How E-Commerce Data Helps Us Understand Customer Behavior

By abhishek.verma75000 · October 15, 2026

In this article, we will explore an e-commerce customer segmentation dataset containing 50,000 records for 2026. We will look at what the data reveals…

The Consumer segment generates the highest total customer profitability at about $1.3 billion, while the Premium segment has the highest average customer lifetime value at roughly $57,998 per customer. So different segments lead each metric. I've generated a side-by-side chart comparing total profitability and average CLV across segments, along with supporting tables with the segment breakdown.

Beauty has the highest average order value at $511.49, followed by Health & Personal Care ($507.61), Fashion ($506.97), Toys ($506.75), and Electronics ($505.92). For average total spent per customer, Electronics comes out on top at $51,601.10, followed by Fashion ($50,751.03), Home & Kitchen ($50,632.93), Office Supplies ($50,512.68), and Pet Supplies ($50,424.83). I've created side-by-side bar charts comparing both metrics across all 12 categories, a scatter plot relating order value to total spent, and summary tables of the results.

Higher churn risk goes hand in hand with more complaints, more returns, and lower satisfaction. Customers in the Very Low risk group average 5.98 complaints, 7.90 returns, and a satisfaction score of 3.81. At the other extreme, Very High risk customers average 11.73 complaints, 14.66 returns, and a satisfaction score of just 1.52. The Low, Medium, and High categories fall in between, with complaints and returns climbing steadily and satisfaction dropping at each step. A chart and summary tables were generated to show these differences across risk levels.

Average total spend and churn risk are essentially flat across tenure bands. Average total spent hovers around $50,000 in every band, ranging from $49,712 (61-72 months) to $50,960 (97-108 months). Average churn risk stays between 26.5 and 27.5 in every band. A dual-axis chart (bars for spend, line for churn risk) and summary tables show these results by tenure band, from 1-12 up to 109-120 months.

Among 41,093 High CLV customers, 2,270 have a high churn risk score (60 or above), with an average score of 68.1. Together they account for $134,036,504.90 in total spending, which is about 5.4% of all High CLV spend. Most of this exposure sits in the 'High' churn risk category (1,832 customers, about $109.1M), followed by Medium (251 customers, about $14.0M) and Very High (187 customers, about $10.9M). The biggest individual accounts at risk are Diamond-tier customers, each with roughly $185K-$198K in lifetime spend. I generated two charts and several supporting tables: one chart shows High CLV spend by churn risk score band, and the other ranks the top 10 at-risk customers by total spent.

Mobile App delivers the best ratio of customer lifetime value to customer acquisition cost, at 568.2x. Its average customer lifetime value is about $57,975 against an average acquisition cost of just $102.04, across 12,542 customers. Marketplace (564.2x) and Online (563.1x) follow closely, while In-Store ranks last at 558.2x. The overall ratio across all channels is 563.4x. I've generated two charts and supporting tables: one showing the CLV:CAC ratio by channel (with the overall average marked), and one comparing average CLV against average CAC for each channel.

I computed a Pearson correlation matrix across seven metrics: email open rate, click-through rate, conversion rate, total purchases, total spent (USD), satisfaction score, and days since last purchase. The result is displayed as an annotated heatmap, where each cell shows the correlation coefficient and the color scale runs from -1 (strong negative, red) through 0 (no relationship) to +1 (strong positive, blue). Supporting tables were also generated, including the ranked pairwise correlations used to identify the strongest relationships and the marketing-versus-spending comparisons. Use the heatmap and the ranked tables to see which pairs have the largest absolute correlations and to judge whether email open rate, click-through rate, and conversion rate move together with total purchases and total spent. Coefficients near zero indicate marketing engagement has little association with spending, while values beyond roughly 0.3 in absolute terms would indicate a meaningful link.

I broke the dataset down by loyalty tier and income bracket, producing summary tables that cover customer counts, average total spending (USD), and average satisfaction scores for each combination. These tables let you compare how spending and satisfaction vary across the different tier and income groups, and you can scan them to spot which combinations sit at the top and bottom of average spending. The results were generated as data tables for you to review.

Using the 1.5 x IQR rule, none of the three columns contained outliers. return count had Q1=4 and Q3=15 (IQR=11), giving bounds of -12.5 to 31.5 with 0 outliers (0.0%). complaint count had Q1=3 and Q3=11 (IQR=8), giving bounds of -9.0 to 23.0 with 0 outliers (0.0%). avg order value usd had Q1=$259.07 and Q3=$752.70 (IQR=$493.64), giving bounds of -$481.39 to $1,493.16 with 0 outliers (0.0%). Because no values fall outside these bounds, there are no extreme values to compare across behavior segment categories, so no customer group stands out as more likely to produce outliers. Summary tables of the bounds and outlier counts were generated.

The top 10 countries by customer count are very closely matched, ranging from about 2,944 (Spain) to 3,007 (United States) customers. The United States, Pakistan and Singapore take the top three spots. Average customer lifetime value (CLV) ranges from $55,178 (Spain) to $59,716 (Singapore). Satisfaction scores are tightly clustered around 3.0 (2.97 to 3.06), and average returns sit near 9.4 to 9.6 per customer. A sorted bar chart compares average CLV across the 10 countries, and a scatter plot of CLV vs. satisfaction shows the quadrants. Using the top-10 medians as thresholds (CLV about $57,489, satisfaction about 3.013), three countries rank high on CLV but low on satisfaction: Mexico (CLV $58,510, satisfaction 2.97), Germany (CLV $58,601, satisfaction 3.00) and the United States (CLV $57,616, satisfaction 3.01, only just below the satisfaction median). These are highlighted in red in the charts.