Justify building a new feature with evidence

Q: Justify building a new feature with evidence

This question evaluates a data scientist's product analytics, experimental design, and strategic decision-making skills, covering market sizing, competitive analysis, privacy-aware data and infrastructure assessment, risk identification, and A/B testing planning.

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 Onsite 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

Case Prompt: 10-Minute Go/No-Go Recommendation for a New Feature

You are the data science lead supporting a large-scale consumer messaging product. Your VP asks you to justify building a new product feature from scratch and present a go/no-go recommendation in 10 minutes.

To make the discussion concrete, assume the candidate feature is "Smart Reply Suggestions" in 1:1 chats (the app shows 1–3 short, tap-to-send replies inferred from the last message). You may limit scope to English and mobile clients.

Address the following:

1) External Analysis

Market sizing: estimate TAM, SAM, and SOM, and outline assumptions and formulas.
Competitive landscape: identify at least 3 comparable offerings, what they do well/poorly, and likely defensibility.
Market signals: cite at least three signals you would gather and how you would quantify each (data sources, metrics, thresholds).

2) Internal Assessment

Capability fit: engineering/ML feasibility (e.g., on-device vs server inference), latency, cost, and quality.
Data availability and policies: what data is needed; constraints due to encryption, privacy, and user consent.
Risks: integrity, misuse/abuse, safety, localization, and regulatory risks.
MVP proposal: minimal viable launch surface (languages, platforms, entry points), and key dependencies (teams/systems).

3) Evidence & Decision Plan

Pre-launch work: offline evals, prototypes, research; how you would estimate impact before an A/B test.
Experiment plan: primary success metrics, guardrails, segments, and ramp strategy.
Milestones and thresholds: define what success looks like at 6 weeks and 6 months, and the exact decision thresholds to proceed, pivot, or kill.

Deliver a crisp recommendation (go/pilot/no-go) supported by the above.

Justify building a new feature with evidence

Quick Overview

Justify building a new feature with evidence

Case Prompt: 10-Minute Go/No-Go Recommendation for a New Feature

1) External Analysis

2) Internal Assessment

3) Evidence & Decision Plan

Write your answer