There are updates to this page that haven't been applied because you've entered text. Refresh this page to see updates.
Hide this message.

IBM thinks Google, Amazon are too amateur to serve enterprise computing needs. Is that true? Do you think IBM and HP can make a real impact?

In the business week cover story (page 5),we see Ric Telford, IBM's vice-president for cloud services saying that Google, Amazon are too amateur to serve the enterprise customer. While, I think this is same as saying PC's wont work in the late 70's, how much chance do you think companies like IBM and HP stand in serving the enterprise customer on Cloud?
Mauricio GodoyMauricio Godoy, IBM Cloud Computing - Analyst ... (more)
I would like to make a quick clarification about the statement that originated this thread. At no point did an IBM executive label Google or Amazon as amateurs. 

The reporter has surrounded his own statement with quotes from IBM  but we are not quoted directly saying that about anyone nor did we offer that opinion as part of the interview process.
Ernest MuellerErnest Mueller, Have implemented several cloud... (more)
Speaking as an enterprise cloud user, there is some truth to that - but not enough. Let me start by saying my company does business with all these companies (Amazon, Microsoft, Google, IBM, HP) and we value our partnership with all of them. But from an "enterprise cloud customer" point of view, they all have their challenges.

With Google, it depends on what you mean by "cloud" of course... Google Apps is industrial strength, but Google App Engine is not, it's more of a cool tech toy, we only use it for non-mission-critical services like our public health dashboard. They're just feeling their way through how to do enterprise sales; I'm a techie and even I thought the presentation they came in and did for Google Apps needed to be a lot more polished.

Amazon AWS has great technology and a great service.  We host some of our paid SaaS applications on AWS. Dealing with them as an enterprise customer can be very frustrating. Getting a sales rep in the first place and paying with a PO is doable once you figure out how to get through the wall-of-robots, but even my Amazon sales rep has started trying to steer me to an Amazon reseller when I want to do "fancy" stuff like prepay, or get flexibility with reserve instances, or whatnot. (The inflexibility of reserve instances, and my inability to even prepay or get any kind of discounting, is extremely unhappy from an enterprise customer POV). But, they have the best cloud tech, so we deal with that. They are also not prepared to pursue or take advantage of more high level partnerships; they seem to be very intent on staying a focused utility supplier and don't want to entangle themselves with anyone (not saying that's wrong, just the way it is).

Microsoft Azure has an up-and-coming technology that's a lot better than I expected it to be (I'm an old UNIX/open source guy). We are moving towards hosting some of our paid SaaS applications on Azure. They are... Kinda easy to work with from an enterprise point of view, at least they have the financial mechanics down. You have to deal with "Microsoft Messaging (tm)" and their unending efforts to convince you and your management that everything would magically work if only you'd just rip out every piece of technology you have and replace it with Microsoft technology.  The unending cross-selling, jockeying for negotiating position behind every single phone call, etc. is very tiring from a technical staff perspective. And they admix marketing message with technical documentation to the point where most people I meet are confused about Azure's actual technical capabilities.

We talked to IBM about their cloud offerings.  IBM is "enterprise friendly" in a certain sense - they can talk high level partnerships and such.  They do try to sell you millions of dollars in services all the time, and we're more of a do it yourself company, so that doesn't go over well here (some companies seem to take joy in handing over briefcases of cash for work a single $75k/year sysadmin could do, but we're not one of those).  And their cloud offerings are still mostly in the planning stage, or "we'll build something out for you for those millions of dollars in services."  We ended up not piloting anything on the IBM cloud because at the time it wasn't really low hassle enough.  (Approximately the same thing happened with Verizon, though their tech was actually more impressive.) Here I think some of the "enterprise focused" companies don't understand the value prop of cloud to us enterprises - we want less hassle as much as the startups do. If I have to put together a 12 month engagement plan, it's not "cloud", it's traditional IT outsourcing.

HP has no real public cloud offerings.  If you go click on their "Enterprise Cloud" link on their Web site you get links to buy HP ProLiant blade systems and/or professional services. We use plenty of HP software in house, but this is one of the worst cases of just taking your existing product line and slapping "cloud" on it.  If "enterprise friendly" means "can sell stuff to wooly-minded execs who have no idea about cloud", it's certainly enterprise friendly. I may be biased because I consider "private cloud" to be basically a big lie, or at best a reminder that in your own data center you should probably be using virtualization and automated provisioning. HP's solutions are quite good for that, but they're not to be confused with what most people mean when they say "cloud."

