IBM’s journey to the cloud has been a complicated process. As an analyst, I have had a bird’s eye view of the evolution of IBM’s cloud strategy. Like many other enterprise technology companies, IBM has had to balance its need to preserve on- premises revenue with its desire to satisfy customers’ demands for cloud services. I recently attended IBM’s cloud analyst summit, where key executives put the evolving IBM cloud strategy into perspective. In the midst of massive changes in technology and customer requirements, Robert LeBlanc, Senior Vice President and leader of the cloud strategy, succinctly stated the overarching strategy: “Our solutions are cognitive, our platform is the cloud and our focus is industry.” In this blog, I will provide my perspective on what this strategy means for businesses.
What Does Cognitive Mean to Business?
Of all of the aspects of IBM’s strategy, cognitive business is probably the most complex. Typically, we have come to think of Cognitive computing as the technology behind the Watson platform. Watson is a set of underlying data driven technologies that incorporates machine learning, natural language processing, advanced analytics, and developmental APIs. Watson gains insights and learns based on both data and collaboration with subject matter experts. The environment combines dynamic learning, hypothesis generation and evaluation. Watson is delivered to clients as a cloud-based platform intended to create targeted and often industry-focused solutions.
The Insight X-Factor
More recently, with the use of machine learning and other advanced analytics techniques, it is now possible to bring unstructured and semi-structured data (i.e., IoT data) into the mix. Unstructured or semi-structured data exists in everything from emails, articles and books to sensors, images, videos, and even smells. A cognitive approach allows organizations to start to apply analytics and deep learning to these typically untapped, but highly valuable sources of data. When an environment can understand and learn from unstructured and semi-structured data, the system will begin to have the knowledge, insights, and intuition that subject matter experts with years of experience possess.
Smart CEOs have known for decades that, that if they could gain insights into the massive amounts of data inside their organizations as well as other data sources, they could gain a massive competitive advantage. It has long been understood that the winners in business are those organizations that can turn that data into differentiated strategies. Because of the massive amounts of compute power and storage required to understand all of a company’s available data, the bottom line is that without cloud computing companies could not afford to turn data into knowledge.
What Differentiates a Cognitive Cloud?
So what does it mean that IBM’s cloud environment is cognitive? If we take IBM at its word, a cognitive cloud could be profound because it is not just focused on compute and storage but also on advanced data analytics. It would surprise no one to say that nearly all businesses are run on their data and that the most successful businesses have a deep understanding of their data. Businesses have become pretty good at analyzing their structured data to better understand customers – who they are and what they buy. In recent years, with the advent of predictive analytics and near real-time capabilities, businesses have gotten much better at predicting customer behaviors and presenting customized offers.
Cloud as a Platform
While IBM has been a player in cloud computing for almost a decade, its relationship to the cloud has been complicated. Some of the common questions about IBM and the cloud that I’ve heard and sometimes asked have been: How much will IBM invest in the cloud? What will the impact of cloud be on hardware and on-premises software? Can cloud revenue grow fast enough to offset declining on-premises licensing fees? I am sure that there have been hundreds of meetings among IBM executives discussing these fundamental financial questions.
What was most intriguing about this recent analyst cloud meeting was LeBlanc’s assertion, “everything first gets delivered on the cloud.” IBM did not come to this decision lightly. However, faced with the growth of Amazon’s cloud business and the growth of Microsoft’s cloud offerings, IBM could not forego lucrative cloud revenue. The company also could not afford to ignore the reality that public cloud services affords IBM the most cost effective and pragmatic way to offer new business models and services. Likewise, without a solid cloud strategy, emerging companies would have abandoned any potential partnership with IBM in favor of vendors that have well-articulated cloud roadmaps.
This move to the cloud does not mean that IBM is abandoning on-premises solutions. Rather, based on customer demand and use cases, software will continue to be delivered on premises. There is plenty of evidence that many customers will continue to demand that core applications operate on premises. There are customers that are simply not ready for a wholesale move to the cloud (public or private). There are other customers that, because of security or regulatory concerns, will demand that certain applications run on premises. Still other companies are becoming vendors of cloud services for their own customers and want to control capital expenses, so they will run their own private cloud environments.
The Cloud Architecture
One of the interesting aspects of the cloud analyst briefing was the articulation of the cloud platform. In its early days, IBM’s cloud platform was a collection of offerings. Almost everything from servers to a variety of software offerings were labeled as cloud (much to the confusion of customers). That has changed under the direction of Bill Karpovich, the general manager of the cloud platform. He explains that the cloud platform is a single entity. That offers “a full spectrum of services from infrastructure to industry focused software.” The foundation of the IBM cloud is SoftLayer, the cloud platform company founded in 2005 and purchased by IBM in 2013. SoftLayer is best known for giving customers the ability to deploy on bare instead of only on virtualized images. Over the past several years, IBM has enhanced its platform with native data and business process integration services, support for containers and microservices, Platform as a Service based on Bluemix (a dev/ops platform based on cloud foundry), security and data privacy services. IBM also hinted that, with the recent acquisition of Gravitant, it will be introducing a hybrid cloud management suite.
IBM as an Open Source Cloud Player
What is behind this pronouncement? The cloud strategy is predicated on several key fundamentals, chief among them is the adherence to standards. Since the late 1980s, IBM has used open systems and standards as a defensive weapon. In that era IBM was knocked off its pedestal as the undisputed computing leader. To regain its footing, IBM embraced de facto standards such as Unix and TCP/IP. It took IBM several years to convince a skeptical market that it could move away from the proprietary world where they controlled every element to it controlled to a company based on openness. Remember that the development environment Eclipse was originally a propriety IBM development platform. Today, if we look at the foundation of IBM’s cloud offerings, openness is at the core; this includes the use of the Linux operating system, OpenStack, and open source tools such as Cloud Foundry, Chef, Puppet, and Jenkins. In addition, IBM has opened its own IP such as the Watson APIs so that partners can easily develop solution.
Leading with Public Cloud
One of the most surprising announcements was the focus on the public cloud. While IBM has dabbled with offering public cloud services for years, the primary focus has been on private cloud services. This seems to be changing. Not surprising, IBM feels pressure from Amazon Web Services to give customers what they want to buy – self-service public cloud resources. It will be interesting to observe how IBM is able to provide a cohesive set of offerings that can move back and forth between public (theirs and third party offerings) and private clouds. I suspect that this is why IBM is investing heavily in hybrid cloud management.
IBM is going “all in” with industry-based solutions. This is a dramatic change from the early 2000s, when IBM determined that it would no longer offer packaged solutions and would instead focus on enablement and middleware. While we don’t expect IBM to create traditional, monolithic packaged applications, it will create modular services and will revitalize its services organization by offering IT as a Service based on deep industry expertise.
The Take Away
I would summarize IBM’s cloud strategy and the analyst meeting with five takeaways:
- IBM is betting big on solutions-focused clouds incorporating both public and private services, complete with its own integrated platform with open APIs.
- Cognitive services that add deep learning algorithms combined with industry-specific knowledge and data is IBM Cloud’s key differentiator.
- Hybrid cloud management will be the way that IBM brings its myriad of offerings together to support customer demands.
- IBM will focus on public cloud services to become more competitive line of business customers and emerging partners.
- Cloud based dev/ops will become one of the lynchpins of IBM’s Bluemix platform in order to draw developers into the IBM environment.