Ginni Rometty is in the process of changing the direction of IBM, and you can see that progress in the latest IBM Annual Report—particularly in the Letter from the Chairman. Unlike most companies, IBM was designed to change, but that doesn’t mean the process isn’t painful. It does mean, though, that this large firm can identify the need to change direction more quickly and survive the effort better than most in its class. The last major direction change was when the market moved from mainframes to PCs and servers, and most of the other mainframe companies and the one server company that was most famous for driving this trend (Sun Microsystems) didn’t fare as well.
Strangely enough, the current change of direction, to focus on cloud services, looks a lot like the old mainframe market in structure (i.e., a few very large companies providing the capability as a service) than it does the client/server cycle we completed last decade.
Let’s explore IBM’s shift in direction.
Next month will be the 50th anniversary of the mainframe, which was supposed to have died out 20 years ago. The mainframe was a centralized computing resource which, though on premise, was paid for like a service (IBM only leased back then) and the user experience was very appliance-like. If you had a problem as a user, you called someone to fix it and your terminal (once they moved beyond card readers and embraced real-time computing) turned on and off like a TV. Mainframes were outstanding for I/O, which is what allowed them to survive 50 years. With the move to the cloud, the mainframe concept is actually coming back into its own, though the hardware underneath the concept is vastly different.
Today, IBM remains the dominant mainframe vendor, though it positions them more like a super server. It’s also its most profitable solution component. However, it and the company need to evolve massively again and that is why Ginni Rometty is shifting the focus of IBM.
When you are providing services to a vast number of users, whether on premise or through cloud services remotely, hardware is only part of the solution. The flexibility required means you have to go far beyond simple I/O and deliver a set of packaged, flexible services that can be automatically assigned and billed (interoffice or externally) easily at a competitive price. This is one area where the old mainframe and mainframe solutions weren’t up to the task and where a great deal of improvement has been made over the last couple of decades.
While the concept of the cloud suggests that a company should be geographically flexible with resources to optimize on price, you still have to conform with regulatory and policy restrictions. This is especially true if you are providing a solution to larger companies or government entities, which are bound by them.
One of the biggest drivers going forward for cloud services is analytics, which can provide answers to users for an increasing variety of questions. This can range from an Apple Siri-like question to a mission-critical inquiry on customer behavior, or a doctor attempting to diagnose a disease.
It is this kind of intelligence that will define the winners and losers as we exit this decade.
IBM’s Annual Report
If you read through the annual report, and pay special attention to Ginni Rometty’s letter, a couple of things jump out and dovetail with these requirements. One of these things is how aggressively IBM is using social networking inside the company. Apparently, the company has 300K IBM users and 200K social communities inside of IBM working on projects. A core part of the effort appears to have shifted to the social sciences with acquisitions like Kenexa. I’ve often been astounded at how little behavioral science is applied to social networking/collaboration attempts, which is why most companies have trouble today. IBM is going to the heart of that problem and addressing it.
On intelligence, the company is making a massive investment in Watson and now others (most notably Google) are moving to enter what is likely to be the definitive technology for next-generation analytics, intelligent systems. This one part will likely go to the heart of IBM’s shift and it will receive over one billion dollars in investment.
As noted above, the company needs distributed resources, and IBM is investing a billion dollars into increasing the number of its SoftLayer data centers to 40 worldwide. This will provide both the performance these regions want and the compliance they require for cloud services.
Another area getting a significant investment is a platform as a service (PaaS) offering that specifically targets the opportunities that services like Amazon Web Services have identified.
Wrapping Up: The IBM of Tomorrow
Ironically, the IBM of tomorrow, after the shift in direction, structurally will look very similar to the IBM prior to the last major change in focus. The company will still build hardware, but increasingly the solutions will be provided as services rather than sold as a package. Those services will be very different though—massively distributed, far more flexible and easy to use, providing infinitely higher performance, more reliable and secure, and far more intelligent than anything IT users have ever seen.
This will be a change for the ages and still only one of many that IBM has made during its uniquely long life cycle. I’m not sure which is more interesting: the eventual intelligent cloud services offerings or the fact that IBM can regularly make such massive changes.