Interviewer view · keep this screen to yourself

Partner case: should Jabuti Pay launch a credit card?

You run the case. Read the prompt, answer questions from the notes below, and share data only when the candidate asks for it or gets stuck. Score at the end.

Case timer

00:00

Target length:

1. Read the prompt aloud

Read it slowly, then pause. Let the candidate ask questions before they structure.

Jabuti Pay is a Brazilian digital wallet with 12 million monthly active users. The founder wants to launch a credit card to all users next quarter. The partner asks: "You are advising the board. Yes or no, and why?"

Format note: Partner-led: judgment first. The partner expects a view within a few minutes, then pushes on it twice.

2. Answers to clarifying questions

Give these answers only if the candidate asks. If they ask something not listed, give a sensible answer or say it does not matter here.

If asked: How many active users does the wallet have, and how many might take a card?
Answer: 12 million monthly active users. The founder expects about 5 percent to take a card within two years.
If asked: What does each card earn and cost?
Answer: Assume an average balance of BRL 2,000, income of 25 percent of the balance a year from interest and fees, funding cost of 12 percent a year, and operating cost of BRL 80 per card a year.
If asked: What share of balances do we expect to lose to defaults?
Answer: Nobody knows yet. Card lenders in Brazil often see high losses, and many of our users have short credit histories.

3. The hypothesis a strong candidate states

Listen for an early, testable guess like this one. It does not need to match word for word.

Card profits depend mostly on credit losses, so my hypothesis is that the answer turns on one number, the loss rate, and that we should launch small to learn it before launching big.

4. A model structure

Compare the candidate's structure with this one. A different split can be just as good if it is clean and fits the problem.

Launch the card now, later, or as a test?
- How many cards, and what each earns before losses
- Key: The loss rate that makes it break even
- What we know about our users' risk
- How to launch: all at once or a test first

5. The working, step by step

Each step shows how a strong candidate works it out. Share a new fact from it only when the candidate asks or is stuck, and let them do the math: the result in the dark box is what they should reach.

Step 1: How many cards
What a strong candidate does: Candidate: "At 5 percent of 12 million users, the card would reach:"
Cards issued: 12,000,000 × 0.05 = 600,000
Step 2: Earnings before losses
What a strong candidate does: Candidate: "Per card, income is 25 percent of BRL 2,000, minus 12 percent for funding and BRL 80 of operating cost:"
Margin per card before credit losses (BRL a year): 2,000 × 0.25 - 2,000 × 0.12 - 80 = 180
Step 3: The break-even loss rate
What a strong candidate does: Candidate: "So the card breaks even if we lose this share of balances each year. This is the number the whole decision rests on."
Break-even annual loss rate (% of balance): (2,000 × 0.25 - 2,000 × 0.12 - 80) ÷ 2,000 × 100 = 9
Step 4: The bad case
What a strong candidate does: Candidate: "If losses are 12 percent, each card loses BRL 60 a year. Across 600,000 cards that is:"
Annual result at a 12 percent loss rate (BRL): 600,000 × (180 - 2,000 × 0.12) = -36,000,000
Step 5: The good case
What a strong candidate does: Candidate: "At 6 percent losses, the same launch makes about the same amount the other way. So the launch could make or lose about BRL 36 million a year, depending on one number we do not know."
Annual result at a 6 percent loss rate (BRL): 600,000 × (180 - 2,000 × 0.06) = 36,000,000
Step 6: Pushback 1: the rival
What a strong candidate does: Partner: "Our biggest rival launched a card last year and grew users 40 percent. We cannot wait." Candidate: "Growth matters, but the rival's user growth does not tell us whether its cards make money. If our loss rate is 12 percent, faster growth means losing money faster. I would not change the answer, but I would move quickly on a test so we do not fall far behind."
Step 7: Pushback 2: a new fact
What a strong candidate does: Partner: "Our data team says our payment data predicts risk better than a bank's credit score." Candidate: "That is new and it matters: it makes a loss rate below 9 percent more believable. It does not prove it. So I would update the plan, not the answer: a test with the 50,000 users our model rates safest. Even if losses reach 12 percent, the test costs at most about BRL 3 million a year, which is a fair price for the one number the decision rests on."
Worst-case annual cost of the test (BRL): 50,000 × (2,000 × 0.12 - 180) = 3,000,000

The recommendation to listen for

At the end, say: "The CEO walks in. What is your recommendation?"

Do not launch to all users next quarter; launch a 50,000-card test now and scale only if losses stay below about 9 percent. First, the card breaks even at a loss rate of about 9 percent of balances a year, so everything depends on that rate. Second, the swing at full launch is large: about BRL 36 million a year of profit at 6 percent losses, or a BRL 36 million loss at 12 percent. Third, the test costs at most about BRL 3 million a year and checks the data team's claim directly. Main risks: a rival signs up the best customers first, and the safest test users may understate losses for the full base. Next steps: build the risk model, choose the 50,000 test users, and agree in advance the loss rate that triggers a full launch.

Risks a strong answer names: Test users are chosen as the safest, so their losses may understate losses for the full base; A rival may sign up the best customers first; Interest rates may change, which moves the funding cost.

Next steps: Build and back-test the risk model on past payment data; Launch to 50,000 users with low starting limits; Agree now the loss rate and time period that will trigger a wider launch.

Strong versus weak

A strong answer

Found the single number that drives the decision (the break-even loss rate), showed the swing in both directions, held the view against a growth argument that contained no profit fact, and updated the plan when the partner added a real fact.

A weak answer

Said yes because the rival grew fast, or said no without showing what loss rate would make it work, and did not offer a way to find out.

Score the candidate

Score each criterion from 1 to 5. A 2 or a 4 sits between the descriptions.

This case has no exhibit. Score Exhibit reading on how the candidate used the data you gave them: did they pick out the number that matters and say what it means?

Structuring

1: No clear structure, or a memorized framework that does not fit the problem.
3: A sensible structure, but with some overlap, a gap, or no clear priority.
5: A structure built for this problem, with no overlap and no gaps, a stated hypothesis, and a clear first step.

Math

1: Frequent errors, no units, or needed heavy help to finish.
3: Right answers but slow or messy working, or one error caught late.
5: Accurate and clearly narrated, with units, a sense check, and what the number means.

Exhibit reading

1: Read the chart aloud without finding the number that matters.
3: Found the key number but gave only part of the so-what.
5: Pulled the right number quickly and said what it means for the case.

Synthesis

1: Summarized what was done, with no clear answer.
3: Gave an answer, but with weak reasons or no next step.
5: Answer first, then two or three reasons with numbers, the main risk, and next steps.

Communication

1: Hard to follow: long silences, no signposting.
3: Clear most of the time, with some rambling.
5: Calm and concise, signposted each step, and checked in with the interviewer.

Total

0 out of 25

Score all five criteria to see the band and the feedback template.