Pricing and Licensing Practices

Having the right products is critical to success in the enterprise server market, but Microsoft has also changed the way it prices and licenses those products to reflect traditional enterprise purchasing patterns and the practices of its enterprise data center competitors.

Until 2000, Microsoft’s licensing programs showed the company’s focus on workgroup and desktop systems, and on server applications used by desktop users on a corporate network. These licensing models worked fine for small businesses and workgroups, but they don’t account for how enterprises use server applications such as databases. An enterprise database that relies on batch updates might generate reports for thousands of individuals, but might have only a few actual “users” who access it directly, for example. As a result, Microsoft has moved toward more sophisticated licensing models, such as per-processor licensing and annual subscriptions, that better reflect enterprise preferences.

Per-Processor Licensing

Licensing applications per processor has been around since the early days of mainframes. But although Windows NT and its successors have always been capable of running on computers with more than one processor, multiprocessor PCs were relatively rare until the late 1990s, and Microsoft’s licensing models generally did not take the number of processors into consideration. Some applications had “enterprise editions” that could be used on servers with more CPUs (e.g., the Standard Editions of Windows NT and SQL Server 6.5 were limited to four processors, while the higher-priced Enterprise Editions could scale to eight), but generally the company’s licensing was based on the number of people who accessed an application, such as a database or e-mail, running on a server. Enterprises had to buy a Client Access License (CAL) for each person accessing the application.