A practice prompt we wrote. No company or candidate report names it, so it carries no company tag.
How to answer
Answer as if you are speaking to the product manager on the call. What the answer shows is whether you trade scope in the open or absorb it quietly. Leo Mehr’s August 2025 post on Ramp’s engineering blog says FDEs must be able to give an enthusiastic yes to rational customer requests while pushing back against unreasonable ones. Source 1Forward Deployed Engineering (Leo Mehr, Director, Engineering)PublisherRamp Builders (engineering blog)Source typecompany blog This answer is how you tell the two apart in front of the customer.
On the call, in this order:
- Status before the new ask. Open with where the pipeline stands and its date.
- Find the decision behind the request. “Who opens this dashboard, and what do they decide when they do?” A request tied to a real decision gets a yes and a date. “Nice to have” goes on the list with no date.
- Show the trade, in days. Don’t argue about whether it is small. Name what it depends on and which date moves if you start it now.
- Show the change log. Every request since kickoff, with its estimate and status. The log names the pattern so you don’t have to accuse anyone.
- Ask the person who owns the date to decide. Give options with a recommendation. If the product manager can’t move the date, the sponsor decides, in writing, this week.
- Close with the decision out loud and send the recap the same day.
- Stop the next one. Agree on an intake rule with the sponsor: new requests go into the log at the end of the call, you size them before the next call, and the sponsor ranks the log every other week. Ask who is asking him for these. If it is his own stakeholders, give them a view of the pipeline’s progress so their requests have somewhere to go.
Say the trade as a sentence: “If I start this now, cutover moves from Thursday to next Wednesday.” Then stop talking.
The polite yes is the trap: you build the dashboard at night, the pipeline slips, and the customer learns about a trade that was made for them. The opposite mistake is a flat no that ignores why he keeps asking.
The same call at an AI lab. There the request sounds like “can the assistant also handle refunds?” The same moves apply, and the trade has scope nobody mentions: a new intent needs its own eval cases, a bar it must clear, and a regression run over the intents that already work before it ships. Size it in days with that included.
Follow-ups
What the interviewer may ask next, once your first answer is on the table.
- He says the new dashboard is ‘small’. How do you show the trade?
- The sponsor is not on the call. How do you get a decision?
- Write the change-log entry for this request.
Where answers go wrong
- Absorbs each request quietly and misses the pipeline date.
- Argues about whether the request is small instead of showing what it depends on and which date moves.
- Says no without asking what decision the dashboard serves, so the product manager brings it back next week, louder.
Answer this in two minutes
Write the answer you would say out loud. The clock starts with your first word.
Compare with the model answer
Model answer
The setting: a grocery chain, a pipeline loading store sales into their warehouse, cutover on Thursday. Marcus, the product manager, has asked for a new dashboard on every weekly call since kickoff, and this week’s is the fourth. I’d have the change log on screen before he joins.
Status first. “Before new items: every store is loading except the two on the old register system, and we’re on course for Thursday’s cutover.”
The decision behind it. He asks for a promotions dashboard. I ask, “Who opens it, and what do they do with it?” He says category managers use it on Monday mornings to pull failing promotions. That is a real need tied to money, so it gets a date, not a no, and I say the yes before the trade: “Yes, that’s worth building, and I want your category managers to have it. The only question is when, so let me show you what it touches.” I also ask who has been asking him for dashboards. It’s the category managers, because they can’t see when their data will arrive. That goes in the recap too.
The trade. He says it’s small. I don’t debate size; I show what it touches: “It reads the promotions table, which the pipeline doesn’t load yet. So it’s the promotions load, then promo-code mapping and a reconciliation check against finance, then the dashboard. If I start it this week, cutover moves from Thursday to next Wednesday.”
The log. “This is the fourth request since kickoff. Each one made sense. Together they’re about two weeks of work, and none is in the signed scope.” Said flatly, that names the pattern without blaming him.
Options and a recommendation.
- A: Pipeline first. After cutover the queue runs in order: last week’s inventory-aging report, then promotions, which lands the week after next.
- B: Dashboard now. Cutover slips from Thursday to next Wednesday.
- C: Pipeline first, and promotions takes the inventory-aging report’s slot right after cutover, landing next week. Inventory aging moves back a week. Cutover holds.
“I recommend C, since you’ve told me the Monday promotions review costs money every week it’s missing. Not B: a dashboard over a half-loaded pipeline shows numbers finance will dispute, and that costs more trust than a week.”
The owner. If he wants B, it isn’t his date or mine. “Thursday is the date Priya gave her CFO, so moving it is her call, and so is the queue order. I’ll send her the options today and copy you.” The entry I send:
CR-4 Promotions dashboard
From Marcus (PM), weekly call, Tue
Need Monday review of failing promos
Deps promotions table (not in scope)
Size 4 days: promo load, mapping,
finance recon, dashboard
A After cutover, behind CR-3;
live the week after next
B Start now; cutover moves
Thu -> next Wed
C After cutover, ahead of CR-3;
live next week; CR-3 slips
a week
(recommended)
Owner Priya (VP merch, sponsor)
Due Decision by Wed noonClose. “So: cutover Thursday, the promotions dashboard logged as CR-4, and a decision from Priya by Wednesday. From next week, new requests go straight into the log at the end of the call, I size them before the next one, and Priya ranks the log every other Friday. I’ll also send your category managers a daily load status, so they can see when their data lands.” Then I send the recap within the hour.
If Marcus is right that the dashboard matters more than the date, Priya will say so and B happens with everyone’s eyes open. What I won’t do is build it at night and let Thursday slip without a decision.