Capital One Interview Questions — Page 33

Capital One·Data Scientist·Onsite - Behavioral / Leadership

Jul 2026

Behavioral round for a Data Scientist role at Capital One. One meaty question about proactive helpfulness that sounds deceptively simple but has a lot of sub-parts hiding inside it.

Tell me about a time you went out of your way to help colleagues at work even though it conflicted with your own priorities. Walk me through who was involved, what tangible resources you built to make the help last, what you had to put on the back burner and why it was worth it, how you measured the actual impact, and what you'd change if you had to do it over.

“This question has way more layers than it looks.”

View Post

Capital One·Data Scientist·Technical Phone Screen

Jul 2026

Capital One Data Scientist interview that leaned heavily into credit card unit economics. The whole session was basically one long quantitative case with a few layers peeled back one at a time. More finance-y than I expected for a DS role.

Given a credit card with 1% rewards on all spend, compute the annual revenue, costs, and profit per customer assuming monthly spend of $500, a revolving balance of $2,000, 10% net interest margin, 2% interchange, $50 annual operating expense, and 3% loss rate on average balance. Based on these numbers, should the bank issue the card?
Solve for the minimum monthly spend S* that makes annual profit exactly zero, first assuming the revolving balance stays fixed at $2,000, then assuming 20% of annual purchase volume also becomes revolving balance on top of that $2,000 baseline.
Stress test the break-even spend calculation: if interchange drops to 1.6% and the loss rate rises to 5%, what happens to S* under both the fixed-balance and the balance-scales-with-spend scenarios?
In an incremental analysis for a rewards change on an existing customer, under what conditions can you ignore the outstanding balance term entirely, and what does that simplification mean for estimating the break-even level of new spend?

“This is the foundation question and I actually felt okay on it.”

View Post

Capital One·Data Scientist·Onsite - Product Sense / Strategy

Jul 2026

Capital One data scientist loop with a single massive product case question that covers basically everything: market framing, revenue modeling, metrics, prioritization, experiment design, and a finance pushback scenario. It's a lot to hold in your head at once and the sample size calculation at the end will catch you off guard if you haven't practiced it recently.

Pick a consumer digital app you love and explain the product to someone who has never heard of it. Cover the core jobs it does for users, who the main audience segments are, its top three competitors, and what makes it different from those competitors.
Walk through all the current revenue streams for the app you chose and explain the unit economics. Then propose one entirely new revenue stream and build out a six-month ROI model for it.
Define three success metrics for the new revenue stream you proposed. Each metric needs a clear numerator, denominator, and event plus time window definition. Also define two guardrail metrics with explicit thresholds.
Come up with six concrete product improvements for the app, then choose one to ship first and explain why you'd prioritize it over the others.
Design an A/B test for the product improvement you just prioritized. Specify the unit of randomization, who is eligible and who gets excluded, how assignment works, and what bias mitigations you'd put in place. Then compute the minimum sample size and estimated test duration using these inputs: 500,000 daily active users, 15% baseline conversion per user-day, expected absolute lift of 2 percentage points, 80% power, two-sided alpha of 5%, and a 1:1 split. Show your formula and state your assumptions.
Finance pushes back and says the improvement is too expensive to build. Walk through your cost-benefit model, identify the breakeven point, and explain what decision you'd make if interim results after two weeks are underperforming the minimum detectable effect.

“I picked Duolingo and immediately regretted it because the competitive landscape is kind of shallow if you don't prep it in advance.”

View Post

Capital One·Product Manager·Onsite - Product Sense / Strategy

Jul 2026

PM interview at Capital One with one product strategy question about Capital One Shopping. Pretty focused session, just the one question but it required knowing the product well enough to sell it internally.

How would you explain the value of Capital One Shopping to a new internal business stakeholder who isn't familiar with it?

“I fumbled the opening a bit because I started talking about the user-facing features before anchoring on why it matters to Capital One as a business.”

View Post

Capital One·Software Engineer·Technical Phone Screen

Jul 2026

Capital One coding screen for a software engineer role. One algorithmic question, pretty focused, nothing too wild.

Given a string, count how many length-3 substrings have the same first and last character (case-insensitive). The middle character can be anything.

“Not a hard problem once you see it.”

View Post

Capital One·Data Scientist·Technical Phone Screen

Jul 2026

This was a case-style interview for a Data Scientist role at Capital One that leaned heavily into energy investment modeling. Six questions deep, full of unit economics and sensitivity analysis. Not what I expected walking in.

