Evaluate credit-limit increase profitability
Company: Capital One
Role: Data Scientist
Category: System Design
Difficulty: medium
Interview Round: Onsite
## Business/Analytics Case: Credit Limit Increase Strategy
You are a data scientist supporting a consumer credit business.
### Scenario
The company is considering a **credit-limit increase** program for a specific customer segment (e.g., customers with 6–12 months tenure and mid FICO). You must recommend whether to launch, and how to size/target the program.
### What to do
1. **Clarify the objective**: Is success measured by revenue growth, profit, lower default rate, higher approval rate, retention, or a combination?
2. Build a simple **P&L / unit economics** model:
- Use a profit identity such as:
\(\text{Profit} = \text{Revenue} - \text{Loss} - \text{Operational Cost}\)
- Define what counts as revenue (e.g., interest, interchange, fees) and loss (e.g., charge-offs, fraud, cost of funds).
3. Identify key **levers** and trade-offs (e.g., limit size, eligibility rules, APR/pricing, risk policy, model thresholding).
4. Propose an **evaluation plan** (data needed, segmentation, experiment design) and explain how you would sanity-check numbers with quick mental math.
### Output
Provide a structured recommendation: what to launch (or not), who to target, what metrics to optimize, and how to measure incremental profit and risk.
Overview: Evaluate whether to launch a credit-limit increase program for a customer segment, including whom to target and how much to increase limits. Define success metrics and an evaluation plan that accounts for incremental revenue, losses, costs, and risk.
Read the full Capital One Data Scientist interview experience this question came from
