Calculating The IW 500

By Lou Bertin

nformationWeek determines the rankings of the IW 500 based on extensive and ongoing data collection and analysis by Computer Intelligence in La Jolla, Calif. The research firm compiles data on eight specific technology-usage categories for about 1,500 companies-including all the companies' sites-and calculates total scores. The top 500 companies are presented in charts starting on page 92. ( See IW500 menu )

CI assembles its rankings based on in formation provided by the companies themselves. In cases where only partial information is provided, CI constructs a model based on technology usage trends among similarly sized companies in the same vertical-market category. Companies that provided no information are not listed. (Significant IT users that didn't make the IW 500 are listed on page 286-- see PDF file here ).

The categories CI uses are: use of LANs and networks; leading-edge technologies; distributed processors; high-performance PCs; scientific and technical hardware; mainframes; communications equipment; and client-server technologies.

CI refers to these technology-usage rankings as hotpoints. They represent a comparative scale from 0 to 100 for each category. For example, a site with a PC score of 50 means it's ranked ahead of all other sites in the database with scores of less than 50, for either actual usage or propensity to use PC-related products and services. If a site has a score of 0, it has no apparent technology usage for that hotpoint technology.

CI used the following parameters to develop the hotpoint formulas for each technology segment:

Use of LANs and networks . Quantity of LANs installed; quantity of nodes; ratio of nodes to keyboards and printers; network control systems or software installed; number of installed bridges, routers, and hubs.

Leading-edge technologies . Total non-PC MIPS (million instructions per second); ratio of nodes to keyboards and printers; total direct access storage device (DASD) capacity; use of professional services, outside services, and consulting services; use of videoconferencing; number of 486, Pentium, and 68040 devices as a percentage of the total number of PCs installed.

Distributed processors . Number of remote locations; number of LAN and WAN connections; user access to Lotus Notes or other E-mail systems; number of remote PCs in use; number of remote midrange or larger systems.

High-performance PCs. Number of PCs installed at all sites; ratio of PCs to employees.

Scientific and technical hardware . Total number of high-performance workstations; ratio of workstation MIPS divided by total MIPS; number of Unix workstations installed; plans to add high-performance workstations.

Mainframes . Installed base; plans to add mainframes; non-PC terminals as a percentage of total PCs; ratio of number of terminals divided by employees; rewriting and updating of applications to remain on mainframes; number of non-PC MIPS and terminals, and use of non-PC database software.

Communications equipment . Quantity of analog and digital lines; type of phone system installed; type of phone system planned; number of T1 and/or T3 lines installed or planned; number of multiplexer (MUX) or data service unit (DSU) devices installed or planned.

Client-server technologies . Migration of accounting, application development, industry-specific, database, order-processing, and reporting management applica tions to LAN, high-performance workstation, or server platforms; number and type of critical applications running on LAN, high-performance workstation, or server platforms; plans to rewrite major applications to LAN, high-performance workstation, or server platforms; number and type of installed or planned systems identified as servers.

Overall index scores compiled by CI are the average of all technology-specific hotpoints. Although the hotpoint score is always between 0 and 100, the actual distribution of scores itself varies widely on any given sample. For example, a hospitality company would be unlikely to need scientific and technical computing systems, so a low score in that category would make sense. But an aircraft company's scoring zero in that same category would be suspect.

Therefore, when prioritizing a group of sites within an organization, it's not the absolute score that matters. Instead, the relative score among all ranked sites-and, specifically, all ranked sites in any given vertical market-is a truer measure.

Return to: " InformationWeek 500: IT In The Spotlight"

Back to the IW500 Menu

Send Us Your Feedback

Top of the Page