Mad Dog 21/21: As Ginni Bows Out

January 18, 2017 Hesh Wiener

In 1993, when Lou Gerstner became the boss of IBM, it was failing under inept John Akers. A decade later Gerstner passed IBM’s leadership to Sam Palmisano. Gerstner had revived IBM’s computer business, developed a huge services operation and regained the admiration of customers.

But Palmisano was no Gerstner. The IBM he left to Ginni Rometty in 2012 was, beneath the surface, as loused up as it had been under Akers. Moreover, Rometty, who knows IBM needs to invent a new future, is saddled with underlings who often seem to be looking back rather than ahead.

IBM’s largest strategic problem isn’t the fault of Ginni Rometty, nor of Sam Palmisano, nor Lou Gerstner, nor even John Akers. The information technology industry is not the business IBM helped invent during the middle of the last century. That industry thrived on big, powerful, complex systems and the smaller computers derived from the concepts and technology largely developed and perfected on the biggest iron of each generation. When the IBM of Akers promoted the PC developed under the IBM of his predecessor, Frank Cary, he didn’t at first realize the little machine and the microprocessor that powered it were malignant. By the time Akers left IBM, however, the computer architecture that eventually became known as X86 had metastasized.

Servers: X86 Versus All Others: This chart developed by TPM shows how the server business has been taken over by X86, crushing alternatives, particularly those from IBM.

Gerstner may have seen what was going on throughout computerdom. His plan for IBM was to resist the onslaught of PCs and the servers they grew to become by offering something the PC universe didn’t provide: complete information processing solutions. Instead of breaking IBM up to compete with the many specialized companies emerging in the 1990s, as Akers and his top executives seemed to think was the way to go, Gerstner got IBM to focus on providing total solutions to large commercial and governmental entities. Small customers who wanted to act like these organizations were offered essentially the same package, but scaled back. IBM was no longer selling computers or software. Big Blue was selling IBM solutions comprised of hardware, software, expert services and IBM-managed data center facilities. This was a brilliant idea while it lasted. But it got old. The big growth opportunities were elsewhere.

Google, Amazon, Facebook, and their ilk created a new information technology industry. Two leading legacy players, Apple and to a lesser extent Microsoft, also seem to have caught on. Other IT companies, notably the large and powerful server makers, IBM, Hewlett Packard Enterprise, and Dell, have not yet found ways to fully participate in the emerging cloud and mobile oriented IT world, although they are trying. Cisco Systems has grown a viable computer business from its strong position in data communications technology, but growth has stopped. And Oracle, still preserving the shank of what was formerly Sun Microsystems’ systems business, is casting about.

Lou Gerstner: Credited with saving IBM from financial ruin and creating the company’s huge services business, Gerstner now is part of history; the information technology business is not the one in which he achieved so much.

The data showing the information technology world has changed is abundant and irrefutable. One key measure is the server market. It is relatively easy to gauge, and the two market analysis giants, IDG and Gartner, seem to agree about two things:

One is that most of the revenue from server sales, perhaps 90 percent, comes from the X86 portion of the market. All the other architectures together, including Power, mainframe, and Sparc, come to just north of 10 percent of the market.

The other is that the server business in general is shrinking.

As server sales shift to X86 and generally decline, the vast industry segments tied to proprietary processor hardware get hurt. Trade in peripherals, systems software, middleware, applications software, data communications apparatus and the significant portion of the services business tied to glass house type computers (whether the glass houses are in users’ premises or in service vendors’ data centers) remains large but it is nonetheless dwindling.

Because proprietary server sales are fading even faster than sales of generic machines based on X86 technology, IBM is suffering quite severely. Its legacy operations are wedded to its trade in mainframes and Power servers, not only by direct links to its hardware divisions and their related software groups but also its services business that caters to users of proprietary products. Data from the big research houses show that this has been the case for several years, for all of Rometty’s tenure. This aspect of the information technology business was developing well before Rometty was put in charge of IBM. How her predecessor in Armonk, IBM’s directors, financial analysts and sophisticated investors managed to miss this or ignore it now can be seen as a baffling development. Rometty cannot be lauded enough for confronting the situation, even as skeptics wonder whether she has done all she can to address the matter.

Server Sales In Q3 2016: The decline of the server business was quite severe in 2016, particularly for proprietary architectures.

For the past few years, Rometty has made it clear that she expects the old IBM and its offerings to gracefully fade away. She also seems to believe that there will be some upturns when, for example, IBM releases a new generation of mainframes or Power machines. But nothing about her announced intentions includes the possibility of a revival in legacy operations. The best IBM can do for customers who depend on Big Blue’s unique iron and its surrounding constellation of goods and services is to assure users that it will continue to improve all its products and provide top notch support.

