Data
40 bites tagged Data — interview questions with model answers, and 60-second explainers.
Linking form-component adoption to fewer a11y bugs
Define both variables, segment by adoption level, control confounders, watch correlation-versus-causation. designing a sound measurement, not just claiming impact.
How do you translate increase user engagement into a technical measurement plan?
Align with PM to define engagement, map touchpoints for events, pick a north star and guardrails, then draft technical schema. turning vague goals into metrics.
Client-Side vs. Server-Side Event Tracking
This tests your grasp of data integrity trade-offs. A great answer advocates for server-side tracking for critical events due to its reliability against ad blockers, using client-side only for supplementary UI events. A red flag is treating them as equal.
A key metric dropped 20%. How would you investigate?
This tests systematic diagnosis of critical issues. A great answer segments the drop (by region, platform), then traces data upstream from the dashboard to the source, correlating with technical metrics. A red flag is jumping to code before scoping the impact.
Explain cohort analysis and write a retention query
Tests your grasp of user behavior metrics. First, define a cohort (e.g., May signups). Then, explain why it's better than aggregate retention. Finally, outline a query grouping users by signup week and checking for activity. A red flag is a flawed query.
How would you capture UTM parameters for attribution?
This tests your grasp of the data lifecycle from capture to persistence. A good answer covers client-side parsing, cookie storage, and linking anonymous data to a user record upon sign-up. A red flag is forgetting to persist the data server-side.
DAU dropped 10% overnight. Is this a significant change?
Tests your use of statistical hypothesis testing on business metrics. Outline the process: state a null hypothesis (no change), choose a Z-test, calculate the p-value, and compare to an alpha of 0.05. A red flag is guessing causes before proving significance.
How would you A/B test a redesigned dashboard?
Tests translating a vague goal ('more engagement') into a concrete engineering plan. A good answer defines key metrics first, then outlines user bucketing, instrumentation, and statistical analysis.
Find leading indicators for long-term churn
This tests your ability to translate a business problem into a data investigation. A strong answer defines churned vs. retained cohorts, hypothesizes key early behaviors, and compares their frequency to find a leading indicator.
Why don't analytics and backend user counts match?
This tests your systematic debugging and understanding that 'user' has different definitions. A good answer first defines 'user' in each system, then investigates tracking implementation, timing differences, and filtering.
Client-side vs. Server-side Event Tracking: When and Why?
This tests your grasp of data reliability and security trade-offs. A good answer defines both, contrasts reliability (ad blockers) vs. implementation ease, and uses a critical event like "Payment Processed" to justify server-side's accuracy.
Translate 'increase engagement' into a technical measurement plan
This tests your ability to translate vague business goals into concrete metrics. First, clarify the goal with the PM. Then, propose specific, measurable proxy metrics (e.g., DAU/MAU, session length). Finally, outline the instrumentation plan.
Research Hypothesis: A Testable Bet on Reality
A research hypothesis is a testable bet about user behavior, not just a guess. It frames A/B tests by turning an observation into a statement you can prove or disprove. The footgun is confusing it with a theory; a hypothesis is a starting point, not.
Vanity vs. Actionable Metrics: Measure What Matters
Vanity metrics look impressive but don't inform decisions (e.g., total downloads). Actionable metrics tie to business goals and guide your next move (e.g., conversion rate). This helps product teams focus on real growth, not just impressive-looking charts.
Qualitative vs. Quantitative: The 'Why' and the 'How Many'
Quantitative research counts and measures ("how many?"), while qualitative research explores and understands ("why?"). Use quantitative for A/B tests to get statistical proof, and qualitative for user interviews to uncover motivations.
Time-Series Databases: Optimized for Data Over Time
A Time-Series Database (TSDB) is a database optimized for data where time is the primary key. It's the backbone for monitoring systems, IoT devices, and financial apps. The footgun is using a regular database, which can't handle the unique query load.
Data Augmentation: Getting More from Your Data
Data augmentation creates 'new' training data by making small, realistic changes to your existing data. It's used to fight overfitting in ML models when a dataset is small, teaching the model to generalize rather than memorize.
Data Labeling: Teaching Machines What to See
Data labeling gives raw data meaning so a machine learning model can learn. It's used to prepare datasets for tasks like object detection in images or sentiment analysis in text. The footgun: low-quality labels directly limit your model's performance.
Common Crawl: A Free Snapshot of the Entire Web
Common Crawl is a public library of the internet—a massive, free snapshot of web text and links. It's the raw material for training many LLMs and for academic research on web-scale data. The footgun: it's unfiltered, containing everything from facts to spam.
Instrumentation Plan: Your Analytics Blueprint
An instrumentation plan is the blueprint for what user actions to track. It's like deciding where to put security cameras *before* building, not after a break-in. It's essential for new features and for cleaning up messy analytics.
Experiment Knowledge Base: Don't Repeat Failed Tests
An Experiment Knowledge Base is your company's collective memory for A/B tests, centralizing hypotheses, results, and learnings. It prevents re-running old tests and surfaces past insights.
Metrics Layer: The Dictionary for Your Data
A metrics layer is the central dictionary for your company's numbers, defining what "Revenue" or "Active User" means once for everyone. It ensures teams and AI agents get consistent answers from a single source of truth, preventing conflicting reports.
Event-based Analytics: Tracking User Actions, Not Page Views
Event-based analytics tracks specific user actions—clicks, purchases, signups—not just page views. This model powers product analytics tools to map user journeys and measure feature adoption. The footgun is tracking everything, creating noisy, unusable data.
Data Retention Policy: Your Schedule for Deleting Data
A data retention policy is your company's official schedule for deleting data, not a plan to keep it forever. It's essential for legal compliance (like GDPR) and managing storage costs.
Get Data bites daily.
Five a day, five minutes, offline. With quizzes so it sticks.
Open testing — you’ll join as an early tester.