Before building a financial model for a utility-scale renewables investment, what quantitative factors would you prioritize? List at least six, explain how an increase in each affects ROI, and describe how you'd measure or validate each one.
Given the base economics (capex $50M, fixed O&M $2M/year, variable cost $28/MWh, PPA price $40/MWh, max capacity 800,000 MWh/year), compute the minimum annual MWh output needed to achieve a 10% Year-1 ROI. If that requirement exceeds physical capacity, quantify the shortfall.
If regulation caps Year-1 output at 500,000 MWh, what PPA price (to the nearest cent) is needed to still hit a 10% ROI while holding costs constant? And separately, what is the maximum variable cost per MWh that still yields 10% ROI if price stays at $40/MWh?
Compare two mutually exclusive investment options (Solar at $25M capex, $2M fixed O&M, $0 variable cost, 300K MWh at $40/MWh versus Biomass at $25M capex, $1M fixed O&M, $30 variable cost, 1.1M MWh at $40/MWh). For each, compute Year-1 operating profit, simple payback period, and unit margin.
Which investment option would you recommend and why? Use unit economics, sensitivity to price and cost shocks, capacity risk, and operational complexity to justify your answer.
Give a 30-second executive summary (90 words or fewer) covering your recommendation, key assumptions, and the single biggest risk.

“This was the opener and I went way too broad at first.”

View Post

Capital One·Data Scientist·Onsite - Behavioral / Leadership

Jul 2026

Behavioral round for a Data Scientist role at Capital One. One question, but it had a lot of moving parts and I don't think I fully stuck the landing on the ROI piece.

Tell me about a time you proactively helped a colleague or team succeed without being asked. How did you decide it was worth your time, what did you do to avoid creating dependency, and how did it end up benefiting you or your team? How did you measure the return on that effort?

“The dependency part tripped me up.”

View Post

Capital One·Product Manager·Onsite - Product Sense / Strategy

Jul 2026

Capital One PM case interview built around a credit card merchant partnership with a company like Uber. Heavy on unit economics and segment math, with strategic follow-ups that pushed into incrementality, pilot design, and co-funding scenarios. Tougher than I expected going in.

Given three customer segments with different spending behaviors before and after a merchant partnership launch, is the partnership directly profitable for Capital One? Walk through the math.
If the partnership runs at a direct loss, why might Capital One pursue it anyway?
If each acquired customer generates $300 in value across other Capital One business lines, how many new high-value customers would be needed to break even on the partnership loss?
What would you actually recommend Capital One do with this partnership?
How would you estimate whether the spending lift you observed is actually caused by the partnership, versus spending that would have happened regardless?
How would you design a pilot for this partnership before a full launch?
How does the math change if the merchant agrees to co-fund half of the discount cost?

“I fumbled this at first because I wasn't sure whether the 20% discount applied to all merchant spend or only the new incremental spend.”

View Post

Capital One·Product Manager·Onsite - Product Sense / Strategy

Jul 2026

Capital One PM interview with a product sense question about reimagining a physical everyday object. Pretty open-ended, which I wasn't expecting from a financial services company.

Redesign a pillow with 10 new features.

“Ten features is a lot and I burned time trying to hit the number instead of thinking about whether the features were actually good.”

View Post

Capital One·Software Engineer·Onsite - Product Sense / Strategy

Jul 2026

Product analyst case for Capital One, built around a warehouse retailer with a food court open to non-members. Heavy on quantitative modeling and strategy, less behavioral than I expected. The case had a lot of moving parts and I don't think I covered all of them cleanly.

How would you define and quantify 'missed profit' from non-member visits that are limited to food court purchases, and what breakeven equation would you write for the food court?
What are the main business and measurement risks of keeping the food court open to non-members at a subsidized price?
Walk through how you'd evaluate a price increase for the hot dog combo, bringing production in-house, restricting access to members only, and fully shutting down the food court.
How would you estimate the causal lift in retail basket size attributable to food court visits, rather than just the correlation?
Design an experiment to test differential pricing between members and non-members at the food court.
How would you put a dollar value on a non-member converting to a paid membership after visiting the food court?
How does your analysis change if the food court is running at capacity and members are being crowded out by non-member visitors?

“This is where I got tangled.”

View Post