Capital One Interview Questions — Page 37

Capital One·Data Scientist·Technical Phone Screen

Jul 2026

Capital One data scientist interview with a scenario-based question about flight delays. Pretty open-ended, which I wasn't expecting from a finance company. They wanted you to think like an actual analyst, not just recite textbook stats.

You're a data scientist brought in to investigate flight delays for a major airline. How would you quantify and model the delays using historical departure and arrival data, and how would you determine whether a new mitigation strategy has actually made a statistically significant difference?

“This one sprawled in a direction I didn't expect.”

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 was a beast. They basically want your entire career condensed into a tight, metrics-driven story with trade-offs and reuse baked in.

What is your single most significant professional achievement in the past three years? Walk through the context, the measurable target, the constraints you were working under, the key decisions and risks you took, how you measured success, the before and after metrics, trade-offs you accepted, and whether anything from that project carried over to later work.

“This question wrecked my pacing.”

View Post

Capital One·Data Scientist·Onsite - Behavioral / Leadership

Jul 2026

Behavioral screen for a Data Scientist role at Capital One. Just the one question from what I can tell, pretty standard opener to kick things off.

What's the most significant accomplishment of your professional career, and why does it actually matter?

“I had something ready but kept second-guessing whether my numbers were impressive enough for a finance company.”

View Post

Capital One·Data Scientist·Technical Phone Screen

Jul 2026

Capital One data scientist interview that was basically a stats exam. Three heavily computational questions back to back, all requiring you to show your work. Not a vibe check, not a case study, just raw probability and inference.

You have an A/B test comparing two variants on conversion rate. Variant A got 410 conversions from 5000 visitors, Variant B got 470 from 5100. Compute a two-sided 95% confidence interval for the difference in proportions using both the normal approximation with unpooled standard error and the Wilson-Newcombe method. Is B statistically different from A under each approach? Also, if you were running 3 variants against A and needed to control family-wise error at 5% via Bonferroni, what confidence level would you use per comparison, and does your conclusion about B change?
A sample of 40 users has a mean session time of 8.2 minutes and a sample standard deviation of 2.4 minutes. Compute the 95% confidence interval for the population mean using the t-distribution, and state the margin of error using the t value at 39 degrees of freedom.
What minimum per-group sample size do you need to estimate a proportion with 95% confidence and a margin of error no greater than 0.01, using the conservative p equals 0.5 assumption? Show the formula.

“This was a lot.”

View Post

Capital One·Data Scientist·Technical Phone Screen

Jul 2026

Capital One data scientist interview built around a single extended marketing analytics case for the Quicksilver card. The whole thing was basically one big scenario with layered follow-ups, which I didn't expect going in. Math-heavy but also surprisingly strategic toward the end.

Before running any ads for this credit card campaign, what key factors would you evaluate? Think about metrics, customer segments, and risks.
Using the given assumptions (100,000 viewers, $50,000 fixed cost, $0.10 per view, $500 revenue per approved account, 1.4% click-to-apply rate, 50% approval rate), calculate the profit or loss for the unskippable ad campaign.
If you switch to a skippable ad costing $0.30 per completed view, with 80% of viewers skipping, what minimum conversion rate is needed to match the profit from the unskippable scenario?
There's also a third option: run no campaign at all. How would you decide between the three options?
If you had more time, what additional data or analyses would you want to strengthen your recommendation?

“I went straight to click-through rate and conversion and kind of forgot to talk about customer segmentation at all until they nudged me.”

View Post

Capital One·Software Engineer·Technical Phone Screen

Jul 2026

Capital One data engineer technical screen focused almost entirely on gradient boosted trees applied to credit risk. Three meaty sub-questions back to back, no warmup, no small talk. Felt more like an ML engineer interview than a data engineering one, which threw me a bit.

How do boosted tree models handle missing values during training and inference, and when would you choose native handling over imputation?
What causes overfitting in boosted tree models and which hyperparameters or techniques would you use to control it?
For an imbalanced credit default prediction problem, which evaluation metrics would you use and how would you validate that the model generalizes well?

“I knew XGBoost learns a default direction for missing values at each split, so I led with that.”

View Post

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

Jul 2026

Case interview for a Data Scientist role at Capital One centered on a full product launch analysis for a vegan burger line. Heavy on unit economics and break-even math, which I wasn't expecting from a DS interview. Left feeling like I was in the wrong room for about half of it.

Given revenue and cost figures for an existing beef burger product, calculate its total profit, profit margin, and per-unit contribution margin.
Work out the unit economics for the new vegan burger: selling price, variable cost per unit, and contribution margin.
Calculate the break-even sales uplift needed so that cannibalizing beef burger sales doesn't reduce total profit, and separately calculate the break-even quantity to recover the fixed launch costs.
The analysis suggests roughly a 60 to 70 percent incremental sales lift is required to maintain current margins. Would you recommend launching the vegan burger, and why?
What future trends in costs, consumer demand, pricing dynamics, or regulatory and ESG pressures could make the vegan burger market more attractive over time?
What cost-reduction and revenue-growth levers exist to improve the economics, and how would you design a pilot to test whether the launch is viable?

“This part was fine.”

View Post

Capital One·Data Scientist·Technical Phone Screen

Jul 2026

Capital One Data Scientist interview that was basically one long SQL problem covering campaign analytics, cost modeling, and donor targeting. The question was dense enough that I had to slow down and think through the schema carefully before writing anything.

Given a donors, campaigns, contacts, and donations schema, write SQL to compute per-campaign and per-segment metrics for August 2025: total reached, unique donors who donated, conversion rate, gross donations, variable cost, fixed cost, and net revenue (gross minus total cost).
After computing per-campaign metrics, write a single-row summary query that picks whichever campaign (gala vs. online) had the higher net revenue.
Write a query to identify the top 10 prospective gala donors, ranked by total donations in the last 12 months, excluding anyone already contacted for the gala campaign.

“This took me longer than I expected to set up.”

View Post

Capital One·Data Scientist·Technical Phone Screen

Jul 2026

Capital One data scientist interview with a stats-heavy scenario question about airline delay data. Pretty technical for a phone screen, felt more like a mini case study than a standard interview.

You have flight-level delay data. Walk through how you'd statistically identify which factors are driving the delays, what models or tests you'd use, and how you'd validate your approach and interpret the results.

“This one had more layers than I expected.”

View Post

Capital One·Data Scientist·Onsite - Behavioral / Leadership

Jul 2026

Capital One Data Scientist interview with a deep behavioral question on failure that goes way beyond the usual 'what did you learn' fluff. They want numbers, root cause chains, postmortems, and proof the fix actually worked.

Tell me about a recent failure. Walk through the root causes (what was on you vs. external), any warning signs you missed, who was impacted and by how much financially, the postmortem you ran, and what specific control you put in place afterward to make sure it doesn't happen again.

“This is not a soft 'describe a challenge' question.”

View Post