1 Problem Statements Overview

  1. Problem 1: Perform Descriptive Statistics to summarize customer data.  
  2. Problem 2: Identify key drivers of spending using Linear Regression.  
  3. Problem 3: Compare group spending using ANOVA.  
  4. Problem 4: Segment customers using K-Means Clustering.  
  5. Problem 5: Present results through Data Visualization.

1.1 Step 0: Data Preparation

Step Action Description
0.1 Load Data Read CSV file into R. Imports dataset for analysis.
0.2 Handle Missing Values Remove rows with missing Satisfaction Level. Ensures clean dataset.
0.3 Convert to Factors Convert categorical variables to factors. Prepares data for modeling. 
# 0.1 Load dataset
df <- read_csv("C:\\Users\\Shaik Fawaz\\Documents\\E-commerce Customer Behavior - Sheet1.csv")

# 0.2 Handle missing Satisfaction Level
df_clean <- df %>%
  drop_na(`Satisfaction Level`)

# 0.3 Convert character columns to factors
df_clean <- df_clean %>%
  mutate(
    Gender = as.factor(Gender),
    City = as.factor(City),
    `Membership Type` = as.factor(`Membership Type`),
    `Discount Applied` = as.factor(`Discount Applied`),
    `Satisfaction Level` = as.factor(`Satisfaction Level`)
  )

cat("Data successfully loaded and cleaned.\n")
## Data successfully loaded and cleaned.
str(df_clean)
## tibble [348 × 11] (S3: tbl_df/tbl/data.frame)
##  $ Customer ID             : num [1:348] 101 102 103 104 105 106 107 108 109 110 ...
##  $ Gender                  : Factor w/ 2 levels "Female","Male": 1 2 1 2 2 1 1 2 1 2 ...
##  $ Age                     : num [1:348] 29 34 43 30 27 37 31 35 41 28 ...
##  $ City                    : Factor w/ 6 levels "Chicago","Houston",..: 5 3 1 6 4 2 5 3 1 6 ...
##  $ Membership Type         : Factor w/ 3 levels "Bronze","Gold",..: 2 3 1 2 3 1 2 3 1 2 ...
##  $ Total Spend             : num [1:348] 1120 780 511 1480 720 ...
##  $ Items Purchased         : num [1:348] 14 11 9 19 13 8 15 12 10 21 ...
##  $ Average Rating          : num [1:348] 4.6 4.1 3.4 4.7 4 3.1 4.5 4.2 3.6 4.8 ...
##  $ Discount Applied        : Factor w/ 2 levels "FALSE","TRUE": 2 1 2 1 2 1 2 1 2 1 ...
##  $ Days Since Last Purchase: num [1:348] 25 18 42 12 55 22 28 14 40 9 ...
##  $ Satisfaction Level      : Factor w/ 3 levels "Neutral","Satisfied",..: 2 1 3 2 3 1 2 1 3 2 ...

1.2 Problem 1: Descriptive Statistics

Objective: To summarize customer spending behavior and understand data distribution.

Step Action Description
1.1 Summary Stats Calculate Mean, Median, and SD. Measures data central tendency and spread.
1.2 Frequency Plot Create histogram for Total Spend. Visualizes spending distribution. 
# Generate descriptive statistics
summary(df_clean)
##   Customer ID       Gender         Age                   City   
##  Min.   :101.0   Female:173   Min.   :26.00   Chicago      :58  
##  1st Qu.:188.8   Male  :175   1st Qu.:30.00   Houston      :56  
##  Median :276.5                Median :32.00   Los Angeles  :59  
##  Mean   :275.9                Mean   :33.58   Miami        :58  
##  3rd Qu.:363.2                3rd Qu.:37.00   New York     :59  
##  Max.   :450.0                Max.   :43.00   San Francisco:58  
##  Membership Type  Total Spend     Items Purchased Average Rating 
##  Bronze:114      Min.   : 410.8   Min.   : 7.00   Min.   :3.000  
##  Gold  :117      1st Qu.: 505.8   1st Qu.: 9.00   1st Qu.:3.500  
##  Silver:117      Median : 780.2   Median :12.00   Median :4.100  
##                  Mean   : 847.8   Mean   :12.63   Mean   :4.024  
##                  3rd Qu.:1160.6   3rd Qu.:15.00   3rd Qu.:4.500  
##                  Max.   :1520.1   Max.   :21.00   Max.   :4.900  
##  Discount Applied Days Since Last Purchase   Satisfaction Level
##  FALSE:173        Min.   : 9.00            Neutral    :107     
##  TRUE :175        1st Qu.:15.00            Satisfied  :125     
##                   Median :23.00            Unsatisfied:116     
##                   Mean   :26.61                                
##                   3rd Qu.:38.00                                
##                   Max.   :63.00
numeric_cols <- df_clean %>% select_if(is.numeric)
data.frame(
  Mean = sapply(numeric_cols, mean, na.rm = TRUE),
  Median = sapply(numeric_cols, median, na.rm = TRUE),
  SD = sapply(numeric_cols, sd, na.rm = TRUE)
)
##                                Mean Median          SD
## Customer ID              275.887931  276.5 101.3046114
## Age                       33.577586   32.0   4.8780238
## Total Spend              847.793103  780.2 361.6923754
## Items Purchased           12.632184   12.0   4.1460789
## Average Rating             4.023563    4.1   0.5791447
## Days Since Last Purchase  26.614943   23.0  13.4747498
# Histogram of Total Spend
ggplot(df_clean, aes(x = `Total Spend`)) +
  geom_histogram(bins = 20, fill = "steelblue", color = "black") +
  labs(title = "Distribution of Total Spend", x = "Total Spend", y = "Frequency")

