Data Plan for Diagnosing an Engagement Metric That Falls 10% Every Month

Quick Overview

A product manager analytics question: engagement with a product has been falling about 10% every month, and the candidate must explain which data and metrics to examine, and in what order, to find out why. It tests metric definition, data validation, segmentation and structured root-cause reasoning.

Data Plan for Diagnosing an Engagement Metric That Falls 10% Every Month

Company: Google

Role: Product Manager

Category: Analytics & Experimentation

Difficulty: medium

Interview Round: Other

Engagement with the product you own has been dropping by roughly 10% every month. The interviewer wants to know which data and metrics you would look at to understand why, and in what order you would look at them. ```hint Two questions before the why Split the investigation into "is this decline real?" and "where exactly is it concentrated?" before you start listing possible causes. ``` ### Clarifying Questions - Which product is it, and what is the engagement metric: a count of active users, a per-user average such as sessions or time spent, or a ratio such as daily over monthly actives? - How many consecutive months has the decline lasted, and did it start as a sudden step or a gradual slide? - Is the drop visible on every platform and in every market, or was it first noticed in one dashboard or region? - Have related numbers such as total sessions, sign-ups or revenue moved in the same direction? ### What a Strong Answer Covers - A precise definition of the engagement metric and what a repeated 10% monthly decline adds up to over several months - How you would establish that the decline is real rather than an artifact of tracking or metric definitions - A breakdown of the metric and of the user base that narrows down where the loss sits - Internal and external factors lined up against the timing of the decline, each with the data source that would test it - A prioritized order of investigation and the result that would confirm or rule out each hypothesis ### Follow-up Questions - Every segment you check looks flat, yet the overall figure keeps falling. What could explain that, and how would you demonstrate it? - If the cause turns out to be a competitor winning your users, what data would show which users are most at risk? - Once you have found the cause and shipped a fix, how would you decide whether the fix worked?

Overview: A product manager analytics question: engagement with a product has been falling about 10% every month, and the candidate must explain which data and metrics to examine, and in what order, to find out why. It tests metric definition, data validation, segmentation and structured root-cause reasoning.

Google logo
Google
Feb 9, 2024
mediumProduct ManagerOtherAnalytics & Experimentation
0
0

Engagement with the product you own has been dropping by roughly 10% every month. The interviewer wants to know which data and metrics you would look at to understand why, and in what order you would look at them.

Clarifying Questions Guidance

  • Which product is it, and what is the engagement metric: a count of active users, a per-user average such as sessions or time spent, or a ratio such as daily over monthly actives?
  • How many consecutive months has the decline lasted, and did it start as a sudden step or a gradual slide?
  • Is the drop visible on every platform and in every market, or was it first noticed in one dashboard or region?
  • Have related numbers such as total sessions, sign-ups or revenue moved in the same direction?

What a Strong Answer Covers Guidance

  • A precise definition of the engagement metric and what a repeated 10% monthly decline adds up to over several months
  • How you would establish that the decline is real rather than an artifact of tracking or metric definitions
  • A breakdown of the metric and of the user base that narrows down where the loss sits
  • Internal and external factors lined up against the timing of the decline, each with the data source that would test it
  • A prioritized order of investigation and the result that would confirm or rule out each hypothesis

Follow-up Questions Guidance

  • Every segment you check looks flat, yet the overall figure keeps falling. What could explain that, and how would you demonstrate it?
  • If the cause turns out to be a competitor winning your users, what data would show which users are most at risk?
  • Once you have found the cause and shipped a fix, how would you decide whether the fix worked?