Income-Adjusted Customer Spending Analysis
- James Gifford
- Oct 26
- 2 min read
Understanding how spending behavior varies when normalized by local income, and what that reveals about financial patterns across different customer segments.
🎯 Project Overview
This project explores how customer spending intensity changes once adjusted for local income levels.
It asks:
How can customer spending behavior, when normalized by local income, reveal which groups drive disproportionate or unstable spending patterns?
By combining anonymized credit card transaction data with U.S. Census ZIP-level median income, the analysis identifies customers whose spending appears disproportionate to their earning environment, offering insight into potential overextension and market concentration patterns.
⚠️ Key Assumption: Income Modeling by ZIP
Since personal income data wasn’t available, each customer’s earning capacity was estimated using their ZIP code’s median household income (from the U.S. Census ACS 2019 dataset).
While approximate, this approach provides meaningful context for comparing spending patterns across economic segments.
📈 Executive Summary
This analysis combines credit card transaction data with ZIP-level income statistics to understand spending relative to earning capacity.
By adjusting for local income, the dashboards uncover customers whose spending behavior is atypically high within their economic context.
Key Takeaways
The top 10% of customers account for roughly 20% of total spending.
Overspending is most common in shopping categories, suggesting lifestyle-driven or convenience-based patterns.
Low- and mid-income groups make up nearly all overspenders, while high-income customers remain within expected spending ranges.
Volatility and spending intensity don’t show a clear linear relationship. Some overspenders show stable habits, others erratic spikes, suggesting different behavioral profiles within the same group.
🧭 Dashboard Summaries
1️⃣ Customer Overspending Overview
Analyzes spending and income patterns among the top overspending customers.
Key metrics include:
Overspending Intensity Index (5.74) – average spending multiple vs local income
% Overspending Customers (10%) – fixed threshold identifying top spenders
Avg. Volatility (1.92) – variation in monthly spending behavior
Insights:
18% of low-income customers fall into the top spending group vs 4% of mid-income customers.
Overspenders tend to exhibit late-night, high-value, and weekend spending behaviors, consistent with more impulsive or lifestyle-driven activity.
2️⃣ Behavioral & Income Outlier Analysis
Explores income distribution and overspending concentration by category.
Key metrics include:
Median Household Income: $61,507
Avg Relative Spend Index (RSI) for Overspenders: 4.83 (vs 1.76 overall)
Share of Spend from Top 10%: 20.4%
Insights:
Overspending intensity declines sharply as income rises.
“Spending lift” is most concentrated in shopping and online retail categories.
Outlier visualization highlights clear clusters of low- and mid-income customers with high RSI values.
💡 Analytical Value
This approach shows how income-normalized metrics can reveal spending behaviors that absolute amounts hide.
Possible business applications:
Identifying customers overspending relative to capacity
Refining credit limit or loyalty segmentation strategies
Designing targeted financial wellness or education outreach campaigns
Comparing spending behavior fairly across different income regions
🧾 Conclusion
This project demonstrates how combining behavioral and demographic data can reveal spending patterns that are invisible in raw transaction metrics alone.
Even with modeled income data, income-adjusted analysis provides a more realistic picture of customer behavior and financial capacity.
Future iterations could incorporate actual income, credit utilization, or savings data to deepen insight into financial wellness and portfolio risk, while also refining marketing and credit strategies.
Comments