Nothing changes faster than IT. With all the new technologies and technology trends on the horizon, how do CIOs decide which offer the best business value for their companies and how to retool their people in time to take advantage?
Below, two CIOs discuss their differing approaches, which reflect the particular challenges each faces.
PRC Inc., a McLean, Va.-based subsidiary of Northrop Grumman Corp., uses a process called Tech Watch to identify useful technologies, jump-start projects and re-skill people. An architectural lead team watches emerging technologies, maps them to the appropriate architectural groups and numerically rates them for impact on the business, cost to implement and readiness for deployment.
“The smaller the final number, the more likely it is we’ll be using it,” explained Cora Carmody, former CIO at PRC. (Carmody recently left PRC to become vice-president and CIO at Herndon, Va.-based Invensys Software Systems, a division of Invensys PLC.)
The top 10 on the current Tech Watch list – which is “subject to change at a moment’s notice,” Carmody cautioned – are Linux, XML, data analysis/decision support, enterprise portals, business-to-business e-commerce, text analysis, customer relationship management, workflow engines, wireless Web protocols and Digital Subscriber Lines/cable modems.
Once Tech Watch charts an emerging technology, anyone from the IT organization can volunteer to be an advocate for it. (During Carmody’s tenure, if no advocates came forward, she encouraged individuals to take the lead.) The advocate learns about the technology and the vendors and thinks about how it can make a difference to the company.
“We’ve got advocates at all levels,” she said. “We see it as another way to bring people along, give them an opportunity to explore something and push them further.”
Typically, the advocate finds others with an interest in the subject, and they begin to share information and build a technical knowledge base. The advocate may also attend seminars or buy inexpensive tools to learn more. In time, advocates may present a primer on the topic at a brown-bag lunch, an informal training opportunity that’s popular at PRC.
They typically propose a pilot program to a manager in one of the architecture groups to which the technology maps. He can choose to become part of the pilot project, thereby working into a new job in the new technology area. The manager may send them, along with high-potential people in his department, to formal training in the technology in preparation for the project.
Using this approach, Carmody’s group recently implemented business-to-business e-commerce software from Mountain View, Calif.-based Ariba Inc. and realized the tool could also help with internal electronic purchasing. The group used the Ariba workflow engine to automate internal authorization and purchase processes via e-mail.
The new system allowed PRC to cut its internal procurement cycle from 16 days to less than two while decreasing procurement staff. “Think about the savings,” Carmody said.
There are exceptions to the advocacy process when it involves a technology that will have a huge impact. “For something like PeopleSoft, we’re not talking little seminars. We’d need some very serious training and investment,” Carmody said. In that case, she would choose people based on past performance and familiarity with the business functions involved, such as human resources and accounting for an enterprise resource planning system like the one from Pleasanton, Calif.-based PeopleSoft Inc.
In general, Tech Watch keeps the company on top of emerging technologies while enabling people to navigate their careers based on interests and initiative. “Everybody is challenged to take responsibility for his or her career and is given a lot of encouragement,” Carmody said.
Everything Old Is New Again
At State Street Corp. in Boston, CIO John Fiore is in the midst of an initiative to add contemporary software skills to the repertoire of his mainframe staffers while upgrading his mainframes with powerful new open-systems tools. Fiore said the mainframe/open-systems integration should give the bank the best of both worlds: new tools and languages adapted for stable, scalable mainframe systems. In addition, the move should enable legacy developers to pick up new skills while working on a platform they’re familiar with.
During the past five years, the bank has added browser-based technology to the front end of its systems. Fiore chose mainframe staffers to help, based on their performance and subject-matter expertise in the area under construction. He then immersed them in a comprehensive in-house curriculum ranging from object-oriented tools to Internet technologies. After that, the staffers were seeded to work among new hires chosen for their open-systems expertise.
Although the plan has worked pretty well, Fiore acknowledges that certain delays and initial performance problems with the new front ends could have been avoided had he given people more time to learn the skills and apply them on smaller, less-critical systems to go back for retraining and only then to work on the critical applications.
“Of course, that’s easy to say, but time is not always a luxury we have,” he said.
Those lessons will be put to use as the bank gradually integrates languages such as Java and C, relational database technologies, TCP/IP and Web server software, which used to be available only in the open-systems environment, into its OS/390 platform.
The initiative began with a series of feasibility studies to assess when the new technologies were “ready for prime time.” TCP/IP, for example, initially had some performance problems running on the mainframe.
As technologies mature, Fiore plans to integrate them into the mainframe systems to dovetail with other activities. For example, if a major new component is being built for a system, he may introduce just-in-time training for Java or C tools to help with that development.
Since new technologies will be introduced at different points based on their readiness for use and internal development plans, the staff should have time to digest the new tools and concepts one at a time, without suffering the overload that slowed the front-end systems development, he said.