IBM will crush Amazon - no it won't, yes it will

Den Howlett Profile picture for user gonzodaddy November 12, 2013
The furious debate around whether IBM or Amazon will dominate the cloud infrastructure market ratcheted up a couple of notches this week. Who wins? Decide for yourself as I dissect the arguments.

If you are a buyer or CIO/CTO in the enterprise space thinking about cloud architectures, you'd be forgiven for wondering what to make of the recent media comment around IBM's decision to shutter SCE in favor of SoftLayer.

In the last few days I've seen some extraordinary headline grabbing articles suggesting that (first) IBM 'pulled a fast one' on its customers by swapping out SCE for SoftLayer. Sheesh. A few days earlier, the author questioned whether anyone uses SCE. Then Rob Enderle weighed in with a glowing piece entitled: Why IBM Will Win the War With Amazon Web Services. This was so positive towards IBM it was borderline close to PR BS. And I say that advisedly because in the past I have listened and learned a great deal of value from Enderle.

Then comes David Linthicum with a counterpiece to Enderle entitled:Amazon Web Services has no reason to worry about IBM. I rate Linthicum less because he is a new style 'cloud first guy' and that bias isn't made clear in the article.

Interspersed among this, El Reg pipes up with details about why the CIA chose Amazon over IBM. It's ugly reading. What's going on and who is right in what appear to be intractable views at two ends of a very wide spectrum?

The short answer is that everyone is partially right but mostly for the wrong reasons and for the wrong markets. In examining the positions, I've spoken with people who are close to IBM. They're not giving me any inside track but the information I have been given is good enough for me to assess these arguments in a more balanced fashion.

In the red corner - Enderle

Amazon is exposed to marketing that disparages its offerings. It lacks the longevity in market to have the kind of loyalty and brand trust that companies such as IBM enjoy. This means Amazon's voice simply isn't as credible or influential in the space. Like Apple of old, Amazon hasn't yet developed the skills needed to defend against a marketing campaign that disparages it. Worse, Amazon is light with marketing execution in general. This suggests that another vendor could take control of its brand image and do significant damage.

I'm not convinced. That didn't work in the CIA case because the CIA looked a little deeper than IBM gave them credit for. Who outsold whom? Will IBM make the same mistakes again? Customers are now on alert and Amazon has a very credible reference flag in can hold up at RFP time. However, what we don't know is what kinds of workload the CIA is anticipating even though it wants high availability.

If for example this is in non-critical systems then Amazon may well turn out to be a good fit. We won't know until this project gets some distance down the track. At least one colleague believes the CIA backed the wrong horse and that the CIA deal will end in tears. If it does then some - maybe all - bets are off. More to the point, recent history is showing us that marketing - of itself - is not enough.

CIO/CTOs are questioning everything as they think of new ways of using technology to build a different future. In an article entitled Delivering the Organization of the Future, powered by IT, Ben Kepes provides a video of Jonathan Murray, EVP and Chief Technology Office, Warner Music Group talking about how he is refactoring IT. I've posted it at the top because it is an object lesson in how to think about IT as a value enabler in the 21st century. In that video, Murray makes clear that he doesn't care about infrastructure. If that's an indicator of the future then IBM has to become a commodity play around which it delivers services. In that scenario, IBM doesn't have to win anything as it relates to Amazon. It only has to convince existing customers that its infrastructure can be delivered at commodity pricing. That's not easy but it is do-able.

My one major caveat around Enderle's argument is that IBM's marketing may be aggressive and it may be super well funded but it is failing to play to its strengths. That's a big no-no for me.

In the blue corner - Linthicum

Linthicum's argument is weaker than Enderle's for one basic reason:

IBM won't beat Amazon -- or the other major cloud providers, for that matter -- because IBM is, well, IBM. Although IBM does understand the enterprise and certainly has the sales force to go sell stuff to enterprises, it doesn't have a culture prepared to promote public cloud services, which are very different than enterprise hardware and software.

In this non-argument, Linthicum demonstrates an astonishing lack of understanding around how large scale enterprise relationships operate. So for instance, I know several consultancies that actively promote enterprise scale applications development and testing in Amazon's cloud because it is a cheap way to try things out albeit in a known risk fashion.

But there is no way those same SIs will (yet) trust Amazon to deliver fault tolerant cloud infrastructure of the kind needed at in-production petabyte scale. However, given Murray's comments above, it can only be a matter of time before Amazon - if it chooses - gets with the program on reliability alongside scalability. It's one thing to have Netflix barfing from time to time. It's an altogether different thing to have your real-time trading system going into 'unexpected maintenance.'


There is a perfectly good reason for arguing that IBM need not go balls out into the public cloud infrastructure business right now. It has enough on its plate with data center operations for existing mega scale customers and the looming hybrid cloud movement. SoftLayer provides a platform for it to do so in the future and engineer its way around some of the technical anomalies that exist today.

I see IBM buying time while the market fleshes out. When Linthicum noted that IBM was barely a pin prick on the market where Amazon had scooped the lion's share of a $3.5 billion business I could only smile. Has he read IBM's financials recently? That's a fraction of IBM's current business and as noted above, can be substituted for serivces. Amazon on the other hand is providing us with a vision for commodity infrastructure which clearly resonates across many IT scenarios.

While my good friend Vinnie Mirchandani has, for years bemoaned IBM's champagne pricing, I now wonder whether the market is catching up in its understanding of how they can think sensibly about crushing those costs. There can be no other explanation for IBM's aggressive positioning - however wrong headed it looks today.

But then as both parties agree, who wants to bet against an old dog and especially IBM? I don't. Remember, it is barely 20 years since they went through a near death experience. Many of the current leadership lived through that experience. They know the value of lessons from the past. So who wins in this zero sum game? Neither. Both do but in their own spaces and in their own ways.

Featured image credit: Umfrage - fotolia

A grey colored placeholder image