Music has the power to transport us to another time and place. With the rise of music streaming services like Spotify, we now have access to a vast library of music from different eras. In this analysis, I will be looking at the top 2000 songs of all time on Spotify from 1956-2019, aiming to identify trends in music over time and explore the characteristics of the most popular songs.
To collect the data for this analysis, we obtained the Top 2000 Song of All Time dataset from Kaggle, which includes information about the most popular songs on Spotify from 1956 to 2019. The dataset contains over 170,000 entries and includes features such as song title, artist name, release year, and various audio features like danceability, energy, and loudness. After downloading the dataset, we imported it into Google Sheets for data cleaning. We addressed missing values, removed duplicates, and formatted the data for consistency. Finally, we saved the cleaned data as a CSV
file to be imported into SQL for
further exploration and analysis.
After importing the CSV file into the database, I used SQL Azure Data Studio to explore the top 2000 songs on Spotify. I began by categorizing the sub-genres into eight main genres: Pop, Hip-Hop, Rock, Folk/Country, Electronic/Dance, Reggae/Funk, R&B/Soul, and Other.
I then conducted a thorough genre analysis, using SQL queries to identify the number of songs in each genre, the number of songs in each genre each year, and the audio statistics for each genre. I found that Rock was the most common genre in the dataset, with over half of the top 2000 songs, while Electronic/Dance was the most danceable genre. I also analyzed the most popular genre each year, and found that R&B/Soul and Hip-hop was the most popular genre in most years, with notable exceptions in the mid-2010s when Hip-Hop and R&B/Soul were more popular.file to be imported into SQL for
further exploration and analysis.
Data Collection and Cleaning
Introduction
Exploring Our Top 2000 Songs
Top 2000 Songs of All Time on Spotify
Moving on to artist exploration, I identified the top 10 artists with the most songs in the top 2000, and found that Drake had the most songs with 22. I also analyzed the most popular artist per year, and found that Ed Sheeran was the most popular artist in multiple years.
Finally, I delved into song exploration, analyzing the number of songs released each year and identifying the top 10 most popular songs, the most danceable songs, and the songs with the most energy and loudness. Overall, my data exploration in SQL Azure Data Studio provided valuable insights into the top 2000 songs on Spotify, and helped me identify trends and patterns in the data.
Dashboard Creation
After exploring the top 2000 songs on Spotify in SQL Azure Data Studio, I used Tableau to create a dashboard that visualized my findings. I began by planning out the dashboard sheets I needed, including a home page, and renaming them accordingly. I then reviewed the queries I had written to determine the data I needed for each visualization. I then turn those tables from the query and turn it into CSV files to be used for Tableau.
For genre analysis, I used the genre audio statistics as a whole and a yearly table pulled from SQL to create visuals for the most danceable, popular, and energetic genres. I also created a timeline for BPM and which genres use it the most, and the most popular genre each year. For each genre, I created analyses showing the number of songs in that genre, the most popular song, and their audio statistic averages in the form of gauges. I also included a timeline chart that shows the yearly audio statistic trend and a list of the sub-genres in the filtered genre.
​
​
​
​
​
​
​
​
​
​
​
​
​
For artist analysis, I used queries that showed the most popular artist each year, artist audio statistics, artist audio statistics each year pulled the tables as a CSV file, uploaded it into Tableau and created a bubble chart that shows the top 10 artists in the top 2000, as well as charts that show the most danceable, speechy, happy, and sad artists based on valence scores. I also included a timeline of the most popular artist each year.
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
​
For song analysis, I used queries that showed all the songs in the top 2000 and their audio statistics and genre. From that query, I created visuals such as a song count timeline per year, a scatter plot for the most danceable and speechy songs, the most energetic and loudest songs, and a chart that shows the top 3 most popular songs. I also included a track list and created a separate page for it. Finally, I put all these visuals in their respective dashboard pages and used the best visual for the home page.
For the last couple of details of the dashboard, I created navigation buttons for each dashboard page to make it easy to navigate and adjusted formats and layouts such as fonts, sizing, and placement, coloring, and organization to tie it all together. Here is the finished dashboard below.
Results and Analysis
Through exploration of the top 2000 songs on Spotify using SQL Azure Data Studio and visualization in Tableau, we have gained insights into the audio characteristics, genre, artist, and trends of the most popular music on the platform.
​
One of the key findings of our analysis is that the most popular genre among the top 2000 songs is Rock with over half of the top 2000 songs on the playlist. But the genre with the most popular songs is R&B/Soul followed by hip-hop/rap.
​
Additionally, our analysis revealed interesting audio characteristics of the top 2000 songs, such as the trend towards higher energy and danceability scores in recent years, and the valence of songs with the highest and lowest valence scores. We also found that the most popular songs tend to have higher loudness and speechiness scores, indicating a preference for catchy and vocal-heavy tracks.
​
Upon delving deeper into the top artists on the platform, it was observed that a majority of the artists with a danceability score higher than 90 also have an average speech rating ranging from 5-10, except for some notable artists such as Daft Punk, Bob Marley, and Eminem.
​
Overall, our analysis provides a comprehensive overview of the top 2000 songs of all time on Spotify, offering insights into the current trends and characteristics of popular music, as well as highlighting emerging artists and genres to watch.