# Decide confidence level and forecast video views ## Part A — Choosing 95% vs 99% confidence level You are running an A/B test and must choose the confidence level for hypothesis testing / confidence intervals. **Question:** How would you decide between using a **95%** confidence level vs a **99%** confidence level? In your answer, address: - What business and statistical trade-offs change when moving from 95% to 99%? - How this affects **Type I error (false positives)**, **Type II error (false negatives)**, required **sample size**, and **time-to-decision**. - When you would prefer each choice (give realistic product scenarios). - Any adjustments you would consider for **multiple testing** (many metrics/segments) or sequential peeking. ## Part B — Predicting the number of video views You want to predict how many times a video will be watched. **Question:** Describe how you would forecast/predict **video watch counts (views)**. Please cover: - What exactly is the target (e.g., views in next 24 hours/7 days, lifetime views) and at what granularity (per video, per creator, per country)? - What data you would use (exposure/impressions, recommendations, follower graph, seasonality, content features, recency). - What modeling approach you would start with (simple baselines → more complex), how you would evaluate it (backtesting, metrics), and key failure modes (cold start, bots, non-stationarity, viral shocks). ### Constraints & Assumptions - Preserve the scope, facts, inputs, and requested outputs from the prompt above. - If the prompt leaves a detail unspecified, state a reasonable assumption before relying on it. - Keep the answer interview-ready: concise enough to present, but concrete enough to implement or evaluate. ### Clarifying Questions to Ask - Clarify the business objective, unit of analysis, time window, exposure definition, and primary metric. - State assumptions about instrumentation, randomization, sample size, and data quality. - Separate descriptive analysis from causal claims. ### What a Strong Answer Covers - A metric framework with primary, guardrail, and diagnostic metrics. - A credible analysis or experiment design with clear assumptions and bias checks. - SQL/statistical logic for segmentation, variance, confidence, and data validation where relevant. - An actionable recommendation that explains trade-offs and next steps. ### Follow-up Questions - What sanity checks would you run before trusting the result? - How would you handle novelty effects, seasonality, or selection bias? - What decision would you make if metrics disagree?