IBM’s Internet start-up: alphaWorks

December 23, 1998 — How does an 0 billion company retool its product development pipeline to meet the requirements of the fast-paced Internet market? If dedicating all nine full-time staffers to the task doesn’t seem adequate, the tactic nonetheless seems to have paid off for IBM.

Keeping with IBM chairman, Lou Gerstner’s directive to bring Big Blue’s immense research and development operation closer to the realities of the marketplace, alphaWorks – a cross-functional team of technologists and product managers — has shepherded numerous technologies from the lab to the hands of paying customers and licensees.

The alphaWorks 9 — just enough for a departmental baseball team — resembles a Silicon Valley software start-up by design. Steering IBM’s efforts in such cutting-edge sectors as Java, XML, Internet media, and the extended network of mobile and embedded systems demands a nimble approach.

The roughly 2-year-old group exerts its leverage through an extreme form of matrix management within IBM, and efficacious use of the Web in dealing with developers and potential customers. For example, the group managed to streamline the process for naming product and technologies within IBM.

The group’s well-received Web site (www.alphaworks.ibm.com) functions as a showcase for emerging technologies and as a developer community forum of sorts. AlphaWorks also distributes finished products such as the Jikes Java compiler, which was recently released as open-source code. Various IBM product divisions, involved in the development process, may also incorporate technology.

Headquartered at IBM’s Almaden Research Facility in San Jose, CA, alphaWorks speeds the transition of technologies from research projects to products, while injecting reality-checks and market testing early in the process.

“Sometimes the scientists haven’t asked the big [business] question. AlphaWorks is a bridge to make the technology a reality,” said Chris Bahr, alphaWorks manager. “We can also wed internal resources with the extended developer community.”

The group, which uses consultants when needed, and relies on interns to extend its reach, solicits technical proposals from IBM’s labs and works with researchers, product development teams, and marketing staff to screen ideas, compose business briefs, and generally market-test ideas at an early stage, explained Bahr.

For scientists and product teams, working with alphaWorks is voluntary, Bahr noted. “We’re [only] in a position to advocate,” he said.

For many researchers, however, the process of making business cases for technology can be eye-opening, according to John Wolpert, director of emerging technology development.

“It’s almost like business school for computer science students,” Wolpert said.

Beyond its current areas of focus — Java, XML, Pervasive Computing, and Internet Media — alphaWorks is also charged with identifying the next killer technology, according to Wolpert.

The group’s chances for success depend in part on its ability to secure more resources for the coming year and beyond. In little danger of growing too big too fast, alphaWorks is like many of its start-up cousins in lobbying its backers for more money and more bodies.

Goals for 1999 also include a significant increase in the number of proposals from research, technologies posted on the alphaWorks Web site, products brought to market, and requests from product groups.

To date, the group’s efforts have centered on software, which is fertile ground, in that it consumes more than half of IBM’s research funds, according to Bahr. Of IBM research’s roughly .5 billion annual budget, about billion goes toward software research, he noted.

IBM, the home of blue-suited technocrats and bureaucrats, seems to have shed enough of its reputedly plodding ways of old to establish an effective scheme for funneling some of its high-voltage technical innovations into the market.

The alphaWorks approach, although appealing to the more free-wheeling segments of the IBM customer base and developer community, probably won’t displace or seriously alter many established product development methods, according to Ann Thomas, an analyst at the Patricia Seybold Group, a market research company in Boston.

“For example, the IBM software group follows a strict process of continually going out to customers and getting their requirements, developing solutions, rigorously testing them, polling partners, IBM sales and others, and repeating the process again and again. It’s not just a matter of posting cool and interesting stuff,” Thomas said.

Such firmly entrenched practices cast doubt on whether alphaWorks’ Internet start-up ways will be widely emulated, according to Thomas.

“I’m not convinced that more groups within IBM are going to be as shoot-from-the hip,” she said. “[AlphaWorks] has basically been given a start-up charter, so they don’t have to be as concerned about code quality. But IBM as a company is one of the most, if not the most, concerned with code quality in the world.”

Still, the group is likely to continue to be an important catalyst for IBM’s efforts in Internet-related markets, according to Thomas.

“AlphaWorks will to do a ton of work in Java and XML, and they’ll probably start doing stuff in streaming media and security,” Thomas said. “I’m sure that [the group’s backing] goes all the way to the top. It’s a great way to see if there’s a market for research technologies.”