Services research is a hot area at IBM, as the corporate giant looks to eek greater profitability from its services division.
The idea behind services research is to build a stable of repeatable industry offerings and capabilities using new and existing IBM technologies. By structuring services -- taking out some of the judgment calls and reducing the emphasis on labor -- IBM hopes to streamline delivery while at the same time improve the quality of its services engagements.
Leading the Big Blue charge is Robert Morris, who in mid-2004 gave up his research role at IBM for a spot in services. Morris had been the head of IBM's Almaden Research Center in Silicon Valley when he embarked on a two-year stint in Global Services. Today he's back on the research side of Big Blue's business, spearheading efforts to help the services organization adopt a structure that reduces its dependency on individual consultants' skills and increases its use of standardized technologies and processes.
For example, IBM had existing fraud management technology that it now can easily wrap into a project to deliver a claims management system for an insurer. "That's an example of a technology asset inserted into what otherwise might be a routine implantation of a claims system," Morris says.
Over the last two years, IBM has worked to set up development labs around the world dedicated to services and build out a range of standard processes and methods. In September, IBM began unveiling service products that take advantage of these standardization efforts. Among the first to debut are a network convergence bundle aimed at helping customers determine their convergence readiness and an IP telephony offering focused on designing, deploying and managing IP telephony infrastructure.
The service products contain blueprints for project elements -- such as requirements definition, implementation methodologies and testing plans -- so IBM's services staff can duplicate the tactics they used for customers around the globe and minimize labor-intensive customizations.
Products such as these complement the work Morris did in the services division, where he focused on assets innovation -- trying to figure out how to work technology assets into IBM's service business in order to deliver higher quality services more quickly and affordably. While there, he not only worked directly with services clients but also made efforts to jumpstart an internal transformation at IBM. A key part of his job revolved around getting IBMers from different parts of the company, such as hardware, software, research and consulting areas, to collaborate.
"IBM has so many assets that could be used for services, but being such a large company, it takes somebody to create links that might not otherwise have been made. That was a big part of my job," Morris says.
IBM's 60-year-old research division, which consists of 3,200 scientists across eight labs and six countries, has traditionally contributed technologies to IBM's software and hardware business units. The emphasis on using R&D assets to better services operations is a new focus that has emerged as IBM's services arm has grown to become the company's leading revenue-producing business unit.
"We saw something was happening in services" in 2004, when services accounted for almost half of IBM's business, Morris recalls. "Research was doing some work, but really not very much. Also, technology was not playing anywhere near as much of a role in our services business as we knew it could."
Today IBM's services engagements continue to account for more revenue than the company's software or hardware lines of business. In 2005, Global Services contributed US$47.4 billion, or 52 percent, of IBM's $91.1 billion annual revenue.
But while more than half of IBM's revenue comes from services, profit is another story. Software and hardware have much higher profit margins than services: In the most recent quarter, services had a 28 percent profit margin compared with 38 percent for hardware and a whopping 85 percent for software.
If IBM can find ways to be more efficient in services, the gains will bolster its bottom line. That's the thinking that led 16-year IBMer Morris to move to the services side of the house and learn the business. Two years later he's back in the research division and heading up services research.
Today a substantial portion of the research division is working on services-related issues such as security, problem determination and workforce management, Morris says. "It's very exciting. It's an incredibly open territory to bring technology to these problems. We are having an incredible impact on business."
Not every new research project winds up a candidate for bundling into other services engagements. But Morris and his team consider the broader applicability of projects such as a fire management solution IBM researchers are developing for state authorities in the western United States. IBM's mathematics experts created algorithms to help optimize the deployment of fire trucks, planes and firefighters as available road access and other conditions change.
"It's teaching us some things about the fast solving of very large mathematical problems," Morris says. IBM eventually will determine if it makes sense to use the algorithms to develop a disaster-management system for the public sector. "We've been working on disaster management for years in the context of data center, but we've never done it in the context of a forest. It broadens our horizons and opens our view to new classes of problems."
For researchers, tackling services projects requires a different style of work. Instead of working somewhat in isolation, in a lab or an office, researchers are much more involved with clients and spend considerable time in the field. "It's quite fascinating. It's a real rush to get into a client situation and learn what they need to create a competitive advantage," Morris says. "It requires a different mindset, but pretty much everyone who tries it gets hooked on it."
Part of what's addicting for researchers is the opportunity to get more timely feedback on their efforts, he says. In the past, a research project might take a couple of years to develop into a shipping product, and it might be another year or more before researchers really got information about how that product is being used by enterprise customers and might be improved.
"It was a very long feedback loop. Now it's an almost instantaneous feedback loop. We go out there, we brainstorm with our clients, and we come up with solutions on the spot. Because of the Web and because of software technology, and things like [service oriented architecture], we can implement those solutions very quickly," Morris says.
There's more immediate gratification, he says. "Once you get hooked on that, it's hard to go back to the three- or four-year gratification mode."