Two months ago, IBM dropped the quiet bombshell that it would invest $1 billion over the next three years on research and development in services technology. At the time, this seemed like a drop in the bucket of IBM’s total planned $17 billion R&D spending over three years. But the announcement has generated heightened awareness among other services organizations of the importance of leveraging technology to reduce costs and improve service. And if IBM’s services technology R&D vision plays out, its cost could represent the company's best-ever investment of $1 billion.
Let's face it: Services have become the biggest piece of the enterprise IT pie. Far more dollars are spent on consultants, systems integrators, outsourced managed services, outsourced support, and application development than on all the packaged software that enterprises buy. People provide these services — living, breathing, mistake-making, coffee-break-taking human beings. Services vendors are striving to develop proprietary technologies to leverage those people, to make them more efficient, productive, and empowered.
Big Blue's billion-dollar investment
We recently spoke with Jarir Chaar, program director of services research at IBM’s 3,000-person R&D division. Chaar, who joined IBM 12 years ago right out of college, oversees the contributions of up to 1,300 researchers, many of them with doctorate degrees, working to develop technology that can support IBM’s huge Global Services organization, which now includes the former PricewaterhouseCoopers Consulting.
"Services is becoming the key piece of the IBM Corp., so you have to put your key guys and gals on it," he said.
Chaar explained that services technology R&D is about to enter a “hockey stick growth” phase and that the $1 billion would fund research in addition to the company's previous efforts. "Today it's a very labor-heavy organization,” Chaar said, regarding IBM Global Services. “Don’t expect this to disappear overnight … but we're giving our services organization more firepower to distinguish themselves in the marketplace."
That firepower includes a broad range of technologies in various stages of incubation and deployment, including some that are well-understood, such as systems management software, and some that are just percolating in the labs, such as advanced mathematical solutions to logistical problems.
Chaar outlined four broad areas where IBM has focused its services-enabling technology R&D effort: managed infrastructure, application development and management, e-business on demand, and consulting.
Chaar cited technologies within managed infrastructure to help IBM Global Services professionals efficiently configure networks; manage virtual private networks; manage storage services (especially SAN infrastructure security); automate service level management; and automate client software updates, data recovery and security, and systems management.
Within application development and management, Chaar singled out software for matching developers’ skills with project opportunities and for doing human resource planning globally, with an eye to maximizing quality and reducing engagement time. He also cited better software for performance modeling and analysis, and for testing and Q and A . "We really have found systematic ways of allowing you to decide which level you fit in when we talk about the Software Engineering Institute’s Capabilities Maturity Model (CMM) to objectively assess the maturity of an enterprise as well as propose remedies to get you to the next level,” he said.
Chaar noted two promising areas: specification-based testing and software to support legacy-code analysis. "How do you make sense of the glue that is legacy code?” he asked rhetorically. “How do you extract rules with some semantics out of it and componentize the code so you can reduce the complexity of understanding and managing it? If we are successful making sense of legacy and taming it, this will make our customers very happy."
On the ‘e-business on demand’ front, Chaar focused on technologies such as policy-based computing and resource virtualization. "There is no closed-form genetic solution,” he explained. “You have to understand the complexity and topology of the infrastructure that you're managing so you are able to institute the appropriate policies. The whole challenge of e-business on demand will keep us very gainfully employed for the next five to 10 years, so customers can be sure IT is put at their service rather than lurking as this monster.”
And finally, for the numerous consultants who are wrestling the monster, Chaar honed in on support, including e-learning, skills matching, and sexy technologies that help IBM’s top consultants get “permission to talk to C-level executives,” such as cutting-edge annotation technologies for life sciences applications.
A more human approach
Does all of this add up to real substance or just a very impressive sales pitch? To compare IBM’s effort with another large services firm, we talked with Brad Rucker, executive director of enterprise applications at EDS. Rucker manages 1,000 people and the development of all the technology supporting EDS’s huge services effort, including its AT Kearney consulting division.
Rucker described a research focus more geared to the human rather than the systems side of the equation, with an emphasis on tools for knowledge management, collaboration and communication, and training. He divided the world into "tools and processes for leveraging people and knowledge, increasing the speed of communication, and evaluating and improving your potential impact on the client.”
Rucker claimed much of the technology to support services is still nascent. "Most knowledge engineering firms don’t even realize that they need [workforce management software] yet, just like most people didn’t realize they needed word processors.”
EDS uses Evolve as a workforce management system, and employs client-facing and internal résumé systems, up to the executive level, to match expertise with projects. Because geographic proximity and passion are the two biggest multipliers of productivity and client impact for consultants, Rucker claimed, these systems are highly effective.
EDS built an internal system, dubbed TechLore, for knowledge management of code fragments, white papers, design docs, internal and external Web sites (“One of big learnings was not everything has to be held natively,” he said). But in the future, he predicts that "you'll start to see a suite mentality around knowledge management, workforce management, [and] communication and workflow tools," with disbursed project-type tools getting baked into workflow and communications tools such as Outlook.
Also on Rucker's radar are “service excellence dashboards” for measuring impact on clients, and tools to support the development, mentoring, and recognition of new technical talent and their associated intellectual property.
Rucker said that the tight economy has accelerated the development of technology to support services. "Services firms can't afford to do business the way they did when people were beating down their door,” he said. “Now you have to demonstrate value … . You actually have to show wood chopped."
But he also said there’s a tricky balance between encoding knowledge in software and support tools, and solving problems from scratch every time. "When is the value great enough to walk away from what you know? It's always a judgment call,” Rucker said. “Definitely when you have to mutate your known solution so much that it becomes a kludge. It's an issue of building judgment into systems.”
There’s also a dose of social engineering that’s key to making these systems work, Rucker claimed. Knowledge management systems only work if they enable individual contributors to get name recognition (a lesson he learned from the video game industry when Atari’s developers fled to Activision because they couldn’t get individual credits on Atari’s games). And EDS went so far as to give out cash bonuses for the best initial contributions to the knowledge database to kick-start the process.
Finally, free-spirited developers and consultants who resist being tethered to these systems must be coached on why the systems are worthwhile. "You have to walk through the theory of manufacturing and replaceable parts with [them],” Rucker explained. "We train them to focus their creativity on the flex points in an application,” such as UI and flow, rather than the rote. "Creativity does not include creating something that’s been created 300 times before."
Who wins in the end?
So the quest to develop technology that supports more efficient services delivery continues, but not without raising a handful of fundamental questions about what its impact will ultimately be. Can the individual pieces of technology, from systems management and development tools to knowledge management tools, be brought together to have a substantial coordinated impact on operations and efficiency? Or will they remain fragmented islands that developers and consultants work around or underutilize?
Will the promised productivity benefits, if they materialize, be passed on to customers or kept by the consultants? And do service providers and consultants in the IT world have the right financial incentives to adopt these technologies if they’re getting paid by the hour? Isn’t this really all about automation and replacing billable hours with software?
Stay tuned to see how IBM’s billion-dollar bet pays off. We’ll know in three years.
Enterprises should be aware that these enabling technologies are still highly fragmented and immature, and their economic benefits unproven. CTOs should push their consultants and service providers to adopt as much intelligent automation as possible, yet look for providers who are adept at getting the best of both worlds from software and human beings.