Meta·Data Scientist·Onsite - Product Sense / Strategy·Senior

Senior

Jun 2026

Summary

Interviewed for a Data Scientist role at Meta and got a visualization-heavy product analytics question that I was not fully prepared for. The whole thing felt more like a product design exercise than a data science interview, which threw me a bit.

Questions Asked (1)

You're given a spaghetti chart of regional trends for a single metric and asked to redesign it into an executive dashboard. Walk through your layout choices, which 6-8 metrics you'd include for an early-stage B2B chat app, how you'd make regions comparable, your annotation rules, and your visualization guardrails. How does your redesign surface small-country trends that the original chart buries?

Product Analytics & MetricsProduct Sense & IdeationData Modeling

Author's notes

This was a beast of a question and I kind of underestimated it at first.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the executive audience and the decisions the dashboard should drive, then propose a layout that separates global summary from regional detail. Select 6-8 metrics that cover acquisition, engagement, retention, and monetization for an early-stage B2B chat app, and explain how you'd normalize regions for fair comparison. Finally, describe annotation and visualization guardrails, and show how small multiples or normalized views surface small-country trends.

Pro tip: Executives care about 'so what' — always tie each metric to a business decision or action, and use a consistent color scale across regions to avoid misleading comparisons.

1. Clarify Audience and Purpose

Ask who the executives are and what decisions they need to make (e.g., where to invest, which regions need attention). This shapes metric selection and layout.

2. Design the Layout

Propose a top-down structure: global KPIs at the top, then a regional comparison view (e.g., small multiples or a heatmap), and finally drill-down details. Ensure the most important metrics are visible without scrolling.

3. Select 6-8 Metrics

Choose metrics that cover the funnel: acquisition (new sign-ups), activation (first message sent), engagement (DAU/MAU, messages per user), retention (weekly retention), monetization (paid conversion, ARPU), and growth (net new active teams).

4. Make Regions Comparable

Normalize metrics by population, user base, or time zone; use per-capita or per-user rates; and consider indexing to a baseline (e.g., global average) to highlight relative performance.

5. Annotation and Visualization Guardrails

Annotate only significant events (product launches, outages) and use clear, consistent labels. Avoid clutter: limit colors, use accessible palettes, and ensure axes start at zero where appropriate.

Key Points to Mention

Use small multiples or faceted charts to give each region its own panel, preventing small countries from being overshadowed by large ones.
Normalize metrics (e.g., per-user, per-capita) and consider indexing to a global baseline to make regions comparable.
Select metrics across the AARRR framework: acquisition, activation, retention, referral, revenue, tailored to B2B chat (e.g., messages sent, active teams).
Annotation rules: only annotate major events, use consistent color coding, and provide context in tooltips or footnotes.
Visualization guardrails: avoid dual axes, use appropriate chart types (line for trends, bar for comparisons), and ensure colorblind-friendly palettes.
To surface small-country trends, use a sortable table or a ranked bar chart alongside the map, and highlight outliers with callouts.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.