Much has been said and written (including by me) about IBM's cloud ambitions and its ability to deliver upon those ambitions.
But time has moved on and, to its credit, IBM has also changed its approach. Its much-heralded acquisition of cloud vendor SoftLayer was seen as a transformative move by the company. Watson, IBM's analytics offering, has gone from merely a brand building initiative (who can forget Watson famously winning Jeopardy?) to something more. Watson is being used as the foundation for a serious of vertical-specific offerings that IBM believes will pay real dividends.
At the same time, however, IBM sometimes can't help itself. It has an ongoing battle with public cloud-heavyweight Amazon Web Services (AWS) over who is the bigger cloud vendor. , to be honest, that is something of a one-sided battle. IBM comes out on a regular basis with press releases heralding its incredible cloud revenue while AWS merely grins in a bemused way. The grins are well-founded, the majority of commentators and analysts pour scorn on IBM's assertions of dominance - it is easy, they opine, to be the biggest cloud vendor when you redefine what constitutes "cloud" on a regular basis.
So in the light of all of this turbulence, I was interested to have a conversation with Robert LeBlanc, SVP of IBM's Cloud business. The broad-ranging conversation was a look at IBM's cloudy future, competing with some serious names in the public cloud and that old chestnut, whose cloud revenue is bigger.
On innovation and IBM's public cloud ambitions
Clearly AWS is a fast-moving vendor and the touchstone by which all other vendors are measured. According to LeBlanc, enterprises see the innovation that AWS is enabling for "cloud-native" organizations and want to transform what they do.
They want to bring additional value to their businesses via the cloud - but need to do so within the context of a hybrid model. According to LeBlanc, and a justifiable argument, is that there is a desire to connect the cloud back to the enterprise but within the context of existing assets. No enterprise is going to lift trillions of dollars of existing assets into the cloud.
New development will happen in the cloud, but existing assets will stay put. While some commentators pour scorn on this "bimodal" approach to IT, it is a pragmatic response to conflicting drivers.
Data and analytics are the big plays
Watson is perhaps the most visible analytics play for IBM, but recent partnerships that Big Blue has announced with
IBM has also recently announced that both
Giving up on the public cloud?
Recently there has been an on-again, off-again situation with regards HP and its public cloud ambitions. I asked LeBlanc if IBM was also ready to concede the public cloud market to AWS.
LeBlanc was quick to state that IBM is not giving up on the public cloud space. Their growth, he believes, will also come from the public cloud space, but in the context of a hybrid story. As he says, most enterprise building in the pure cloud space will still need to integrate back into the enterprise. It is this hybrid cloud story that Big Blue is banking on. The IBM strategy of building out a very broad data center footprint speaks to this.
LeBlanc pointed out that building out a robust, world-class network of data centers is not for the faint of heart. While many of us ridicule IBM's announcements of "yet another billion dollars worth of data center investment", clearly the company is investing heavily in core infrastructure.
LeBlanc is a firm believer that IBM's hybrid story, when combined with the ability to do analytics on data in-country, will give the company an edge against AWS and the other pure cloud vendors. When I pressed LeBlanc on whether, given every vendor's broadening cloud footprint, regional granularity is a defensible position, he was adamant that it was. LeBlanc pointed out that one of the key reasons that Coke went with IBM was its requirement for broad geographic options.
On the cost of traffic
I asked LeBlanc whether, given IBM's strong hybrid story, they would begin to zero rate international traffic. Such a strategy would help drive the message of the value of distributed infrastructure without so much of the cost penalty. He didn't answer directly but did point out that high availability is important. IBM runs its own dark fiber so has control over cost and performance. The company has done some internal benchmarking and believes that it has a cost advantage when running real workloads.
On the AWS-envy thing
I am personally sick of IBM's continual claims to be the biggest cloud vendor. I 've not seen any independent commentator who gives their claims any real credence. I put this to LeBlanc who defended the company's approach saying that it is hard for IBM to more accurately split out infrastructure from other cloud business. As he said, if a customer uses IBM's Kinexa HR system sitting on top of SoftLayer, should that be regarded as SaaS revenue, IaaS revenue or both? I understand LeBlanc's position but still don't feel it justifies IBM's aggressive "we are bigger" messaging push.
On enabling the new apps - moving up the stack
LeBlanc pointed out that IBM has 500 engineers actively working on both OpenStack and Cloud Foundry. The company is continually helping its customers move up the stack. In some cases this might mean the ringfencing of existing applications and opening them up to mobile access, in others this will mean high-level integration and data services. LeBlanc considers IBM's broad portfolio - from infrastructure all the way up, to be a key enabler of their customers innovations.
All in all, it was a positive conversation. IBM certainly has its challenges, but it seems to be doing well to turn the ship (and have no doubt; this is a very big and hard to maneuver ship) around.
