What do experienced data scientist know?

Question

As a data scientist, what tips would you have for a younger version of yourself?

Answer

1. "Honey badger doesn't care": In majority of the cases, your business stakeholders won't really care on how you did it. You really have to distill the insight and explain the insight in laymen's terms.
2. "Be good to the future you": On the other hand, people who will be maintaining or taking over your code do care how you did it. More often than not, that sucker may be the future you. So leave plenty of comments and documentations for the future you.
3. "People often don't know (exactly) what they want". Ok, it doesn't quite get to Steve Jobs' extreme, but you will often have to help your stakeholder refining the questions, often over many iterations. The stakeholder often ask questions that has the right big direction, but may not be specific enough for you to "hit the ground running". For example, when you were asked to "build a model", you might need to first ask (1) what business problem are we trying to solve? and (2) is building the model the highest-leverage way to solve this particular problem?

As painful as it might be to admit, some business questions may not have data science solutions, and the faster you can find whether a data science solution is worth pursuing, the better.

(Originally answered on quora: What do experienced data scientists know that beginner data scientists don't know?)