Evaluate a credit-card acquisition partnership

Quick Overview

This question evaluates cohort-level unit economics, discounted cash-flow (NPV) modeling, and sensitivity/break-even analysis related to customer acquisition and credit-card portfolio economics.

Evaluate a credit-card acquisition partnership

Company: Capital One

Role: Data Scientist

Category: Analytics & Experimentation

Difficulty: medium

Interview Round: Onsite

You’re evaluating a co-branded partner that can deliver 50,000 new credit-card customers. Assumptions (Year 1 unless stated): - Activation rate: 80% (only activated users generate economics). - Monthly spend per activated user: $600 for 12 months. - Interchange revenue: 1.5% of spend. - Rewards cost: 1.2% of spend. - Annual fee: $95 collected in month 1 from 25% of activated users. - Servicing cost: $1.00 per user per month. - Expected credit loss (bad debt): $24 per activated user over 12 months. - Partner economics: $70 CPA to partner + $100 signup bonus per activated user. - Discount rate: 10% annually (assume monthly discounting at 10%/12). Tasks: 1) Compute the 12-month NPV per acquired user and total NPV for the cohort. Show all components: interchange, rewards, fees, servicing, credit loss, and acquisition costs. 2) Derive the break-even CPA (holding other assumptions fixed). 3) Sensitivity: Recompute NPV for (a) interchange ±0.3 pp, and (b) credit loss ±$10. Identify which variable the decision is most sensitive to and state your recommendation.

Quick Answer: This question evaluates cohort-level unit economics, discounted cash-flow (NPV) modeling, and sensitivity/break-even analysis related to customer acquisition and credit-card portfolio economics.

|Home/Analytics & Experimentation/Capital One
Capital One logo
Capital One
Oct 13, 2025
mediumData ScientistOnsiteAnalytics & Experimentation
16
0

Cohort NPV and Sensitivity for New Credit-Card Customers

Context

You are evaluating a co-branded partner expected to deliver 50,000 newly acquired credit-card customers in Year 1. Only activated users generate spend-based economics (revenue, rewards, credit losses, fees). Use monthly discounting at an annual rate of 10% (i = 10%/12 per month). Compute NPV over the first 12 months only.

Assumptions (Year 1 unless noted):

  • Cohort size: 50,000 acquired users; activation rate = 80%.
  • Monthly spend (per activated): $600 for 12 months.
  • Interchange revenue: 1.5% of spend.
  • Rewards cost: 1.2% of spend.
  • Annual fee: $95 collected in month 1 from 25% of activated users.
  • Servicing cost: $1.00 per user per month.
  • Expected credit loss (bad debt): $24 per activated user over 12 months.
  • Partner economics: 70CPA(topartner)peracquireduser+70 CPA (to partner) per acquired user + 100 signup bonus per activated user.
  • Discount rate: 10% annually with monthly discounting at 10%/12.

Assumptions on timing (state clearly if you choose differently):

  • CPA paid at acquisition (month 0).
  • Signup bonus and annual fee occur in month 1.
  • Spend, rewards, credit losses, and servicing occur monthly (months 1–12). Credit losses are spread evenly over the 12 months.
  • Servicing cost applies to all acquired users (activated and not). If you assume it applies only to activated, multiply by 80% instead.

Tasks

  1. Compute the 12-month NPV per acquired user and the total NPV for the 50,000-user cohort. Show all components separately: interchange, rewards, fees, servicing, credit loss, and acquisition costs (CPA and signup bonus).
  2. Holding other assumptions fixed, derive the break-even CPA.
  3. Sensitivity: Recompute NPV for (a) interchange ±0.3 percentage points, and (b) credit loss ±$10 per activated user. Identify which variable the decision is more sensitive to and give a recommendation.
Loading comments...