Why IBM desperately needed to buy SoftLayer

IBM’s acquisition of SoftLayer is a bid to make the IT giant relevant in a world where Amazon Web Services has come in from left field to snarf up workloads that IBM would very much like to own. That’s a big problem for Big Blue.

ibmlogoIncreasingly, IBM(s ibm) is not just competing with age-old hardware and software rivals like Oracle(s orcl) and HP(s hpq), but also with Amazon(s amzn). And, going forward, IBM will butt heads more with Google(s goog) and Microsoft(s msft), which have staked big claims in public cloud infrastructure.

Armonk, we have a problem

IBM’s issue: It says it has a public cloud presence in SmartCloud but most of the world doesn’t know about it. Granted, IBM can sell SmartCloud to its existing (and large) customer base of Fortune 500 companies, but if it wants to be relevant at all to newer, nimbler and more innovative customer accounts, it needed to do something. And, despite IBM’s claims to the contrary, many of those big existing enterprise customers are also either thinking about or actually putting more of their work on AWS. As a recent Morgan Stanley report put it, AWS is a very real threat to IBM and the rest of the legacy IT superstars.

IBM says its current SmartCloud business — without SoftLayer — was expected to generate $7 billion in revenue in 18 months. Not too shabby on the surface, but I would bet that number derives from a melange of IBM hardware, software and services that others might not really consider “cloud” at all. Softlayer is privately held and the most recent available numbers had it reporting total sales of about $335 million for its 2011 fiscal year.

Dennis Quan, IBM’s vice president of SmartCloud, told me the plan is to build a “compelling IaaS layer that leverages IBM strengths in open standard-based private cloud, enterprise workloads and use of Openstack married with the speed and scale of what SoftLayer has today.”

To non-IBMers, this sounds like a matter of glomming together at least two disparate sets of technology. A Frankencloud of sorts.

No question, IBM had to move to shore up SmartCloud. A former IBM executive, who did not want to be named, told me that the underlying SmartCloud code is a nightmare mish mash of aging technology from IBM’s past. It could not compete with what Amazon’s AWS offers. As a result, he said, IBM started looking seriously at cloud acquisitions to remedy that situation. Both SoftLayer and Rackspace(S rax) were on the list of potential targets.

There are lots of questions around the new IBM Global Cloud Services unit that will combine SmartCloud and SoftLayer efforts, as commenters to our earlier story pointed out:

  1. SoftLayer caters to younger, smaller, cooler companies — game makers, database-as-a-service companies including Slideshare, Kixeye, SendGrid and Cloudant. IBM SmartCloud customers tend to be, well existing IBM customers — enterprises. Will culture shock result?
  2. IBM has been a loud proponent of OpenStack since joining the OpenStack Foundation last year. SoftLayer offers an OpenStack Swift-based storage option but has strong allegiances with Citrix, which backs the CloudStack open source effort. So what happens there?
  3. Will the new division start building or acquiring additional new services to compete with AWS and to attempt to lock customers in, as AWS does with its higher-level services?

Is it too late?

IBM said it will continue to support SoftLayer’s private, public cloud and bare-metal infrastructure options — which could appeal to enterprise customers who need more than “four different VM sizes” on a list of resources, as Erich Clementi, SVP, IBM Global Technology Service put it on a conference call on Tuesday.

Update: Lydia Leong, Gartner research vice president, said SoftLayer is a good opportunistic buy for IBM.  SoftLayer “offers dedicated hosting systems that IBM does not and is highly automated, highly agile, cost-conscious and fast. It takes credit cards and is SMB centric so there’s not a lot of product overlap with IBM,” she told me after IBM’s conference call. 

IBM is a big, important company, but its ability to turn out innovative  stuff has been constrained by a hairball of legacy technologies. The question now is whether it will take what is good about SoftLayer and infuse that into the rest of the IBM cloud (one hopes!) or muddies what is great about SoftLayer. And, to IBM’s point, we are still early in the cloud adoption cycle and the stakes are huge.

Gartner says the market for public cloud services will grow 18.5 percent to $131 billion in this year from $111 billion in 2012. That’s a big opportunity and IBM is right to try to accelerate its growth there. The question is whether Amazon’s head start and early domination of this market means IBM will be an also-ran.

SoftLayer CTO Duke Skarda will be on hand to talk about cloud adoption and the changing landscape at GigaOM’s Structure conference later this month.

This story was updated at 12:31 p.m. PDT with Gartner comment.