Developers and testers are comfortable with the idea of receiving input from human users, checking it for validity, and producing output for human users. However, the human user is only part of the equation. Other, invisible software users also submit input and receive output but are less understood and sometimes forgotten altogether. But there is a price to pay for ignoring these invisible users. The first step in proper accounting of these users is to understand their nature and how they can cause software applications to fail.