1. Introduction to data driven decision making
Welcome to this course on data driven decision making. My name is Irene Ortner and together with Profesor Bart Baesens and Professor Tim Verdonck we will guide you through this course.
2. Aim of this course
We assume that you are already familiar with basic SQL queries.
Still, we will quickly refresh some basic SQL statements
and you will learn how you can apply them to extract valuable business insights from your data.
Then you will learn about new SQL techniques to summarize data such as the SQL OLAP extensions the CUBE, ROLLUP and GROUPING SETS operators which are specifically developed as business intelligence tools.
3. MovieNow: an online movie rental company
Throughout this course, we will work with a Postgres database from a fictional movie rental company called MovieNow.
MovieNow offers an online platform for streaming movies. Customers can rent a movie for 24 hours.
For all movies, the company stores additional information such as the genre or the main actors.
MovieNow also stores information about customers
and movie ratings.
4. MovieNow data structure
Here we will give an overview of the tables in the database.
First, in the 'customers' table, we have a column 'customer_id', a number which is a unique identifier for each customer, then we have name, country, gender and date of birth. The final column is the date when the account for MovieNow was created.
5. MovieNow data structure
The columns for the table 'movies' include a unique identifier movie_id, the title of the movie, the movie genre, the runtime, the release year, and, finally, what it costs to rent the movie.
6. MovieNow data structure
The table 'renting' records all movie rentals. 'renting_id' is a unique identifier for each movie rental. The column 'customer_id' tells us which customer rented the movie and 'movie_id' tells us which movie the customer rented. The rating a customer gives after watching the movie is stored in the column 'rating' which has values between 1 and 10, where 10 is the best rating. The final column is the rental date.
7. MovieNow data structure
The table 'actors' contains information about the actors in the movies. Besides the unique identifier 'actor_id', we have the actor's name, year of birth, nationality, and gender.
8. MovieNow data structure
The table 'actsin' shows which actor appears in which movie. Besides the unique identifier actsin_id, it includes movie_id and actor_id.
9. Objectives of data driven decision making
Data-driven decision making achieves short-term and long-term goals.
We can extract valuable information from data to support operational short-term decisions. For example, the popularity of certain actors helps MovieNow decide whether to purchase certain movies. Also, last month's revenue can be important information supporting decisions regarding short-term investments.
For long-term decisions, data-driven support can provide information regarding customer growth and successes in certain regions in the past, which can inform company decisions regarding when and where the market can be expanded in the future. Also, knowing the long-term development of overall revenue helps MovieNow plan long-term investments.
10. KPIs: Key Performance Indicators
Key performance indicators help a company (or its subdivisions) define and monitor success.
Revenue is a trivial indicator of success. For MovieNow, this is calculated as the sum of the price for rented movies.
For the subdivision customer relations management, the KPI 'customer satisfaction' could be quantified by the average rating of all movies or the KPI 'customer engagement' as the number of active customers in a certain time period.
In this course, you will learn how to define KPIs, extract information from the data, and monitor it over time. This will help decision making using hard facts.
11. Let's get started!
Now you can start exploring the database yourself.