Hard SQL questions are the last step towards becoming a SQL expert. For this course, you don’t need to learn any new syntax. Instead, you need to have a lot of practice with what you learned in previous courses.

The main difference between medium and hard SQL questions is how straightforward they are. An intermediate question asks you to write a query that retrieves a specific set of data. Instead, hard questions require you to think what is the data you need in order to solve the question as well. Hard questions tend to be open-ended or ambiguous, as are most real-life situations for data analysts. When evaluating hard questions in job interviews, companies are also evaluating your communication and critical thinking skills.

Hard questions also tend to require more complex queries because they build on everything you learned previously. To get to this level, you need a good understanding of how databases work and have practice with concepts such as primary key, secondary key, or normalization.

We can think of hard questions as having two steps instead of just one:

1. How can you turn an ambiguous question into a precise definition of the specific data points you would need to retrieve in order to answer it?
2. How would you write a query that retrieves said data?

Some examples of hard SQL questions are:

• A group of tables show the results of an A/B test. How can we decide if it was successful?
• Given a group of tables that shows data from advertisements, sales and marketing metrics, perform a funnel analysis.

In summary, the SQL hard level includes understanding and visualizing the data you need to answer open-ended or ambiguous questions and being able to write complex queries to retrieve said data.

Getting to this level of proficiency with SQL will allow you to address any analytics questions that might arise in a business - and in an interview. Good luck!