Embedded Files
Overview
Overview
Scenario: Predict album sales based on advertising budget.
Utilized a dataset to build a regression model and identify key factors influencing album sales.
Data
Data
The dataset contains 200 rows of data, each representing a different album, with the following columns:
TotalSales: Total album sales (in thousands)
AdvertBudget: Advertising budget (in thousands)
AirplayTime: Number of times the song played on the radio
AttractivenessScore: Attractiveness of the band (rating)
Disclaimer: This project was created using fictitious data and is not intended for actual business use. It does not reflect or represent actual data or insights.
Exploratory Analysis with Visuals
Exploratory Analysis with Visuals
Positive correlation between Advertising Budget and Sales
Positive Correlation between AirPlay Times and Sales
No relationship between Attractiveness Score and Sales
Linear Regression
Linear Regression
Predicting the Impact of Advertising Budget on Album Sales
Objective: Predict the impact of the advertising budget on album sales.
Outcome: The model demonstrated a strong fit based on the F-statistic and p-value, confirming that the advertising budget predicts album sales statistically significantly.
Prediction: With an advertising budget of $135,000 for the new Breaking Benjamin album, the expected album sales are 13,034 units.
Observation: While the F-statistic and p-value indicate a strong model, the scatter plot of the data shows a significant deviation from the regression line. This suggests that there may be additional factors affecting album sales. To improve the model’s accuracy, a multivariate regression approach will be applied, incorporating other variables such as airplay time and band attractiveness
Linear regression results summary
Linear regression results summary
Multiple Regression
Multiple Regression
Objective: Improve the prediction of album sales by incorporating additional factors, such as airplay time and band attractiveness, alongside the advertising budget.
Outcome: After applying the multivariate regression model with advertising budget, airplay time, and attractiveness score as predictors, the model showed improved explanatory power for predicting album sales.
Observation: While the F-statistic and p-value still indicate a strong fit for the model, the scatter plot in the initial linear regression showed significant deviations from the regression line. This suggested that additional variables, such as airplay time and band attractiveness, might influence sales. The multivariate regression approach helps address these factors and improve the model’s accuracy.
Multiple Regression Results Summary
Multiple Regression Results Summary
Insights
Insights
Advertising Budget: Advertising budget is a key predictor of album sales. Investing in marketing is a critical factor in driving album sales.
Airplay Time’s Role: Increased radio exposure significantly boosts sales, highlighting its marketing value.
Improved Prediction with Multivariate Model: The multivariate regression model improves the prediction of album sales by considering multiple factors, demonstrating that album sales are influenced by more than just the advertising budget. This emphasizes the need for a holistic marketing strategy that includes both traditional and intangible elements (e.g., image, airplay).
Page updated
Google Sites
Report abuse