The simplest way to describe entropy is a measure of disorder in a system. A system with a high level of disorder is said to have high entropy.
A more software-relevant way to describe entropy is the additional information required to describe a software system. The task of writing software is antientropic i.e. software developers reduce the amount of entropy or reduce the amount of information required to interact with a system.
Great software has very little entropy: all users (whether consumptive or productive users) are able to use it by supplying only what is required with the software making informed guesses or having sensible values for auxiliary inputs. The quantity is entropy is tied to the software's feature scope. When a software cleanly covers a feature scope without partially overlapping other feature scopes and has low entropy then it will be maximally useful and be termed successful.