Examining the Impact of Advertising Expenditure on Sales Revenue Using Regression Analysis in SPSS

In the contemporary competitive business landscape, organizations are spending a lot of resources on advertising to market their products and services. The relation between the advertisement expenditure to the sales revenue is critical both in making an effective decision in marketing and maximization of the investment returns (Tellis, 2012). The effectiveness of advertising has been economically studied, and the results have repeatedly shown that advertising can be of key importance in changing the tastes and preferences of consumers (Vakratsas & Ambler, 1999). The existing studies have identified that advertising spend has a positive correlation with the sales of various industries, though the extent to which this association can be observed can be determined by the circumstances in the market and which media are used (Sethuraman et al., 2018). This paper aims to discuss the relationship between monthly advertising expenditure and sales revenue in a regression analysis using the Statistical Package of Social Sciences (SPSS).

Data Analysis

The data used in the analysis was obtained from a retail company that was in the consumer electronics industry. The data set will consist of 20 months, and two variables (two important variables) will be the monthly advertising spending (in a thousand dollars) and the monthly sales revenue (in a thousand dollars). The expenditure on advertising was between $15,000 and $50,000 per month, and the sales revenue was between $150,000 and $490,000 per month.

Descriptive Statistics

Table 1: Descriptive statistics

The descriptive statistics show that the average funds spent by the company on advertising are $32,250 (SD = $10,541) per month, and average sales revenues are $312,750 (SD = $98,423) per month. The large standard deviation in the two variables indicates a high degree of variation on a month-to-month basis, and this gives a strong platform on which it can be determined that changes in advertisement expenditure are related to monthly changes in sales revenue.

Correlation Analysis

In order to determine the strength and the direction of the relationship between sales revenue and advertising expenditure, the Pearson correlation coefficient was calculated.

Table 2: Correlations

**. Correlation is significant at the 0.01 level (2-tailed).

The correlation analysis reveals a very strong positive correlation between advertising expenditure and sales revenue (r = 0.922, p < 0.001). This correlation coefficient means that 92.2% of the linear relationship between advertising spending and sales is explained by the association between them. The statistical significance (p < 0.001) is strong evidence that this relationship did not occur by chance. A correlation coefficient higher than 0.90 indicates an extremely strong relationship and can be interpreted as advertising expenditure being a powerful predictor of sales revenue for this retail company.

Regression Analysis

In order to explore more on the predictive relationship between the advertising spend and sales revenue, a simple linear regression analysis was conducted. The independent variable was the advertising expenditure, and the dependent variable was the sales revenue.

Model Summary

Table 3: Model Summary

a. Predictors: (Constant), Advertising Expenditure (1000s)

The model summary shows that advertising spending and sales revenue are a very well fit. The R-squared value is 0.850, which means that approximately 85.0% of the variation in monthly sales revenue can be explained by variation in advertising expenditure. This is an extremely high coefficient, indicating that advertising expenditures are a very strong indicator of the sales performance. The adjusted R-squared value of 0.842 shows that the model has true explanatory power. The standard error of the estimate is $39,129, which suggests that the model gives accurate predictions.

Analysis of Variance (ANOVA)

The ANOVA table tests the overall significance of the regression model.

Table 4: ANOVA^a

a. Dependent Variable: Sales Revenue (1000s)

b. Predictors: (Constant), Advertising Expenditure (1000s)

The ANOVA results are very convincing in supporting the importance of the regression model (F(1, 18) = 101.910, p < 0.001). The F-statistic is used to test whether advertising expenditure accounts for significantly more variance in sales revenue than zero. With a very small p-value, the odds for getting such a high correlation relationship by chance are essentially zero. Sum of Squares decomposition says that 156,015.625 thousand squared dollars variance was explained by the regression model, and only 27,547.875 thousand squared dollars variance remained unexplained, showing that around 85% of the total variance of Sales Revenue was explained by Advertising Expenditure.

