By Robert Tasker
Over the past two years or so, we at the Yankee Group have been involved in a series of surveys, personal interviews and programs designed to identify the human and technological boundaries that are likely to shape corporate information technology (IT) organizations over the balance of this decade. I believe that the ultimate form and function of the most successful IT departments will follow and support the form, function, objectives and requirements of the corporation as a whole.
The results of our work to date have been both expected and surprising. Here are a few of the more remarkable trends and directions.
Organizationally, the results of our surveys and other research indicate that a trend will continue toward flatter, more loosely defined organizational structures, occurring first in the functional departments outside of IT and then within it. (Could it be that senior management, in its infinite wisdom, was in the early years of the '90s simply preparing IT for the inevitable advent of the "thin computing client?" Naah!) A direct result of this trend will be evermore vaguely defined job descriptions, the eradication of many if not most job titles, and the institution of amorphous "workgroups."
Going forward, the more successful (that is, surviving) IT organizations will display a number of distinguishing characteristics. Almost certainly the organization will be smaller in number of personnel, and available personnel will be responsible for an ever-widening range of duties and functions. Flexibility will be rewarded.
The trend toward outsourcing of IT functions will continue. The reasons are both simple and compelling. Large outsourcing providers, such as EDS, IBM and CSC, and some smaller high-quality providers, such as Keane Associates and PKS, Inc., will always have better-trained staff and more of them than the average company's IT department. The best and brightest trained in IT skills will seek employment with those companies for which IT is a core business. Moreover, outsourcing--when evaluated without emotion and political overtones--frequently makes strong economic sense.
A quickly emerging ancillary trend is that of outsourcing non-IT-driven business functions. More frequently than ever, I am observing corporations outsourcing what have traditionally been in-house business functions such as purchasing, human resource administration and even sales staffing. The major new twist here will be that the corporate CIO or other senior IT management will be charged with ensuring that the IT function provided by the process outsourcer is on target regarding service level agreements (SLAs), and that proper, cost-effective technological platforms are maintained for the outsourced functions.
On the technical front, the so-called thin client (in effect a stripped-down PC or workstation) will find strong acceptance, as any valid cost-saving concept should, but acceptance will not be universal. Functionally, this strikes many observers as a throwback to the days of green screens and 3270 bit-streams, but with an attention-grabbing GUI added. I don't believe this view is correct. With the support of "fat servers," download ability over the Internet and the utility of platform-neutral tools such as Java applets, appropriately configured thin clients should prove to be very cost-effective. There are a great many PCs and workstations currently configured with much more software functionality than the users of those machines will ever require. The purchase and installation of these systems cost hundreds if not thousands of dollars, and they probably will never produce a reasonable return on investment (ROI) to the corporation.
(As an aside, it is becoming obvious that the real winners in many of the fat client configurations are the PC software companies. They have generated a seemingly endless flow of revenue from their equally endless upgrades. Seriously, how many administrators or clerks in a corporation actually need PowerPoint on their PC?)
Speaking of Microsoft, while the various configurations of the Windows NT operating environment have been referred to as the "great leveler," the unthinking acceptance of this product by many IT organizations strikes me as resembling a lemming migration to the sea. Significant aspects of the NT environment seem to go unchallenged, such as single points of failure in the architecture, inefficiency of directory structures and the ability of Microsoft to adequately support installations of tens of thousands of seats. While the various flavors of Unix offered by vendors are not perfect, mindlessly replacing Unix with NT does not strike me as the right answer. However, now more than ever, marketing masters all: I expect Unix to fork over roughly 10 market-share points to NT in 1997.
Fiduciary responsibility will be the mission of the well-managed IT department over the balance of the decade. IT will be more focused on the requirement of demonstrating an acceptable ROI-or some other alternative economic measure, such as the currently popular metric, economic value-add (EVA)-for all future investments in technology, whether hardware, software or services. The more successful corporations have been enforcing rigorous ROI (and other financial) measures prior to IT expenditures for some time; now the idea is gaining wider adherence. Financial analysis will take precedence over selecting the next whiz-bang operating system or hardware platform, and we should all applaud the change.
Integration of heterogeneous systems is the other holy grail. The predominant operational tasks for the IT department will include the following: smooth, seamless integration of heterogeneous, enterprise-wide devices; installation of data mining capability coupled closely with rapid development of applications for the purpose of identifying and seizing target markets; and, perhaps most important, demonstrating to senior corporate management a rapid, tangible impact for substantial improvement in the production of corporate products and services. It is going to be an interesting run to the end of the century.
Robert Tasker is senior vice president for computing research and consulting for the Yankee Group in Boston.
Did something in this column press one of your hot buttons? Then let us hear what you think by sending a response to pubs@uniforum.org. We'll consider it for publication in "Letters to the Editor."