In my opinion, data science is a cross discipline, but with a clear focus.
As the name indicates, there needs to be data in it and science.
There are 3 major components and many minor components. The major components are:
Among the many minor components are:
Data science can often be structured in ‘projects’ or ‘stories’.
The start of any project are great questions in the respective domain. You need quite some domain knowledge to get to such questions or at least to understand available questions in sufficient depth. Further, there needs to be believable reason to answer the question with data and data analysis.
A desirable end of such a data science project is a decision. If there is no decision, a gain in insight is also a valid goal.
The domain focus distinguishes data science from pure statistics or computing. In those topics, KPIs can be computed, e.g. the quality of a fit or growth in CPU or memory consumption. However, it is a ‘third party’ who typically acts on such KPIs. Further, in data science - focusing on the science part - one answer leads to many more questions that are further pursued within the project.