Regression Coefficients

The coefficients table provides the specific parameter estimates for the regression equation.

Table 5: Coefficients^a

a. Dependent Variable: Sales Revenue (1000s)

The coefficients table shows the exact nature of the relationship between the advertising expenditure and the sales revenue. The constant (intercept) of 13.175 is the predicted baseline of sales revenue, as a result of zero advertising expenditure. The unstandardized value for the coefficient of advertising expenditure (B = 9.291) is most important in interpretation. This coefficient states that for every additional $1,000 spent on advertising, the sales revenue is expected to increase by approximately $9,291, holding all other factors constant. This is a huge payoff on advertising money, which can be calculated as such that each dollar of advertising money can be generated into approximately $9.29 sales revenue.

This relationship is overwhelmingly statistically significant (t = 10.095, p < 0.001). The advertising coefficient 95% has a value of 7.354 to 11.228, indicating that we are very much assured that the true value will lie within this range. Zero is out of the range, and this is quite a slim range, which validates the statistical significance and precision. Standardized coefficient (Beta = 0.922) indicates that an increase of one standard deviation in advertising expenditure will increase the sales revenue by 0.922 standard deviations, and therefore a very powerful effect size.

Regression Equation and Practical Application

Based on the regression coefficients, the predictive equation for sales revenue can be expressed as:

Sales Revenue (1000s) = 13.175 + 9.291 × Advertising Expenditure (1000s)

This equation is a powerful tool that can be used to predict sales revenue from advertising expenditure. For example, if the company has invested $30,000 in advertising in a given month, the predicted sales revenue would be: 13.175 + 9.291(30) = 291.905, or approximately $291,905. If the company increases its advertising budget from $30,000 to $40,000, the expected increase in sales revenue would be about 9.291 * 10 = $92,910.

The implications are immense for marketing strategy in practice. The regression formula indicates that advertising expenditure is extremely effective, as for every dollar of advertising expenditure, the advertising expenditure generates about $9.29 in sales revenue - a return on advertising expenditure of approximately 929%. Marketing managers can apply this model to determine the best advertising budget and predict sales from planned campaigns. However, this relationship may not continue linearly at very large levels of advertising expenditure because of the possibility of market saturation or diminishing returns.

Conclusion

The regression analysis that has been done using SPSS gives strong statistical evidence of a very high positive relationship between advertising expenditure and sales revenue. The Pearson correlation coefficient (0.922) shows that there is an extremely positive association, with p = 0.001. As indicated by the regression analysis, the variance in monthly sales revenue (predominantly) due to advertising expenditure is about 85.0% with an F-statistic of 101.910 (p < 0.001), showing that there has been a total significance of the model.

The advertising expenditure coefficient (B = 9.291) shows that an increase in the advertising expenditure of one more $1,000 would be related to an increment of around 9,291 in sales revenue. This is an impressive payoff (929 percent), indicating that this retail company is very productive with regard to advertising. The 95% confidence interval (7.354 to 11.228) also provides additional support to the validity of this estimate, and the narrow interval can be statistically significant and dependable.

These are findings with great strategic importance in the area of business decision-making. The favorable and strong correlation is an empirical sign of why to continue or to increase the amount of advertising capital. The predictive equation will help managers predict the sales results provided by the intended advertisement spending, which will facilitate better planning and budgeting. Its high R-squared value implies that the percentage of variation of sales is mainly caused by advertising, but the 15 percent of unexplained variance means that there are other factors, including seasonality, competitive activities, and market conditions, that contribute to the remaining portion of sales variation.

As much as these findings are striking, there are a number of shortcomings that ought to be recognized. The study relies on data from one company over 20 months, which could not be generalized. The model also assumes that there is a linear relationship within the range that is being observed, but not at extremely high and low levels of amount of advertising expenditure. It is possible that in the future, this model can be increased with the help of other predictors, including the indicators of seasonality, competition levels of advertising, and economic factors, to obtain a better picture of the drivers of sales.