Convince PM to Implement Duplicate Observation Tool
Overview
This question evaluates a data scientist's competencies in product analytics, experimentation design, causal impact estimation, stakeholder communication, and business‑impact quantification by framing a request to justify and size a Duplicate Observation Tool.
Meta
Aug 4, 2025, 10:55 AM
Data Scientist
Technical Screen
Analytics & Experimentation
0
0
Scenario
Meta is considering building a Duplicate Observation Tool (DOT) to detect malicious copy‑cat content (e.g., re‑uploads of the same media or near‑identical text that siphon engagement from originals). You need to make a business case and propose how to size impact.
Tasks
- How would you approach the PM to convince them this feature is needed and estimate its impact?
- Which existing data sources would you analyze to demonstrate the need for DOT?
- For each proposed data source, list key pros and cons.
Hints
- Define success metrics and guardrails.
- Estimate prevalence of duplicate posts and their exposure.
- Analyze user and creator complaints.
- Quantify revenue or engagement loss attributable to duplicates.
- Propose an A/B test (or quasi‑experimental) design to measure causal impact.
- Assess data coverage, quality, and bias across products (Feed, Reels, Video), regions, and languages.
Solution
Login required
Show