Running the Numbers for Windows 2000's

Recent GartnerGroup and Giga Information Group studies about Windows 2000's (Win2K's) total cost of ownership (TCO) have caused an uproar. GartnerGroup says Win2K is too expensive; Giga says Win2K is worth the cost. The reasons behind these disparate claims lie in the studies.

We carefully reviewed both reports and spoke with the analysts involved in the details of the studies. A company will pay $257 per seat to license Windows 2000 Professional (Win2K Pro), yet this cost is the smallest part of Win2K Pro's TCO. When you add the cost of hardware upgrades, training, systems administrator hours, and programmer hours to recode custom applications, the total cost for an enterprise to migrate to Win2K Pro can be hefty.

So how hefty is the cost of the migration? GartnerGroup estimates the cost of migrations from Windows 9x to Win2K Pro to be between $2015 and $3191 per desktop; Giga projects the cost to be $973 per system. GartnerGroup used a $1700 system to project the costs; Giga used a $1000 system. GartnerGroup estimates labor expenses to be between $92 and $407 per desktop; Giga estimates that a company can (in an extreme case) lose 2 administrator hours per desktop system and estimates $140 per administrator-hour—a cost of $280 per desktop in an extreme case. (Many of the labor costs are in testing and developing custom applications.)

TCO is only one part of a two-part business decision to migrate. A business also needs to look at the benefits of migrating. When you consider costs and benefits, the GartnerGroup study falls short because the study offers pages of cost analysis but almost no benefit analysis. GartnerGroup maintains that most of Win2K Pro's benefits are hard to quantify and won't show up in an IT budget. Because IT department budgets don't typically include increased productivity of computer users, GartnerGroup doesn't factor in increased productivity in its study.

However, although a number might be hard to quantify, you should still estimate the benefits of migrating before you make your decision. Giga considers benefits in its study and projects that the stability (i.e., fewer reboots) and increased power of Win2K Pro will increase productivity by 3 to 5 percent. Assuming that the average employee salary is $70,000, a 3 to 5 percent productivity increase would save a company $2100 to $3500 per employee. Giga also estimates that each dollar a company saves on labor is worth $5 of revenue. Thus, the value of Win2K Pro's benefits would fall between $10,500 and $17,500 annually per employee.

You can consider Giga's $2100-per-desktop estimate on savings for a Win2K Pro migration. This figure is almost equal to GartnerGroup's lowest cost estimate for the migration: $2015. The higher estimates are also nearly equal. Giga's estimates show that the benefits of a Win2K Pro migration will most likely provide a Return on Investment (ROI). As analysts, we favor migrating because of the benefits, which Giga's study shows. Win2K Pro is a compelling, straightforward upgrade to a faster and more stable platform than Win9x.

When you consider Windows 2000 Server (Win2K Server), the questions become more complicated—and more necessary for administrators to answer—because Win2K Server isn't a one-size-fits-all solution. Although GartnerGroup and Giga have opposing Win2K Pro conclusions, the two firms' analyses of Win2K Server are nearly identical. Both groups recognize the complexity of the analysis, which includes the large number of variables involved with installing Win2K Server and the infrastructure, expertise, and diverse needs of each organization.

The major difference between the GartnerGroup and Giga Win2K Server TCO studies centers on an organization's policy-management capability. GartnerGroup takes the more conservative position. Mike Silver, a GartnerGroup analyst, argues that the majority of companies might fail to implement the policy-management features of Win2K Server, and thus fail to realize the major economic benefits that Win2K Server offers. GartnerGroup's report factors in a high percentage of possibility for policy-management failure, which dramatically reduces the economic benefit estimate for Win2K Server. Silver said that if a company implemented policy management, the company should realize a good ROI. However, Silver said a company needs to be cautious: If the company doesn't have policy management already in place, the company probably has a reason, and that reason can prevent a company from implementing policy management in a Win2K Server migration.

Giga views policy management as the basis of Win2K's ROI power. Giga asserts that insufficient expertise will result in unsuccessful policy management; therefore, Giga urges its companies to train employees sufficiently. Both GartnerGroup and Giga also emphasize a substantial training program for users.

Overall, the studies offer similar conclusions with regard to a Win2K Server migration. Don't migrate unless you have a sound business reason (e.g., to gain functionality), and use third-party management tools when you migrate. Finally, both GartnerGroup and Giga recommend that you wait until Microsoft removes the kinks from Win2K before you migrate.