Product sense: user, problem, solution, tradeoffs

A good “design a feature” answer earns its solution — it names a specific user and their real need first, and volunteers the trade-offs last.

The idea

Every “design an X” question has the same winning shape. Pick a specific user segment. Name their underlying job-to-be-done — the thing they’re really trying to get done. Put two or three solution directions on the table before you commit to one. Then volunteer the trade-offs and edge cases of your pick.

The classic fail is jumping straight to a feature. It signals you solve before you understand — and interviewers are scoring your reasoning, not your idea.

Try it — run the answer

Prompt: Design a feature to help new parents on a streaming service.

user need breadth choose trade-offs

interviewer signal

Start by choosing who, exactly, this is for.

How it works

Answer in this order, and say the order out loud so the interviewer can follow your reasoning:

1. USER      pick a narrow, specific segment
             "new parents in the first 6 months", not "users"

2. PROBLEM   name the job-to-be-done underneath
             "find something calming to watch one-handed at 2am,
              without waking the baby" -- not "they want more content"

3. SOLUTION  breadth THEN depth
             list 2-3 directions, then pick one and go deep
             (jumping straight to one direction = the classic fail)

4. TRADEOFFS volunteer them; never claim "no downsides"
             cold start, metric cannibalization, consent/creepiness

When to use it

Situation	Why it fits / the trade-off
“Design a feature / product for X”	The shape signals structured product thinking. Costs 30–60 seconds of setup before any feature.
“Improve <existing product>”	Same shape — pick a user whose need is under-served, then generate directions.
“What’s your favourite product and why?”	Reverse it: name the user and job it nails, and the trade-offs it accepts.
You have a strong idea already	Resist opening with it. State the user and need first, or a great idea reads as a lucky guess.

Watch out for

Opening with the feature. “I’d build a kids’ mode” before naming a user is the single most common way these answers fail.
A segment that isn’t one. “Users,” “everyone,” or “people who like TV” give you nothing specific to design for.
Naming a surface want, not the job. “They want more content” is true of everyone; the real job is what they’re trying to accomplish in the moment.
Depth with no breadth. Committing to one solution instantly hides your range. Name two or three, then choose — and say why.
Claiming your idea has no downsides. Volunteering the trade-offs and edge cases is where senior candidates separate themselves.

Worked example

“Design a feature to help new parents on a streaming service.”

“Let me pick a specific user: parents in the first six months, often watching one-handed at 2am while feeding. Their real job isn’t ‘more content’ — it’s finding something calming they can start with almost no attention and one free hand, without waking the baby. A few directions: a hands-free / voice mode, a late-night wind-down row, and a sleep timer that pauses when you doze off. I’d go deep on the wind-down row because it’s the lowest-effort win. Trade-offs I’d watch: cold start for brand-new accounts with no history, whether it cannibalizes prime-time metrics, and that inferring ‘new parent’ from behavior needs consent so it doesn’t feel intrusive.”

Notice the order: user, then job, then breadth, then a chosen depth, then the trade-offs — volunteered, not dragged out.

Check yourself

The interviewer says “design a feature for commuters.” What’s your first sentence?

You’ve picked a solution direction. What earns the most points next?