Rae's Roots

An interactive interface using Netflix data to analyze what qualitative and quantitative movie attributes correspond to increased viewership.

Using Netflix’s What We Watched the First Half of 2025 dataset of movie titles, runtime, and views, analyze and examine what attributes lead to higher viewership (e. g. what makes people click on a movie while scrolling through searching for what to watch?).

I investigated the Netflix data through some simple aggregation to uncover trends and found that 72% of movie titles have 5 words or fewer. I found this interesting as it shows that more value can be placed on each individual word when considering user behavior like clicks, which ultimately leads to increased viewership. I then filtered the dataset to the top 50 highest occurring words and created overarching and overlapping word categories manually based on personal word association.

I also found trends in runtime, with most movies being around 80-140 minutes but some outliers. In considering the act of scrolling through Netflix searching for something to watch, I decided to investigate this link between words and runtime to see if these factors impact views by catching users' attention to click play on a movie.

I created a heatmap that buckets titles based on word/word category and runtime and fills color opacity based on average number of views. The first iteration showed a tooltip with an overview of the data that makes up a cell on hover and a panel to the right of the heatmap that gets populated with individual title/view details upon user interaction.

After investigating these general trends, I used my existing dataset to fetch qualitative data from The Movie Database API with the goal of uncovering how higher viewership relates to more verbal and visual descriptive factors. I integrated movie posters and descriptions into my existing heatmap, and used Wink.js library to quantify the sentiment of movie titles and descriptions to add to the details available upon interaction. I also redesigned my interface to be more user-friendly and use a more neutral color scale after red/green felt more intuitive to display sentiment analysis. Using hover and click, users can explore details about individual movies, words, word categories, and views. 

Catching Attention

Case Study Coming Soon