The third family of data science questions tests your ability to apply your technical data science competency to real-world problems, especially in the arena of business case questions.

In business-oriented questions, interviewers evaluate your communication skills and your ability to structure an answer. Because of this, there are several useful frameworks for answering these kind of questions that will help structure your thought processes into a solid answer.

Let’s look at each kind of question in more detail:

Data Science Behavioral Questions

Behavioral questions seek to evaluate how well you would do on the job by testing skills that go beyond the technical side of things. The main topics you’ll be tested on are:

• Technical communication
• Culture fit
• Adaptability
• Technical competency
• Past experience

Behavioral questions investigate your previous experiences or how you approach diverse problems. Some examples are:

• Give an example of an analysis that you did that drove business impact.
• How do you make technical topics accessible to non-technical audiences?
• Tell me about a data science project you have worked on. What challenges did you experience? How did you respond?

The best answers in behavioral interviews are like stories. They are framed from beginning to end and include plenty of interesting detail. Your goal will be to satisfy your interviewer’s question and also to provide material for them to ask further questions in a way that shines you in good light.

Product Metrics & Analytics

Product metrics and analytics questions assess your product intuition and ability to make data-driven product decisions. Often, the interviewer will also be specifically evaluating how well you understand the company’s main product offerings, so these questions require special preparation for each company you’re interviewing at.

They say data scientists function as an intersection between product managers, analysts, and software engineers. A product metric and analytics question evaluates your capacity for exactly that kind of intersectional work.

Interviews in big tech companies such as Meta and Google focus mainly on this kind of inquiry. They want to understand if you know how to contribute to building products. This can be a very challenging situation to perform well in, given how unnatural it feels to understand and answer a business case question in under an hour.

As a framework for answering, you can always turn back to a typical product development life-cycle:

1. The product manager asks the data specialist for product insight.
2. The two collaborate on building this feature based on data.
3. They build, test, and launch the feature.
4. They perform monitoring and evaluation in order to improve and analyze the successes or failures of the product.

Depending on the role, product interviews will come up at different frequencies. For example, in machine learning roles, product interview questions will almost never come up. But if you are interviewing for a product role, you’ll likely encounter even more than one product interview during the hiring process.

Product questions may come in different forms:

Investigating metrics questions ask you to provide insights about changes in relevant metrics. For example:

• Why are Facebook friend requests dropping by 10 percent?

Measuring success questions ask you to provide metrics for evaluating how well a feature performs. For example:

• How would you measure the success of the Facebook marketplace?  or How would you measure the success of Yelp reviews?
• Let’s say we want to add/change/improve a new feature to product X. What metrics would you track to make sure any changes have a positive impact?

Data-driven decision questions ask you to make product development decisions based on data. It’s likely that you need to define the metrics you’ll use for your strategy.

• If you are the product manager for Facebook and you see that comments are down by 10%, yet reactions are up by 15%, how would you deal with it?
• We want to grow x metric on y feature; how would we do that?

These questions may also come in the form of case studies, which we’ll look at in the next section.

Business Case Questions

Business case questions test your problem-solving skills by providing you with a business scenario and a problem at hand and asking you to offer solutions. They’ll sometimes provide you with a dataset to investigate before you offer your solutions.

These case questions seek to understand if you can leverage data in order to solve business problems. For example:

• A bank wants to create a new partner card, e.g. Whole Foods Chase credit card). How would you determine what the next partner card should be?
• How would you assess the value of keeping a TV show on a streaming platform like Netflix?

What is difficult with these questions is that there are no clear right or wrong answers. Instead, case study interviews require you to come up with a hypothesis and then produce data to support or validate your hypothesis. In other words, it’s not just about just your technical skills; they also test you on creative problem-solving and your ability to communicate with stakeholders.

For these questions, sometimes you’re not even required to write any code; just make the decision on which code you would need to write - and be able to describe it precisely.

In the next section, we’ll dive deeper into how to prepare for data science interviews.