🌏 中文版
Today's Focus
In Metrics & Analytics interviews, naming a north star metric is just the pass bar. What actually separates candidates is whether they volunteer where things break if the whole company chases that number. Interviewers often bake a hidden resource-allocation twist into the prompt — prioritizing A implicitly means sacrificing B — precisely to see whether a candidate can only pick a metric, or can name the trade-off in the same breath.
This matters because real product decisions are rarely "should we optimize this metric" yes/no questions. They're allocation questions: is it worth moving resources and user attention from one place to another? Being able to state a north star metric and its trade-off together is a direct signal that a candidate has actually governed metrics inside a real organization, not just memorized a framework.
Framework Cheat Sheet
The Three-Condition North Star Metric Check
When screening north star metric candidates, run each one through three conditions instead of picking whichever number sounds most important:
| Condition | Question to ask | Red flag if it fails |
|---|---|---|
| Reflects core value | When this number goes up, does the user actually get more value? | You picked a number that's convenient to report but users don't feel (e.g., raw impressions) |
| Leading indicator | Does it move earlier than lagging indicators like revenue or retention? | The metric only moves once a quarter — too slow to be a daily team signal |
| Actionable by the team | Do the product and engineering team's day-to-day decisions directly move this number? | The metric is mostly driven by external forces (market, seasonality) that the team can't touch |
How to use it in interviews: When you get a "define a north star metric for X" prompt, throw out one or two candidates, then screen them live against these three conditions out loud. Naming which condition a candidate is weak on demonstrates your reasoning process better than just announcing an answer.
Guardrail-Metric Pairing
Once you've picked a candidate north star metric, immediately pair it with a guardrail metric — the number that breaks first if the north star is over-optimized:
- Find the zero-sum relationship: Is this metric's rise built on reallocating a finite resource (user time, engineering capacity)?
- Name what's being sacrificed: If resources really are being pulled from elsewhere, what exactly is on the losing side — another feature, another user segment, another team's goal?
- Attach a number that turns red first: Give the sacrificed side a concrete metric, and define a threshold where crossing it means stop and reassess instead of celebrating the north star's rise.
How to use it in interviews: The most common follow-up is "what if this metric keeps rising but another one starts dropping?" Volunteering the guardrail metric up front answers that follow-up before it's asked, which reads as more convincing than reacting to it defensively.
Today's Practice Question
The Question
"Define a north star metric for Instagram Reels, and explain the trade-off if you prioritize Reels over Stories."
(Source: Meta PM analytical-thinking interview question, from the Meta Product Manager Interview Guide, aced.io (formerly Exponent))
Breaking It Down
- Clarify the question: Ask the interviewer what stage Reels is at (early adoption push vs. mature format that needs to prove revenue), and what "prioritizing Reels" concretely means — engineering resources, ranking/distribution weight in the algorithm, or ad inventory allocation. Each resource implies a different trade-off.
- Define the users: Split into creators and viewers, then split each again — core users who use Stories for daily contact with close friends, versus browsing-mode users pulled in by the recommendation algorithm to kill time on Reels. Shifting resources to Reels affects these groups in different directions.
- Structure the analysis: Run candidate metrics through the three-condition check. "Total Reels watch time" reflects value but isn't leading enough. "Completed Reels plays per DAU" satisfies all three conditions at once — it reflects value (you have to finish watching), it's leading (moves earlier than retention), and it's actionable (both the ranking algorithm and content supply directly move it) — making it the stronger candidate.
- Propose the trade-off: A user's total time in the app is zero-sum. If Reels watch time rises without total app time rising, something else is getting squeezed out — Stories posting and viewing is the most direct candidate. Pair the guardrail: attach "Stories posts per user" to "completed Reels plays per DAU." If Stories posting rate drops below a threshold, it signals that core users' daily contact habit is eroding, not just that attention moved around.