IBM’s growth, according to the company’s management, will come from its foray into cloud computing, its exploitation of Watson and other artificial intelligence type technologies and other efforts it calls strategic initiatives. There are a lot of separate things in this category, some of them real headline-grabbers, like IBM’s interest in blockchain technology and the bitcoins it can create, or IBM’s attempts to apply its big data and AI knowhow to problems in medicine.

What seems to be missing from this picture is an implementation of the core idea behind Lou Gerstner’s success: IBM can and should be a one-stop source of all the vital information technology.

Gerstner used this concept to get IBM’s enterprise customers to build their commitments to IBM and to replace their disparate outside services vendors and often also their internal personnel with IBMers. A byproduct of Gerstner’s strategy has been the movement of IBM’s pool of technical employees from the USA and Europe to India.

When Gerstner first began building up IBM’s services operations, IBM service personnel sometimes worked in the user’s facilities. But that has changed. Today, most of IBM’s services employees are in India, while most of its customers remain in the Americas, Europe and the stronger economies of Asia. And IBM’s key rivals in services are also heavily invested in Indian operations and often in Indian management, too.

The movement to India is only one of the big differences between the IBM of Lou Gerstner and that of Ginni Rometty. A second huge change is IBM’s decision to move into cloud computing. In the cloud, IBM doesn’t have a prominent offering of Power or mainframe. Instead, it has more or less the same kind of cloud operation, X86 server farms, provided by Amazon, Google and others. Sure, each of the big cloud vendors has its own way of operating, but the basic offerings all have a lot in common. They all allow customers to pretty easily add or remove computing capacity. They all support Linux. And so on. The cloud offerings are about as generic as PCs, which means that they are not identical but they sure are similar. It remains to be seen how IBM will compete against outfits that seem to be able to deliver cloud technology at lower prices than IBM while making pretty good money.

Into The Cloud: Computing, even legacy work that has been in glass houses for decades, is moving to the cloud.

So far, IBM’s cloud business is largely independent of the company’s legacy operations. It is based on SoftLayer, which IBM bought and used as the foundation of its cloud business. SoftLayer wasn’t connected to IBM’s proprietary legacy technologies in any way . . . and, aside from marketing slogans, it still isn’t. As it stands, IBM cloud services are not more a part of the historic IBM than the services of Amazon or Google. If an IBM enterprise customer wants to deploy a cloud-based application serving users with mobile devices and turns to SoftLayer, the project won’t have any more of a tie-in to that customer’s glass house Power or mainframe or IBM i central systems than a similar application created in Amazon’s cloud or Google’s cloud or Microsoft’s cloud.

That gap between IBM’s cloud offering and its proprietary systems may help preserve its legacy glass house operations, but it won’t necessarily help IBM get on a fast track to new growth. It’s very hard to forecast the way the cloud computing world will grow, but right now it looks like Amazon, Microsoft, and Google are going to get bigger quite a bit faster than IBM’s cloud operations. IBM seems to be doing okay, but skeptical observers believe that some of IBM’s reported growth is due more to a reclassification of existing services operations than real growth. Two other IT giants, HPE and Oracle, also seem to be having a hard time taking advantage of the opportunities that seem to be more abundant in cloud related computing than elsewhere in the industry.

It is entirely possible that IBM has a plan that includes ways to integrate legacy customers’ in-house and services based operations with the lively cloud. But so far there is no sign that this is the case. The same can be said about HPE and Oracle. They all know the cloud is there. They can see that cloud technologies offer the capabilities and economies to foster new ventures as diverse as Uber, various smart home offerings, photo and video exchanges, Twitter, and Facebook. They can see that none of these very lively businesses is based on traditional glass house computing capability. Sure, some of these new businesses have their own data centers, but the computing facilities are a lot more like the ones built by Amazon than any corporate data center populated by IBM’s servers, or equipment from HPE, or various boxes living in Oracle country.

So, for IBM’s customers, for the company’s shareholders and for kibitzers fascinated with Big Blue, the key question at the start of a new year is: Just what is Rometty going to do? This is almost certainly going to be her last year at the helm of IBM. She has talked about IBM’s future in pieces. Can she somehow cement the various efforts she initiated to the legacy operations she inherited? Can she make IBM into one strong company from an old one that’s slipping and a new one that’s still struggling to prove its mettle?

It would be a shame if she just slipped away quietly, failing to provide some guidance to her successor, to her customers, to her employees and her investors. Like other observers, we are hoping her final bow will be done with a flourish.