What is the classification system in the software industry? In this project, I map the extensive categorization scheme of the software industry for the years between 1990 and 2002, using identity claims from software companies through their press releases. From the 268,963 press releases of software companies, I extract category claims of 4,835 software organizations. From this, I map category similarity (or overlap) based on organizations that claim to be members of multiple categories. Network plots illustrate the category maps over time. These show that classification systems in the software industry evolved rapidly, and that there is a great deal of overlap and connection among categories.

These plots also illustrate whether categories are high- or low-leniency. Category leniency is a measure that indicates how constraining, unambiguous, and well-defined categories are. High-leniency categories in this time period include enterprise, e-business, and platform. Low-leniency categories include accounting software, document automation, and digital imaging.

My research indicates that whether categories are high- or low-leniency affects new category creation, response to technical invention, customer appeal (in terms of revenues), and venture capital financing.