- Define success: Success isn't "the Reels metric keeps rising." It's "the total in-app time pool grows, and Reels is capturing newly added attention rather than attention cannibalized from Stories." Monitor whether overall DAU and total app time per user are growing in step — if only the Reels metric rises while total time stays flat and Stories drops, that's one metric's shine masking a loss elsewhere.
Sample Answer (how you'd actually say this in an interview)
Clarifying and framing: "I'd first confirm what stage Reels is at — still in the adoption-push phase, or mature enough that it needs to start proving revenue — because that decides whether the north star should lean toward reach-and-adoption or depth-and-monetization. I'd also confirm what 'prioritizing Reels' concretely means: ranking weight, engineering investment, or ad inventory. Those three resources imply completely different trade-offs."
Candidate metric and trade-off analysis: "I'd propose 'completed Reels plays per DAU' as the candidate north star, since it satisfies reflects-value, leading, and actionable all at once. But I wouldn't stop there — a user's time in the app is zero-sum, so if Reels watch time rises without total app time rising in step, that time is likely being pulled from Stories. I'd pair this candidate with a guardrail: Stories posts per user, because a drop there signals core users' daily-contact habit eroding, which hurts long-term stickiness more than a simple shift in watch time."
Defining success: "So I wouldn't measure success by whether the Reels number went up alone — I'd watch three numbers together: completed Reels plays per DAU, total app time per user, and Stories posting rate. Only if total time also rises and Stories posting rate doesn't meaningfully drop can I say Reels is capturing newly added attention, rather than trading Stories' existing user experience for a better-looking Reels number on paper."
Self-Check List
Use this table to check whether your answer missed anything:
| Check item | Covered? |
|---|---|
| Clarified which specific resource "prioritizing Reels" refers to, before answering | |
| Screened north star metric candidates against the three conditions (reflects value, leading, actionable) instead of just announcing an answer | |
| Named the zero-sum relationship and identified exactly who is sacrificed | |
| Volunteered a guardrail metric rather than waiting for a follow-up question | |
| Defined "success" by watching both the north star and the guardrail metric, not just whether the north star rose | |
| Bonus: named an overall metric to monitor (e.g., total app time) to distinguish "newly added" attention from "reallocated" attention |
Today's Case
Meta: Instagram shifted over half its ad inventory to Reels — a real number that demonstrates what "prioritizing a resource" actually costs
According to data from market intelligence firm Sensor Tower, more than half of all Instagram ad impressions ran on Reels in 2025, up sharply from 35% in 2024. In the U.S., Reels' share of total time spent on Instagram also rose from 37% in 2024 to 46% in 2025. The same data shows Instagram's daily active users grew 2% year-over-year, driven primarily by Reels usage. In other words, Meta didn't just shift ranking weight toward Reels — it reallocated ad inventory along with it, betting that the time Reels captures represents newly added attention from a growing overall pool, rather than attention simply traded away from Stories and Feed's existing user experience.
Interview relevance: This case turns today's practice question from a hypothetical into a real resource-allocation decision — Meta itself is answering "what gets sacrificed when you prioritize Reels." Whether overall daily active users grow in step is exactly the guardrail-level metric interviewers want candidates to volunteer. When answering this type of question, you can cite this case directly to show that large companies care less about a single format's number and more about whether the overall pool is being diluted.
Further Reading
- Meta Product Manager (PM) Interview Guide | Sample Questions (2026) — the original source of today's practice question, with the full breakdown of each Meta PM interview round and real follow-up patterns
- North Star Metric: Benefits, Challenges & Tips — extended explanation of the three-condition NSM check, with real north star metric examples from multiple companies
- Most of Instagram ads now run on Reels — the full data report behind today's case, citing Sensor Tower's 2025 figures for Reels' ad share and time-spent share
References
- Meta Product Manager (PM) Interview Guide | Sample Questions (2026) — source of today's practice question on the Instagram Reels north star metric
- Most of Instagram ads now run on Reels — source of today's case: Meta's 2025 Reels ad share and time-spent share figures
- North Star Metric: Benefits, Challenges & Tips — source for today's framework cheat sheet: the three-condition NSM check and real company examples
Loading...