So in the end, there is currently a mix of companies who have trouble interacting with enterprises in the traditional way but have real useful cloud technology, and companies that work in a more established way but are mainly rebranding existing technology as "cloud" to cash in. This is obviously a continuum with no "right answer", going from "super high tech but you can only pay with credit card" to "$50,000 toilet seat with a 'cloud' sticker on it."  You have to decide what suits your needs best (and as an enterprise, it's never just one - I imagine in 5 years we'll be using something sold as "cloud" from every one of these supplier partners).  Which of these are "enterprise friendly"?  In the end, I buy stuff from all of them, so all of them are.  But all of them also have large strides to make - some on the financial and relationship aspects, and some on the raw technology aspects, and both are important to me as an enterprise architect.
Krishna SankarKrishna Sankar, "Not all those who wander are ... (more)
  • It is not a war between amateurs and professionals
  • It is about compliance, control, legal & policy
  • In that sense, many of the current needs of traditional enterprise applications are different from what  Amazon and Google provide - for example, an enterprise is not going to run their current internal ERP in the cloud
  • Two things will change this neutral equilibrium - a) Enterprise applications (and the compliance environment) will evolve to include the cloud computing model; b) AMZ, Google and the rest will start adding enterprise features to their offering. AWS already has compliance like HIPPA and has VPC for enterprise extension
  • Today [March 8,2011] Randy Bias had a presentation[1] at the Cloud Connect on this topic.
  • While I don't totally agree with his statements and the inferences, he has some good points viz. public cloud model is a viable option for many of the green field applications, provided they are re-architected to be cloud-aware from many aspects - including performance, failure models, latency models, scalability as well as compliance/control.
  • Lastly, there are certain enterprise applications that are well suited for cloud (those are the ones running at AWS/GAE), and there are some that should never be run in a public cloud - most probably IBM is better suited to know which is which, than Amazon or Google !
Ray DePenaRay DePena, Cloudsmith @
3 upvotes by Andy Lemke, Kenik Hassel, and Quora User.
I do believe that Amazon and Google, to a degree, are underestimating the extent of the enterprise challenge. I have a blog post coming out on this very topic.

Amazon and Google have great technology, no doubt, but their level of maturity when it comes to the enterprise has a long way to go. IBM, HP, Oracle, and other traditional IT players have been in the enterprise for years.

That is a battlefield that will not be ceded easily. As I point in out my post, traditional players are vertically organized - they know the industry and business challenges, have built long standing relationships, have dedicated account teams for large enterprise that are experts in their client, industry, relationships, and technology.

If Amazon and Google are to compete in the enterprise, they cannot approach the enterprise from a customer segment mindset (which is what made them successful in their respective segments), they will have to organize to battle for the enterprise.

Understand the regulatory and compliance challenges specific industries face, and how to work with large enterprise customers.

Large enterprise isn't going to rip out their infrastructure just because a Cloud rep comes calling.

To succeed, they will have to bring on board quite a few large enterprise veterans lest they bring a knife to a gun fight.
Matthew McCleanMatthew McClean, Currently working on Cloud Com... (more)
I don't agree with the view that Amazon and Google are too amateur to serve enterprises as mentioned by IBM. First of all the Amazon AWS is the de-facto standard in IaaS services today so that should speak for itself. Many large enterprises are already using AWS services, for various usages from dev & test to data analytics, and even large companies such as Netflix and Zynga are moving their entire IT infratructure to AWS.

There are many reasons why I believe internet companies such as Amazon & Google will continue to be successful in cloud . Primarily it will come down to economics. The business model of internet companies such as Amazon are completely different to traditional Enterprise IT companies like IBM, HP, Cisco in that their solutions are based on using very low-cost commodity HW (Google even builds their own servers), using primarily open source software and then focusing on high level of efficiency in their operations thus passing cost benefits to end users. Enterprise IT players typically sell Hardware and Software to Enterprises and this is where they make their highest margins so it will be almost impossible for them to obtain similar price levels. They do not have the same experience of running operations at the same scale as internet companies making it difficult for them to compete on the efficiency side.

There is also a lot of mention from Enterprise IT players on the fact that Amazon do not offer high enough service levels in their SLA agreement. What really matters however is actual uptime information from historical monitoring. A very good article about this can be found on the Cloud Harmony blog here:
Dave KieransDave Kierans, Extensive user of EC2
1 upvote by Andy Lemke.
Cloud services and expertise are no different from any other product that has been offered for sale through the ages.

Positions for or against different solutions that consider various strengths and weaknesses of present days cloud offerings all are fine and valid provided they are evaluated in the context of their intended use.

It would be folly to generalise and say solution X is inferior to Y because we pre suppose its planned use. This is particularly true for enterprise level consumers of cloud resources. This also goes to the technical leads who roll out high visability projects

Any CIO/CTO worth their salt will be very aware of the short comings of all the competing solutions but, also have to evaluate its suitability against other criteria such as company policy or regulatory compliance issues for example.

If companies like Google, Amazon and Microsoft are happy about moving as much of their own business and operational data and systems to their own clouds then that's good enough for me.
1 Answer Collapsed
Write an answer