Output Description: The summary statistics show the spread and central tendency of key numeric columns. The histogram illustrates how spending is distributed — whether most customers spend moderately or a few contribute to high spending.

1.3 Problem 2: Identifying Key Spending Drivers (Regression)

Objective: To identify which factors most influence customer spending.

Step Action Description
2.1 Build Model Predict Total Spend using predictors. Quantifies effects of predictors.
2.2 Interpret Coefficients Examine p-values and R². Determines significance and fit. 
# Build multiple linear regression model
lm_model <- lm(`Total Spend` ~ Age + `Items Purchased` + `Membership Type` + 
                 `Discount Applied` + `Average Rating`, data = df_clean)

cat("Summary of Linear Regression Model:\n")
## Summary of Linear Regression Model:
summary(lm_model)
## 
## Call:
## lm(formula = `Total Spend` ~ Age + `Items Purchased` + `Membership Type` + 
##     `Discount Applied` + `Average Rating`, data = df_clean)
## 
## Residuals:
##    Min     1Q Median     3Q    Max 
## -54.27 -19.67  -1.36  15.86  68.61 
## 
## Coefficients:
##                          Estimate Std. Error t value Pr(>|t|)    
## (Intercept)             -189.0576    27.1938  -6.952 1.84e-11 ***
## Age                        8.1156     0.5210  15.577  < 2e-16 ***
## `Items Purchased`         42.1572     0.9597  43.929  < 2e-16 ***
## `Membership Type`Gold    519.3501    12.4802  41.614  < 2e-16 ***
## `Membership Type`Silver  198.7341     8.3796  23.716  < 2e-16 ***
## `Discount Applied`TRUE   -82.0024     2.5889 -31.674  < 2e-16 ***
## `Average Rating`           7.8595     9.3604   0.840    0.402    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 22.85 on 341 degrees of freedom
## Multiple R-squared:  0.9961, Adjusted R-squared:  0.996 
## F-statistic: 1.443e+04 on 6 and 341 DF,  p-value: < 2.2e-16

Output Description: The regression summary highlights which variables (like items purchased or membership type) significantly impact total spend. The R-squared value shows how much variation in spending is explained by the predictors.

1.4 Problem 3: Comparing Spending Across Groups (ANOVA)

Objective: To test if Membership Type or City significantly affect Total Spend.

Step Action Description
3.1 Run ANOVA Fit model with aov(). Tests mean differences among groups.
3.2 Interpret Results Review F-values and p-values. Identifies significant group effects. 
anova_model <- aov(`Total Spend` ~ `Membership Type` * City, data = df_clean)
cat("ANOVA Table:\n")
## ANOVA Table:
summary(anova_model)
##                    Df   Sum Sq  Mean Sq F value Pr(>F)    
## `Membership Type`   2 42183508 21091754   35097 <2e-16 ***
## City                3  3005981  1001994    1667 <2e-16 ***
## Residuals         342   205528      601                   
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Output Description: The ANOVA table shows whether spending significantly differs across membership levels or cities. If p-values are below 0.05, those factors have statistically significant effects on spending.

