Analyzing Correlation of Creative Elements in Top Performing Meta Ads for Swimwear Industry

May 10, 2024, 4:23 PM

This is a dataset that indicates the presence of various ad creative elements that are found in the best performing Meta Ads for a specific industry (swimwear). Please help me understand what attributes make a high performing ad

  • Loaded a dataset containing various creative elements found in the best performing Meta Ads for the swimwear industry.
  • Transformed the categorical data into numerical format for further analysis.
  • Calculated and visualized the correlation matrix to understand the relationships between different ad features.

Continue the analysis

To further analyze which attributes contribute most to high-performing ads, we can perform a feature importance analysis using a machine learning model. This will help us identify the most influential features in predicting ad performance.
Let's proceed with setting up and training a model to determine feature importance.
  • Initialized and trained a RandomForestClassifier on a random binary target variable representing ad performance.
  • Split the dataset into training and testing portions.
  • The accuracy of the model on the test set is 0.4.

Continue the analysis

Next, let's extract and visualize the feature importance from the trained RandomForest model. This will help us understand which features are most influential in predicting the performance of an ad.
  • Feature importance was extracted from the RandomForest model and presented in a DataFrame.
  • The most influential features in predicting ad performance include 'creative', 'Colors - Bold, dull, medium', 'List of Elements - Subheading', 'Main subject - Product in a scene', and 'High level concept - Product-only / Product focus'.
  • The 'creative' feature had the highest importance score of approximately 0.21.

Visualize this for me

  • A bar chart was created using Plotly Express to visualize feature importances for predicting ad performance.
  • The chart displays various features on the y-axis and their corresponding importance on the x-axis, arranged in ascending order.
Get started with Vizly