Framework for Goal Setting & North Star Metrics

Metrics questions are one of the most common question types you'll face in a PM final round. They're also getting harder. Generating a list of plausible KPIs is easy now. What interviewers are actually testing is whether you can connect a metric to strategy, defend why it's the right one, and navigate tradeoffs live.

This lesson walks through a four-step framework for answering any metrics question. Later lessons go deeper on specific metric types, north star selection, and tracking and guardrail metrics, but this is the foundation that everything else builds on.

What interviewers are looking for

Strategic grounding: Can you connect the product's metrics to the company's mission and business model, not just the feature in isolation?
Stage awareness: Do you understand what kind of product you're measuring, and what the right metric focus is for that stage?
A defensible north star: Is your chosen metric tied to the goal, with a clear argument for why it beats the alternatives?
Tradeoff fluency: Can you identify where your metric might mislead you, and what you'd watch alongside it?

The 4-step framework

The 4 Step Metrics Framework

Step 1: Scope

What the interviewer is scoring here: Do you check your assumptions before building an answer on top of them?

Start by confirming your understanding of the product and what you've been asked to measure. State it briefly and invite a correction if you've misread the question.

A simple template: "Just to make sure we're aligned: my understanding is that [product] is [brief description]. Are there any constraints I should know about, like whether we're looking at [parent company/specific feature/particular user segment]?"

This step exists because scope questions often have hidden ambiguity. If you're asked to define success for Airbnb, are you measuring the entire company, or just short-term rentals? What about Experiences? Getting that wrong early means you're solving the wrong problem.

Step 2: Mission

What the interviewer is scoring here: whether you think about metrics as strategy, not just measurement.

After scoping, establish the strategic context in three layers before you touch a single metric.

The company's mission. State it briefly and explain how the product supports it. If you don't know the exact wording, make a reasonable assumption and flag it. What matters is that your metrics ladder up to something real.

The product's role. What does this product do for the company and for its users? What job is it hired to do?

The lifecycle stage. This determines your primary metric focus. New products are validating whether anyone wants this; the focus is on adoption and early retention. Growth products are scaling; the focus is on acquisition and engagement. Mature products are holding the base and deepening value; the focus is retention and monetization. Declining products are managing attrition and thinking about what's next.

Name the stage out loud, and name the metric bucket it implies. This sets up everything that follows.

Step 3: User actions

What the interviewer is scoring here: do you think about the full product ecosystem, or just one type of user?

Before brainstorming specific metrics, do a quick sketch of who uses this product and what they're trying to accomplish. For most products, there are three lanes:

Creators: the users who supply the experience (hosts listing properties, sellers listing items, creators publishing content)
Consumers: the users who receive or consume that experience (guests booking stays, buyers purchasing, audiences watching)
The company: with its own goals, is usually tied to the mission and the business model

The goal here is not a full user journey. Keep it to two or three sentences per lane. You're building a behavioral map that your north star will need to speak to.

The best north star metric represents value for both user types, not just one. If your metric only captures consumer behavior, you're missing half the marketplace. If it only captures transactions, you might be missing quality signals. A strong north star speaks to creators, consumers, and the company's underlying goal simultaneously.

Step 4: North star

What the interviewer is scoring here: can you turn strategic logic into a specific, defensible measurement choice?

Now, brainstorm for a north star metric using the metric bucket you named in Step 2 and the user map you built in Step 3.

A strong north star has three properties:

Customer value: It represents both user types, achieving something meaningful
Business value: It maps to the company's mission and model
Team value: The product team can actually move this metric through their work in a reasonable timeframe

Generate a few candidates, reason through each one, and land on a choice. Be explicit about why you're choosing this one over the alternatives. That's where the answer earns its score.

"Given the mature stage and the transaction focus, my candidate north stars are:

Number of completed bookings per month, which captures both host and guest activity and directly maps to Airbnb's revenue model
Number of unique guests who complete a stay per month, which captures consumer demand but misses host health
Percentage of listed properties with at least one booking in the past 30 days, which captures supply-side health but misses demand depth

I'd go with the number of completed bookings per month as my north star.

Here's why: a completed booking represents host success (they earned income), guest success (they found and stayed somewhere worth booking), and Airbnb's business success (GMV drives revenue and marketplace trust). The team can influence it directly through search ranking, pricing tools, host onboarding, and demand marketing. And it's sensitive enough to move in a normal experiment cycle.

The tradeoff I'd watch: it doesn't tell me anything about quality. Bookings could be rising while guest satisfaction declines, which is a leading indicator of churn. So I'd pair this north star with a quality guardrail, like average guest rating per stay or the percentage of bookings that result in a complaint or refund."

A competent answer picks a reasonable north star and can explain why it connects to the product goal. It shows an understanding of lifecycle stage as a concept, even if that framing isn't applied explicitly to narrow the metric bucket before brainstorming.

A senior+ answer moves through all four steps in a clear, logical sequence and makes the reasoning audible at each transition. In the north star step, it generates multiple credible candidates, explains the rejection criteria clearly, and anticipates the conflicting-signal follow-up before it's asked — for example: "The scenario I'd watch for is bookings rising while host retention falls, which would tell me we're burning through supply rather than building a healthy marketplace." That kind of foresight is what separates a candidate who passes an analytical round from one who stands out.

Common pitfalls

Don't skip the stage diagnosis. It's the highest-leverage move in the answer. Candidates who jump straight to brainstorming metrics end up with a north star that floats free of any strategic logic, and that's exactly where interviewers apply pressure.

Answer the question that was asked. If the interviewer asked for one north star metric, land on one. Walking out with three options signals indecision, not thoroughness.

"Engagement" is not a metric. Neither is "retention" or "growth." Name the specific signal: daily active users who complete a booking, time spent per session, D7 retention rate. Then define how it's measured.

Make sure your metric can actually move. A small feature's impact will get lost in platform-wide DAU. Match the scope of your metric to the scope of what's being measured.

Think about both sides of the marketplace. The most common analytical mistake is optimizing for consumers and forgetting about supply. A north star that only reflects consumer behavior misses half the system.

Habit-based prep advice

Before you practice answering metrics questions, practice diagnosing products. For any product you encounter: apps you use, features you read about, companies in the news. Run through the first two steps quickly: What stage is this product at? What's the right metric bucket? Getting that diagnosis quickly is what makes the rest of the framework feel natural under pressure.

Overview

Mock Interviews & Practice