Interview timeline:
5/14 recruiter call
5/27 first round — Product Mini Case
6/4 Power Day
6/6 recruiter notified me I passed
*There were two weeks between the recruiter call and the first round because the team wasn't scheduling interviews that particular week.
Product Mini Case (1 hour)
Mainly an interviewer-led product casing session. The recruiter told me in advance they'd pick one of three Capital One products to discuss: 1) Capital One Shopping, 2) Capital One Mobile Banking App, 3) Capital One Virtual Card Number. Mine was Capital One Shopping. The interviewer started by roughly asking what I thought the revenue streams were for the Capital One Shopping app, then moved into what I thought could be improved and how I'd test it (I said A/B test). Then he threw a chart at me and asked which email style I'd recommend based on the A/B test results. Finally he had me compare the UX design of two emails, critique them, and give improvement suggestions.
Overall it didn't feel too hard — as long as you can comment well on what you're shown and give some direction, you're fine.
Two hours after the first round, they told me I passed and set up my Power Day.
Power Day (9AM–1:30PM EST) — also over Zoom, four interviews, with a 30-minute break in the middle. Interviewers also try to let you go around the 50-minute mark so you can get some water.
The PM interview stages are: 1) Product Skill (BQ), 2) Product Case x2, 3) Product Discovery
Product Skill
Basically a deep dive on a tech product/feature I'd led before, answered STAR-style — not too hard, but you need to have thought through how you evaluated trade-offs at the time, lessons learned, and how you interacted with stakeholders, since that helps with the follow-ups.
I actually got feedback afterward saying I used too much technical jargon in this round. I thought, isn't this interviewing for a tech team?? But then I remembered the Power Day interviewers aren't necessarily from the org you're actually going to.
Product Case
There were two product cases, and one of them is tagged as HC (Hiring Chair) — meaning if there's disagreement, she's the tiebreaker.
Case 1 —
You are a product manager for a low-code/no-code website builder that lets customers easily create websites and add plugins. You're currently targeting restaurants that want to build an online presence.
Q1 — What customer segment do you think is the target user? I initially just said medium-size restaurants that don't have their own tech team, but the interviewer wanted me to first define what I thought the whole market looked like, so I went along with it and named small/medium/large restaurant customers plus some online ordering platforms that want integrated services.
Q2 — Asked how this makes money. Basically covered what the costs are and what revenue models could work (free w/ ads, freemium, subscription, etc.)
Q3 — Gave me Exhibit A and asked me to calculate 2-year profitability.
$200 onboarding fee
$20 monthly subscription fee
$300 one-time cost to onboard
$100 annual service upkeep
So profit = 200 + 20×24 − (300 + 100×2) = $180
Then he asked what about just year 1? (The insight is that profit is lower in year one because of the onboarding cost.)
Q4 — Asked whether I'd integrate with an external location system or build it myself if we wanted location integration (you can think through the reasoning yourself).
Q5 — Exhibit B gave engineering effort and upkeep numbers and asked which path was better.
Integration
2 engineers, 2 months, $120,000/yr salary, one-time onboarding cost of $100,000, subscription fee (to booking site) $15,000
2×2×120,000/12 = $40,000 (dev cost)
- $100,000 onboarding cost
- $15,000×22 (subscription fee)
= $470K
Build-yourself
4 engineers, 6 months, $120,000/yr salary, $5,000 monthly upkeep
4×6×10,000 = $240,000 (dev cost)
- 18×$5,000
= $220K
The key here:
- He asked for a 2-year horizon, but when calculating you need to subtract dev time — I asked a clarifying question, he gave me the wrong answer at first, then corrected himself and told me to subtract dev time.
- The engineers don't need ramp-up time and leave once development is done.
Q6 — Asked if there's anything Exhibit B didn't account for (e.g. adoption).
Q7 — Exhibit C
Gave a customer base of 100,000, with different strategies driving different growth in customers (integration 5%, self-built 3%). Then using the $180 from Q3 you could calculate how much extra revenue the additional customers from integration would bring, and you'd find it covers the extra cost.
Q8 — Final recommendation.
Case 2 (HC) —
Credit card partnership with, say, Uber.
This one had fewer exhibits, and I made a small mistake but got course-corrected. The goal was to increase engagement to drive spend.
He first asked what partnerships might make sense and which might be more profitable. Then he gave me three customer segments, with before-and-after spend on the merchant, and before-and-after spend on the card.
Segment 1: 100k users, before/after merchant spend $0/$0, before/after card spend $200/$200
Segment 2: 50k users, before/after merchant spend $10/$20, before/after card spend $400/$410
Segment 3: 50k users, before/after merchant spend $0/$40, before/after card spend $300/$500
(I don't quite remember the exact numbers for segments 2 and 3, but the differences are correct.)
Card margin revenue/spend = 1%
Capital One pays for the 20% discount offered at the merchant.
Then he asked what profitability means here, whether this makes sense, and why you'd do this if you're losing money (adoption). Follow-up: if a customer is worth $300 to Capital One's other business units, how many customers would you need to break even? Then he asked how likely you think that breakeven customer count is, and finally had me summarize a recommendation.
Product Discovery
As others have said, it is indeed the DMV improvement case. They show you three customer reviews and a DMV overview. You don't need to calculate anything here, but you're first asked to identify a number of problems, then define a problem statement, then pick one to go into solutioning and storyboard it, then summarize a recommendation for management and say whether you think it solves the problem statement.
Discussion
Loading comments…