Identify non-table data for feature demand

Q: Identify non-table data for feature demand

The question evaluates a data scientist's competency in triangulating product demand from non-table sources, defining and de-biasing signals, specifying metrics and guardrails, and designing evidence-driven experiments within the Analytics & Experimentation domain for a Data Scientist role.

Q: How do I approach Analytics & Experimentation interview questions?

Analytics & Experimentation questions require understanding of core concepts and practice. PracHub provides solutions with explanations to help you master analytics & experimentation interviews.

Q: What difficulty level is this interview question?

This is a hard difficulty Analytics & Experimentation question, commonly asked during Technical Screen rounds at Meta.

Q: What role is this question designed for?

This question is commonly asked for Data Scientist candidates at Meta during technical interviews.

Question

Evaluate Demand for a New "Group Call" Feature Using Non-Table Data and Experiments

Context

You are a data scientist evaluating whether to invest in a new "group call" feature for a large consumer communications product. Assume you cannot rely primarily on existing product usage tables (e.g., call logs, feature flags). Instead, you must triangulate demand and feasibility from non-table sources and then design experiments to validate demand.

Define "eligible users" as users who have access to calling in the app and belong to at least one group/thread with ≥3 members.

Task

Identify specific non-table data sources you would use and for each, describe:
- The signal you would extract.
- How you would quantify it (metrics/models).
- How you would de-bias it.
Discuss at least the following categories: (a) Qualitative channels: customer interviews, support tickets, sales call transcripts, app store reviews, community/social listening. (b) Market/competitive intel: pricing pages, public docs, analyst reports. (c) Operational data: infra costs, incident logs, capacity constraints. (d) User research/UX telemetry: task success, time-on-task, heatmaps.
Design an evidence-driven experiment plan to validate demand, including:
- A fake-door test (exposed to X% of eligible users), a waitlist, and a limited beta.
- For each, specify: hypothesis, eligibility, primary success metric (e.g., new weekly active group-call hosts), guardrails (e.g., crash rate, setup failure rate, CSAT), and minimum detectable effect (MDE) with rough sample size math.
- Provide a realistic timeline using "today" = 2025-09-01.
- Explain how you will segment results (e.g., team size, geography, plan tier), control for novelty effects, and ensure ethics (avoid dark patterns).

Identify non-table data for feature demand

Quick Overview

Evaluate Demand for a New "Group Call" Feature Using Non-Table Data and Experiments

Context

Task

Solution

Comments (0)