Exploratory Data Analysis of Spotify Tracks
Insights from audio features and popularity metrics
Date - 23rd July , 2025
Presented by
Analytics Avengers Next Slide
Objective
Understand the structure of
the Spotify dataset
Identify relationships between
variables (e.g., popularity vs.
danceability)
Detect missing or unusual
data
Visualize key musical features
Next Slide
Presentation Agenda
Dataset Data Data Overall
Info Visualizations Insights Conclusion
Next Slide
Dataset
Brief Overview
This dataset contains detailed information about a collection
Overview
of Spotify tracks, capturing both musical and metadata
features. It offers valuable insights into song characteristics,
artist popularity, and audio attributes, making it ideal for
exploratory analysis.
Purpose of The Presentation
Understanding the dataset lays the foundation for effective
feature engineering and predictive modeling
Next Slide
Univariate Analysis of numerical variables
Next Slide
🧪 DISTRIBUTION OF KEY AUDIO FEATURES
🎧 Features Visualized:
Popularity: Most tracks have low popularity scores, with few achieving high engagement.
Acousticness: Majority of songs are low in acousticness, suggesting a preference for electronic or produced sounds.
Danceability: Values cluster around the mid-range, indicating moderate rhythmic suitability.
Instrumentalness: Most tracks contain vocals, with instrumental pieces being rare. Next Slide
Liveness: Low values dominate, implying few live-recorded tracks.
Loudness: Tracks tend to hover near average loudness levels, with fewer extremes.
Outlier Detection and Range of Audio Features
“Popularity is highly skewed with
many tracks having low scores.
Danceability and acousticness show
moderate variation and few outliers.
Instrumentalness and loudness
show extreme outliers; these may
need further investigation or
cleaning.
Box plots help identify variability,
central tendency, and potential data
quality issues.
Next Slide
📊 Popularity Distribution of Spotify Tracks
Overwhelming concentration at popularity 0 suggests a large volume of
underplayed or niche tracks.
Very few songs reach high popularity scores (above 60), highlighting the
competitive nature of viral success.
The distribution is heavily right-skewed, with long tail behavior. Next Slide
Univariate Analysis of categorical variables
Next Slide
🎤 Top Artists in the Dataset
🎼 Artists Featured:
Shankar Mahadevan appears most
often, suggesting a strong
representation of Indian classical or
fusion music.
Other notable artists include Alan
Silvestri, Daniel Pemberton, and
Ramin Djawadi, known for cinematic
compositions.
Taylor Swift, Justin Bieber, and
Madonna represent mainstream pop
entries.
Blake Neely and Billy Joel round out
the diversity with orchestral and
classic rock styles.
Next Slide
📊 Most Frequent Track Names
"Merry Christmas, Happy Holidays" is the most frequently
appearing track.
Several well-known hits like "Love Yourself" and "Just the
Way You Are" appear prominently.
This may suggest duplicate entries, remixes, or covers within
the dataset.
Next Slide
🌍 Language Distribution of Spotify Tracks
💡 Insight:
The dataset reflects a blend of
global and regional music trends,
with English leading but significant
contributions from South Asian and
East Asian languages.
Next Slide
Univariate Analysis of Temporal variables
Next Slide
📈 Yearly Popularity Trend of Spotify Tracks
Mid-1970s Peak: A noticeable spike in popularity, possibly driven by iconic releases or cultural shifts.
1980s Decline: Popularity dips, suggesting reduced engagement or fewer standout tracks.
1995–2015 Recovery: A gradual upward trend, reflecting the rise of digital music and streaming platforms.
Post-2015 Drop: A decline toward 2020, potentially due to dataset bias or changing listener behavior.
Next Slide
Bivariate analysis - Numerical Vs Numerical
variable
Next Slide
📊 Correlation Heatmap
Most features show weak
correlations with popularity.
Notable observations:
Danceability and
instrumentalness: strong
negative correlation (-0.40).
Danceability and
acousticness: moderately
negative (-0.34).
Popularity and acousticness:
slightly negative (-0.14).
Next Slide
📊 Danceability vs. Instrumentalness
Tracks with high danceability typically show low instrumentalness.
Instrumental tracks are generally less danceable, which aligns with expectations.
This negative correlation is consistent across the dataset, supporting the intuitive link
between vocals/beats and dance appeal.
Next Slide
Bivariate analysis - Numerical Vs categorical
variable
Next Slide
📊 SLIDE 1: AVERAGE POPULARITY BY LANGUAGE
🎧 Features Visualized:
Korean-language tracks show the highest average popularity among all languages in the dataset.
Hindi and English tracks follow, while Malayalam tracks show the lowest average popularity.
Language appears to influence engagement and reach on Spotify.
.
Next Slide
Overall Analysis key-points
Data Quality
👤 Top 🔤 Language 🎶 Track
Artists &
Insights Popularity Characteristics
Tracks
Next Slide
🧠 Strategic Takeaways
This Spotify EDA revealed that most tracks
have low popularity, with a few dominating
global attention—a clear sign of algorithm-
driven discovery. Features like energy,
danceability, and language diversity strongly
influence listener appeal, while custom
metrics like mood and vibe score unlock
deeper storytelling across regions and
decades..
Next Slide
🎯 Closing Message Suggestion
“From rhythms and moods to regional
languages and listener behavior, data has
revealed how music moves the world—and
what makes a track truly resonate. Let’s
keep listening with insight.
Next Slide
Thank You!
Thank you for joining this exploration into Spotify’s sonic landscape