There are many ways that Data Science could add value. In theory.
In practice, I’ve found that there is one way for Data Science to add value, and that way is in supporting a decision making process.
Note that supporting a decision making process is not the same as “making the decision” — ultimately this must rest with a human that is accountable for the decision. This may mean that if you use software to “make decisions,” you are still responsible and accountable as you decided to employ the software as your proxy.
The real opportunity arises when a decision has uncertainty (or ambiguity) and some “data science” — read system and associated analyses — can be applied to 1) quantify the uncertainty, and 2) reduce the uncertainty for the decision. Using information to reduce uncertainty adds value that can be quantified.
What else does Data Science do?
-is it a workflow, or pipeline or database architecture? Maybe in part, but only if it facilitates decision making.
-knowledge management? Maybe in part, but only if it enables decisions based on that knowledge.
-improve understanding? Perhaps, but how do you translate that improved understanding into value? –only by your subsequent actions, i.e., decisions.
In the end, whether you invest in Data Science should really depend on whether you have decisions where you want to reduce the uncertainty associated with different actions or choices.