bia_interview_questions

Questions to ask as a data scientist

Question 1
"Can you share an example of a time the leadership team made a decision based on analytics?"

Why ask question 1
Hopefully the importance of this question is clear.  As an analyst your ability to provide value and grow is heavily dependent upon leadership 
making quantitative decisions.  While a highly analysis-aligned culture is extremely rare, your prospective employer should at least have one example
of leadership makign decisions based on an analysis.

Question 2
"Is the analytics team centralized or federated?"

Why ask question 2
There are pros and cons to both organizational structures, but some work by Gartner around 2014 showed analysts to be happier and more impactful in centralized teams.
If you are the analysts for a particular team (e.g. a marketing analyst within marketing) you may feel isolated from your peers, have few chances to learn from other analysts, 
and/or possibly feel pressured to make your department look good.  A centralized team of analysts can address these problems.
tl;dr - If you can choose, I'd recommend looking to join a centralized team.

Question 3
"How are requests for analytics work processed?  Can anyone submit a ticket or something similar?"

Why ask question 3
You want to understand the process for an employee to ask you for work.  If any employee can ask for a dashboard at any time you will be swamped with low-impact busy work.
Ideally, access to analysts such as yourself is reserved for a select group of decision makers with whom you will partner closely. 

Question 4
"On a spectrum of 100% ad-hoc requests to 100% premeditated where does most of the analytics work reside?"

Why ask question 4
This question is trying to get at something really really important. Essentially, you are trying to determine if analytics is just used as data gophers, 
fetching data for decision-makers and then being dismissed, or if they are being used as actual analysts and thought partners.
In my experience, analytics is most useful to a company and more rewarding to the analyst when the analysis is premeditated.
Premeditated simply means that there is a line of inquiry that the company wants to explore with several hypotheses to test.
The analyst is then made aware of what the company is trying to decide and is allowed to use their analytics skills to help inform that decision.
An analytics team that does a ton of ad-hoc work is typically not investigating targeted lines of inquiry to explore important questions; they're just fetching data for people.

Question 5
"Can you share the employee satisfaction scores for the analytics team for the past 2 or 3 quarters?"

Why ask question 5
This question might get you some weird looks, but in my opinion, any company that cares about employees at all should have some measure of employee satisfaction. 
If they aren’t measuring this, how do they plan to even improve or at least maintain it? 
If they do measure it but can’t reveal that information in an interview, that is reasonable and hard to argue with.  
In that case, just ask follow-ups like “How often do you measure employee satisfaction?”