Product managers at an investment bank build tools for professional users: institutional clients who want to see their trades and positions, bankers preparing material for a deal, and traders who need answers in seconds. Prepare stories about saying no to a powerful user, working with compliance from the first day, weighing a regulatory deadline against a client request, and measuring success when each user matters a great deal. What to show here: client trust first, with compliance as a partner in the work. Each question below has what the interviewer is checking, an answer frame and a sample spoken answer to adapt.
How Goldman Sachs hires, in its own words, and its questions for every job →
“A senior user asked for a custom report built only for his team, and said it was urgent. I asked him to show me what he did with the current reports. He was exporting data each morning and rearranging the columns by hand. The real need was a saved layout, which other teams had asked for as well. I explained that I would not build a separate report, but I could deliver saved layouts, and I showed him a rough version within two weeks. He was not pleased at first, but he used it daily. The result was your result.”
Weak answers gave way because of the person's rank, or refused without finding out what was really needed.
“We had an old spreadsheet-based tool for tracking positions that only two people could maintain, and it broke every few months. A newer system existed, but a group of users refused to move. I sat with them and learned why: the old tool had one view they used every morning that the new one lacked. So I built that view first, with their help, and asked them to run both for a month. Then I set a switch-off date, sent reminders and stayed available on the day. Nobody lost a task they needed, and your result.”
Weak answers switch the old tool off on a date and let users complain, or keep both running for ever.
“I would map the journey step by step with the teams that run it, and measure how long each step takes and how long it waits. In processes like this, most of the time is usually waiting: a document requested twice, or a form sent back for one missing field. I would give the client one place to see what is needed and what is done, check forms for completeness before they are submitted, and reuse documents the bank already holds. No check is removed. I would measure the time to first trade and the number of requests sent back to the client.”
Weak answers suggest skipping or loosening checks, or redesign the form without studying where the time goes.
“I would sit with a few junior bankers while they prepare a real presentation and list what they repeat. I expect to see figures copied from spreadsheets into slides, charts rebuilt after each change, and numbers checked by eye. I would rank those tasks by time spent and by the chance of a mistake. My first choice would probably be linking figures in slides to their source, so that an update flows through and every number can be traced. That saves time and lowers errors. I would measure hours saved for each presentation and the errors caught before it goes to the client.”
Weak answers propose a grand new tool without watching the work, or chase speed and forget that a wrong figure in front of a client is worse than a late one.
“With so few users, I would look at each client, not at averages. For every client I would track how many of their people use the tool each week, which features they return to, and whether use is growing or fading. I would add direct feedback from regular conversations, and what the people who cover that client hear from them. For the bank, the question is whether clients who use the tool do more business or stay longer, though I would be careful about claiming the tool caused it. A client whose use drops suddenly deserves a call that same week.”
Weak answers count total logins, or plan large experiments that a few hundred users cannot support.
“A regulatory deadline cannot be bargained with, so that work comes first. But I would not accept the first estimate. I would ask compliance exactly what the rule requires and find the smallest change that meets it fully. That often frees some time. With that time, I would see whether a useful first part of the client's feature can be delivered. Then I would speak to the client myself. I would explain that a regulatory change has priority, show what they will get this quarter, and give a date for the rest that I am confident I can keep.”
Weak answers put the rule at risk to please the client, or tell the client nothing until the date is missed.
“I enjoy building for users who know their field deeply and will tell me exactly where the product falls short. In markets and banking, a product is part of a client relationship, so trust and accuracy matter more than clever features. Goldman Sachs serves companies, investors and institutions, and its tools carry that relationship. The pace of learning appeals to me too, because I would need to understand the products being traded to do the job well. In my past work on your project, I liked talking to every user personally. Here that is simply how the work is done.”
Weak answers use consumer product thinking unchanged, or speak about the firm's name and not the work.
“The fix is only the start. I would first find out exactly which clients opened the tool during that hour, which holdings were wrong and by how much, and whether anyone downloaded a report. Then I would bring in compliance and the teams who cover those clients, because how and what we communicate is a shared decision. Affected clients should hear from us directly, with the corrected values, before they find out themselves. After that I would look at why a bad input reached the screen. I would add a check that holds back values that move beyond a sensible range until someone confirms them.”
Weak answers consider the matter closed once the data is fixed, or contact clients without involving compliance.
No figures on this page: pay depends on the level, the city and the offer in front of you. When the offer comes, the free salary negotiation script helps with that part.
ClapAssist is an AI interview assistant for Mac and Windows. It listens to the interview on your computer and shows you what to say, in short lines you can read while you talk. Your live interview audio and screen are never stored. Your resume and notes are saved to your account so the app fills them in on any computer. It is hidden from screen share on every plan, including Free; only you can see it.