As software ages, it is increasingly unable to leverage new technologies and fulfill evolving user requirements. Firms producing software products therefore get an opportunity to introduce and sell upgrades. Our research looks at the optimal intervals between upgrades, and whether these intervals should increase or decrease over the product's life cycle. Although the accelerating pace of hardware and software developments (e.g. Moore's law) suggests decreasing intervals between upgrades, real business data for several commonly used commercial software products shows the opposite. To investigate this discrepancy, we set up model incorporating the costs and revenues from upgrades for a monopoly software producer. Our analysis confirms that the optimal upgrade intervals are monotonically increasing along the product's life cycle. Understanding of the optimal upgrade intervals allows managers to infer the upgrade costs involved and thus helps them to allocate budgets for managing these upgrades.