Customer Sales Analysis Report
Date: 20-Aug-2025
This report provides an analysis of customer purchases, top-performing countries and
categories, and predictive insights for strategic business decisions.
Prepared By
Sandhya V G
Data Analyst.
Analyzed a consolidated view of 200 customers. Total revenue recorded in the available period
was $500,765.
The leading market was Germany, and the dominant product category was Clothing.
The monthly revenue trend highlights clear peaks aligned with campaign periods, and a Pareto
view confirms that a minority of customers drive the majority of revenue.
Recommendations focus on targeted retention in high-performing regions, bundling around top
categories, and structured experimentation to grow average order value.
Top country by revenue: Germany ($126,550.00)
Top 5 customers contributing the most revenue:
1124: $4,966.00
1151: $4,960.00
1026: $4,904.00
1099: $4,870.00
1170: $4,857.00
Best-performing category: Clothing ($114,830.00)
Regression Analysis
Regression Model: PurchaseAmount ~ Age + AnnualIncome + SpendingScore +
tenure_days Intercept: 3002.79
Coef Age: -6.33
Coef AnnualIncome: -0.00
Coef SpendingScore: -1.95
Coef tenure_days: 0.00
R² (test): -0.0116
RMSE (test): 1302.45
Sales By Country
Total revenue from each country is summarized here, highlighting which markets perform best
and should receive strategic focus.
Country PurchaseAmount
Germany $ 126,550.00
UK $ 117,282.00
India $ 104,321.00
Canada $ 79,291.00
USA $ 73,321.00
Revenue generated differs significantly across countries. The taller bars highlight the top-
performing countries, which can be prioritized for marketing and expansion efforts.
Sales By Category
This table outlines sales per product category. Categories with higher totals suggest areas to
boost promotion and inventory.
PurchasedCategory PurchaseAmount
Clothing $ 114,830.00
Electronics $ 104,959.00
Groceries $ 103,627.00
Books $ 92,978.00
Home $ 84,371.00
Sales by Category (Bar Chart)
Product categories contribute unevenly to overall sales. Categories with higher sales should be
emphasized in promotions and inventory planning.
Monthly Sales
Monthly totals illustrate sales trends throughout the year, aiding in identifying high-demand
periods and slower months.
OrderDate PurchaseAmount
44197 $ 82,368.00
44228 $ 65,023.00
44256 $ 68,079.00
44287 $ 76,892.00
44317 $ 77,334.00
44348 $ 75,018.00
44378 $ 56,051.00
Monthly Sales Trend (Line Chart)
Sales fluctuate over the months, showing peaks and troughs. These trends help identify
seasonal patterns and plan campaigns or stock accordingly.
TOP Customers
The top five customers are listed by total spending. These high-value customers are key targets
for personalized engagement and retention strategies.
CustomerID PurchaseAmount
1124 4966
1151 4960
1026 4904
1099 4870
1170 4857
Pareto Chart (Top Customers)
Customers are ranked by their total contributions to revenue. This helps identify the small
group that generates most of the business, guiding focus on loyalty programs.
Correlation Heat Map
Numerical factors like age, income, spending score, and tenure are compared against purchase
amounts. This provides insight into which attributes most influence customer spending.
index Age AnnualIncome SpendingScore tenure_days PurchaseAmount
Age 1 -0.118694066 0.032216213 -0.086678258
-
AnnualIncome 0.118694066 1 -0.026029801 -0.013353834
SpendingScore 0.032216213 -0.026029801 1 0.014149753
tenure_days
PurchaseAmoun -
t 0.086678258 -0.013353834 0.014149753 1
Correlation Heatmap
Relationships between customer attributes and purchase amounts are visualized here. Stronger
correlations indicate factors that significantly influence spending behavior.
Variable Pair Correlation Insight
Age vs Slight negative relationship. Older customers
-0.12
AnnualIncome tend to earn a bit less (but effect is very weak).
Age vs Slightly older customers spend a little less, but
-0.09
PurchaseAmount again weak effect.
AnnualIncome vs Almost no link. High income does not necessarily
-0.01
PurchaseAmount mean high spending in this dataset.
SpendingScore vs No meaningful relationship — a high “spending
0.01
PurchaseAmount score” doesn’t guarantee high purchases.
Most variables don’t show strong relationships
Other pairs Near 0
with each other.
The relationships between age, income, spending score, and purchase amount are very weak.
This means that customer demographics alone don’t explain how much people actually spend.
In practice, spending behavior seems to be driven by other factors like product preferences,
shopping frequency, and promotions.
Regression
Actual versus predicted purchase amounts are shown here, indicating how well the model
forecasts revenue. Close alignment means the model can be trusted for planning purposes.
Actual Predicted
922 2457.302022
2374 2552.513439
2645 2387.372571
144 2463.907471
2020 2466.542533
1042 2398.62533
370 2565.948464
864 2427.374217
2493 2630.294733
773 2687.22995
2110 2326.719636
2780 2447.755894
3375 2508.434424
3727 2562.39325
3926 2463.861332
3055 2407.715358
4284 2411.412922
3769 2560.573744
285 2456.584203
1669 2705.369947
3737 2493.696812
1783 2714.822107
2962 2478.503924
3693 2564.422462
2269 2385.479826
4835 2567.524196
4479 2692.114596
3763 2398.759083
2201 2594.722182
2085 2546.941552
926 2663.28655
2613 2636.811423
1967 2431.727383
3008 2480.826035
4960 2595.614627
1115 2533.088368
972 2611.032075
2254 2355.862324
4156 2393.054465
2088 2268.071424
Regression: Actual vs Predicted
The comparison between actual and predicted purchase amounts demonstrates the accuracy of
the predictive model. Points close to the line indicate reliable forecasts for future sales.
Insights:
The model isn’t capturing variability well. It tends to predict an “average spend” for
everyone.
This likely means the chosen variables (Age, Income, etc.) are not strong drivers of
sales.
Other features — such as purchase frequency, product category preferences, or
discount usage — would probably improve the model.
Actionable Recommendations:
1) Focus marketing and resources on top-performing country and category.
2) Target top customers for personalized promotions and loyalty campaigns.
3) Use tenure_days and spending score to identify customers likely to return and re-engage
them.
4) Monitor monthly trends to plan seasonal promotions and inventory levels.
