything that technically locks in customers" is untrue. The "often imitated, but never duplicated" AWS API(s) are an obvious lock-in.
90% of 5%
Amazon may have 90% of the web hosting cloud market (and I'm not sure they do), but that market represents less than 5% of what businesses spend on their IT infrastructure. Amazon is great for web scale but poor for anything that need to be secure, integrated with the wide area network or local to you for data protection or latency reasons.
Amazon has a great offering and it has it's place, but the is plenty of room for others.