Introduction:
Advertising plays a central role in shaping consumer behavior and driving product demand. Firms routinely invest substantial resources in various advertising channels; including television, radio, and print media, with the expectation that these efforts will translate into increased sales. However, the relationship between advertising and sales is complex and often influenced by numerous contextual factors. The effectiveness of advertising may vary depending on the medium, the market environment, and consumer responsiveness. (Tellis, n.d.)
The data set I will be using is the advertising data set. It comes from Nielsen’s Ad Intel, which tracks advertising occurrences across the U.S. since 2010. It consists of 200 rows and 4 columns, which are labeled TV, Radio, Newspaper, and Sales. In the past, it has also been used to help businesses analyze the effectiveness of advertising, audience targeting, and media trends.
This analysis aims to investigate the relationship between advertising expenditures and product sales using the advertising data set, which details spending on TV, radio, and newspaper advertising. I will be using an Ordinary Least Squares (OLS) regression to quantify the impact of each advertising medium on sales performance. By estimating a multiple linear regression model, this analysis seeks to identify which forms of advertising are most strongly associated with increased sales and to what extent.
## TV Radio Newspaper Sales
## 1 230.1 37.8 69.2 22.1
## 2 44.5 39.3 45.1 10.4
## 3 17.2 45.9 69.3 12.0
## 4 151.5 41.3 58.5 16.5
## 5 180.8 10.8 58.4 17.9
## 6 8.7 48.9 75.0 7.2##
## Call:
## lm(formula = Sales ~ TV + Radio + Newspaper, data = advertising)
##
## Residuals:
## Min 1Q Median 3Q Max
## -7.3034 -0.8244 -0.0008 0.8976 3.7473
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 4.6251241 0.3075012 15.041 <2e-16 ***
## TV 0.0544458 0.0013752 39.592 <2e-16 ***
## Radio 0.1070012 0.0084896 12.604 <2e-16 ***
## Newspaper 0.0003357 0.0057881 0.058 0.954
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.662 on 196 degrees of freedom
## Multiple R-squared: 0.9026, Adjusted R-squared: 0.9011
## F-statistic: 605.4 on 3 and 196 DF, p-value: < 2.2e-16Analyzing the Data:
First, I loaded the advertising data set into R. Then, I ran an OLS model on the predictors for TV, radio, and newspaper sales. This model will illustrate the coefficients and significance of each predictor. Some of the key findings of this model are that TV advertising has a coefficient of 0.0544. This indicates that for every $1,000 spent on TV advertising, sales increase by ~54.4 units, holding other variables constant. It also has a p value of <2e-16; which indicates that TV advertising has a strong and statistically significant effect on sales, which supports my hypothesis. Radio advertising has a coefficient of 0.107, which indicates sales increase of ~107 units for every $1,000 spent on advertising. Radio also showed a significant positive effect on sales. However, it also interestingly showed a larger effect size than TV advertising. On the other hand, newspaper advertising did not show any meaningful relationships, with a coefficient of 0.00034 and a p-value of 0.954, meaning that spending on newspaper advertising does not significantly affect sales. I also included visualizations to illustrate the relationship between each predictor and sales.
library(readr)
library(ggplot2)
# TV vs. Sales
ggplot(advertising, aes(x = TV, y = Sales)) +
geom_point(color = "steelblue") +
geom_smooth(method = "lm", se = TRUE, color = "darkred") +
labs(title = "TV Advertising vs. Sales",
x = "TV Advertising Spend",
y = "Sales") +
theme_minimal()## `geom_smooth()` using formula = 'y ~ x'
This plot shows a strong positive linear relationship between TV advertising spend and sales. The fitted line indicates that as TV spending increases, sales tend to increase as well, consistent with the high coefficient and significance in my regression model.
# Radio vs. Sales
ggplot(advertising, aes(x = Radio, y = Sales)) +
geom_point(color = "forestgreen") +
geom_smooth(method = "lm", se = TRUE, color = "darkred") +
labs(title = "Radio Advertising vs. Sales",
x = "Radio Advertising Spend",
y = "Sales") +
theme_minimal()## `geom_smooth()` using formula = 'y ~ x'
Radio advertising also has a positive and statistically significant relationship with sales, although it’s slightly weaker than TV advertising. This is reflected visually in the upward-sloping regression line.
ggplot(advertising, aes(x = Newspaper, y = Sales)) +
geom_point(color = "orange") +
geom_smooth(method = "lm", se = TRUE, color = "darkred") +
labs(title = "Newspaper Advertising vs. Sales",
x = "Newspaper Advertising Spend",
y = "Sales") +
theme_minimal()## `geom_smooth()` using formula = 'y ~ x'
Newspaper advertising appears to have almost no relationship with sales, as shown by the nearly flat regression line. This corresponds with the model’s results, where Newspaper has a very low t-value and high p-value (not statistically significant).
Conclusion:
This analysis aimed to evaluate the relative impact of different advertising media: TV, radio, and newspaper; on sales performance using a multiple linear regression model. The findings indicate that both TV and radio advertising are statistically significant predictors of sales, with TV exerting the strongest influence. This supports the initial hypothesis that TV advertising, likely due to its broad reach and visual appeal, plays a more substantial role in driving consumer purchasing behavior. Radio advertising also shows a meaningful positive effect, though less pronounced than that of TV. In contrast, newspaper advertising does not have a statistically significant impact on sales, suggesting diminishing effectiveness of print media in the current marketing landscape. The model demonstrates a high level of explanatory power, with an R-squared value of 0.90, indicating that 90% of the variance in sales is accounted for by the included predictors. These results highlight the importance of focusing advertising investments on more impactful media channels, particularly television and radio, to optimize marketing outcomes.