1.5 Problem 4: Customer Segmentation (K-Means Clustering)

Objective: To group customers into segments based on their spending patterns.

Step Action Description
4.1 Scale Features Standardize numeric variables. Normalizes data for clustering.
4.2 Apply K-Means Run clustering with k=3. Creates customer segments.
4.3 Summarize Clusters Calculate average metrics per cluster. Describes each group’s characteristics. 
data_cluster <- df_clean %>%
  select(`Total Spend`, `Items Purchased`, `Days Since Last Purchase`, `Average Rating`) %>%
  scale()

set.seed(42)
kmeans_result <- kmeans(data_cluster, centers = 3, nstart = 25)
df_clean$Cluster <- as.factor(kmeans_result$cluster)

cluster_summary <- df_clean %>%
  group_by(Cluster) %>%
  summarise(
    Avg_Spend = mean(`Total Spend`),
    Avg_Items = mean(`Items Purchased`),
    Avg_Rating = mean(`Average Rating`)
  )

cat("\nCluster Summary:\n")
## 
## Cluster Summary:
print(cluster_summary)
## # A tibble: 3 × 4
##   Cluster Avg_Spend Avg_Items Avg_Rating
##   <fct>       <dbl>     <dbl>      <dbl>
## 1 1           1456.     19.9        4.81
## 2 2            983.     13.4        4.35
## 3 3            547.      9.57       3.53

Output Description: The cluster summary shows distinct customer groups — for example, high spenders with frequent purchases and low spenders with fewer items. This segmentation can help target different marketing strategies.

1.6 Problem 5: Data Visualization and Insights

Objective: To visualize patterns in customer data for better understanding.

Step Action Description
5.1 Boxplot Compare Total Spend by Membership Type. Highlights variation in spend levels.
5.2 Scatter Plot Plot Age vs Total Spend. Shows age-based spending behavior.
5.3 Bar Chart Visualize satisfaction across cities. Compares satisfaction by location. 
# --- Step 5.1: Boxplot ---
ggplot(df_clean, aes(x = `Membership Type`, y = `Total Spend`, fill = `Membership Type`)) +
  geom_boxplot() +
  labs(title = "Total Spend by Membership Type", x = "Membership Type", y = "Total Spend")

# --- Step 5.2: Scatter Plot ---
ggplot(df_clean, aes(x = Age, y = `Total Spend`, color = `Satisfaction Level`)) +
  geom_point() +
  labs(title = "Age vs Total Spend", x = "Age", y = "Total Spend")

# --- Step 5.3: Bar Chart ---
ggplot(df_clean, aes(x = `City`, fill = `Satisfaction Level`)) +
  geom_bar(position = "dodge") +
  labs(title = "Satisfaction Level by City", x = "City", y = "Count")

Output Description:

  • The boxplot shows that premium or higher membership types have greater average spending.
  • The scatter plot reveals that middle-aged customers tend to spend more, with satisfaction linked to spending level.
  • The bar chart highlights which cities have the highest satisfaction rates.

1.7 Final Findings and Insights

Key Insights from the Dataset:

  1. Spending Behavior:

    • Most customers spend within a moderate range, with a few outliers indicating high-value buyers.
    • Average spending increases with membership level.

  2. Regression Results:

    • The number of items purchased and membership type are the strongest predictors of total spending.
    • Discounts have a moderate positive impact, while age has a smaller effect.

  3. ANOVA Findings:

    • There is a statistically significant difference in average spending between different membership types.
    • Some cities show slightly higher spending, but the effect is secondary to membership.

  4. Customer Segmentation:

    • Cluster 1: High spenders with frequent purchases and high ratings.
    • Cluster 2: Average customers with balanced spend and satisfaction.
    • Cluster 3: Low spenders who purchase infrequently and give lower ratings.

  5. Visual Insights:

    • Premium members are the most satisfied and spend the most.
    • Younger and middle-aged customers show stronger engagement.
    • Certain cities display a higher satisfaction rate, indicating potential market focus areas.

Conclusion

This R Markdown report demonstrates how Descriptive Analytics can transform raw e-commerce data into actionable insights. Through data cleaning, regression, ANOVA, clustering, and visualization, we discovered:

  • Membership type and purchase frequency are key spending drivers.
  • Spending patterns vary significantly across customer groups.
  • Data visualization helps highlight behavioral and regional trends for data-driven business strategy.

```