Evaluating a 15 % reduction in post‑card height

Quick Overview

Evaluates feed UX experimentation for reducing post-card height and diagnosing divergent geographic revenue outcomes. Strong answers measure scroll efficiency, viewability, engagement, ad revenue, device/network effects, and geo-specific decisions.

Evaluating a 15 % reduction in post‑card height

Company: Meta

Role: Data Scientist

Category: Analytics & Experimentation

Difficulty: medium

Interview Round: Technical Screen

##### Scenario: Designers want tighter cards to surface more content per scroll, hoping for higher session depth and ad load. You must measure scroll efficiency, user engagement, and investigate why U.S. revenue rises while Thailand’s drops. ​ ##### Question 1: If post card height shrinks 15 %, how would you measure impact? (Hint: visible posts per view, scroll speed, ad impressions) ##### Question 2: After launch, U.S. revenue up but Thailand down—what actions next? (Hint: local content mix, bandwidth constraints, ad pricing)

Overview: Evaluates feed UX experimentation for reducing post-card height and diagnosing divergent geographic revenue outcomes. Strong answers measure scroll efficiency, viewability, engagement, ad revenue, device/network effects, and geo-specific decisions.

|Home/Analytics & Experimentation/Meta
Meta logo
Meta
Jul 12, 2025
mediumData ScientistTechnical ScreenAnalytics & Experimentation
148
0

Evaluating a 15 Percent Reduction in Post-card Height

You own the feed UX for a social app. Designers propose shrinking each post card's height by 15 percent to show more content per scroll, aiming to increase session depth and ad load. After launch, U.S. revenue rises while Thailand's revenue falls.

Design the experiment, choose metrics, and diagnose the divergent geo outcome.

Constraints & Assumptions

  • Keep ranking constant unless the test explicitly changes ranking.
  • Instrument viewport exposure and ad viewability carefully.
  • Measure user experience and revenue together.
  • Diagnose geographic differences before broad rollout.

Clarifying Questions to Ask Guidance

  • Which post types, devices, and screen sizes are affected?
  • Does reducing height change creative cropping, text truncation, or ad rendering?
  • Did ad insertion, auction, or ranking change at the same time?
  • Are network conditions, content mix, and ad demand different by country?

Part 1 - Experiment and Metrics

How would you design the experiment and what metrics would you track?

What This Part Should Cover Guidance

  • Use user-level randomized A/B testing with treatment reducing card height by 15 percent.
  • Track visible posts per viewport, scroll speed, session depth, dwell time, engagement, hides, reports, ad impressions, ad viewability, CTR, revenue, latency, and retention.
  • Segment by screen size, device, network, country, content type, and user tenure.
  • Include guardrails for readability and accessibility.

Part 2 - Divergent Geo Outcome

After launch, U.S. revenue is up but Thailand is down. What would you do next?

What This Part Should Cover Guidance

  • Compare content mix, ad demand, CPM, fill rate, viewability, bandwidth, device mix, localization, and user behavior by geography.
  • Check whether more impressions lowered ad quality or auction prices in Thailand.
  • Examine latency and rendering issues on lower-end devices or slower networks.
  • Consider geo-specific rollout, rollback, or design adjustments.

Follow-up Questions Guidance

  • What if engagement rises but retention falls?
  • How would you measure whether content became harder to read?
  • How would you isolate ad-load effects from UX effects?
Loading comments...