Growth Metrics & Unit Economics Questions

The core quantitative vocabulary of growth: activation, retention, referral and revenue metrics, growth-accounting frameworks such as AARRR, and unit economics including LTV, CAC, and payback period. Covers defining these metrics precisely and computing growth calculations that reveal whether growth is efficient and durable. The scope is metric definitions and economic math, not go-to-market execution.

HardTechnical

81 practiced

You need to avoid look-ahead/data leakage when building an LTV predictive model. Enumerate common leakage sources specific to growth metrics (e.g., using future spend, campaign tags applied post-conversion), explain how to detect leakage in features or model validation, and propose strict validation protocols to prevent it.

MediumTechnical

78 practiced

Design metrics for a referral program to increase virality. Define the primary metric or North Star, enumerate the submetrics required to understand performance (for example invites_sent, invite_accept_rate, referred-user-retention), show how to compute the viral k-factor, and outline an A/B test for a referral incentive including guardrails and fraud detection methods.

Sample Answer

Primary (North Star): Net New Viral Users per Week (NVU) — number of new users attributable to organic referral virality after accounting for paid/organic non-referral sources. This aligns growth impact to business outcomes.

Key submetrics (definition and purpose):- invites_sent (per referrer, total): volume of outreach.- invite_touch_rate = invites_opened / invites_sent: engagement with invite channel.- invite_accept_rate = invites_accepted / invites_sent: conversion to sign-ups.- referred_activation_rate = referred_users_who_complete_A / referred_signups: quality of referrals (A = meaningful first action).- referred_retention_d1/d7/d30: retention of referred users vs baseline.- referrer_conversion_rate = referrers_who_refer_again / referrers: repeat virality.- virality_lift = NVU / baseline_new_users: program impact.- cost_per_viral_user (if incentives paid).

Viral k-factor computation:k = (avg_invites_sent_per_user) * (invite_accept_rate) * (avg_referral_conversion_rate_into_active_user)More precisely:k = (Σ invites_sent / active_users) * (Σ invite_accepts / Σ invites_sent) * (Σ referred_activations / Σ invite_accepts)Interpretation: k>1 => exponential growth.

A/B test for incentive (treatment = higher reward):- Hypothesis: Treatment increases invite_sent and invite_accept_rate and yields k↑ without lowering referred_retention.- Randomization: user-level stratified by tenure and historical activity.- Metrics: primary — NVU lift; secondary — invite_accept_rate, referred_activation_rate, referred_retention_d7, referrer_repeat_rate, cost_per_viral_user.- Duration & power: run until minimum detectable effect (e.g., 5% lift) reached with 80% power; minimum 2 full viral cycles (e.g., 30 days).- Guardrails / stopping rules: stop if referred_retention_d7 drops >10% vs control or fraud signals exceed threshold or cost_per_viral_user > target CPA.- Segmentation: analyze by channel, geography, user cohort.

Fraud detection & prevention:- Signals: abnormal invites_sent distribution (heavy tails), very low-time-to-accept, many accepts from same IP/device, high invite_accept_rate with near-zero activity afterward.- Methods: rule-based filters (rate limits per user/IP), device fingerprinting, email/phone verification, clustering anomalies on invite patterns, supervised model trained on labeled fraud examples (features: invites/sec, distinct devices, behavioral features of referred users), and graph analysis to detect collusive groups.- Post-hoc validation: mark suspicious invites; run sensitivity analysis of k excluding flagged events.

Final note: report both short-term acquisition lift and long-term LTV of referred users; include dashboards showing funnel and k over time and a fraud-adjusted k.

HardTechnical

86 practiced

Compare survival analysis techniques (Kaplan-Meier, Cox proportional hazards) for modeling churn. Explain when to use each, how to handle time-varying covariates, and how to interpret hazard ratios in the context of product features or retention interventions.

EasyTechnical

81 practiced

Define Customer Lifetime Value (LTV / CLTV). Describe at least three calculation approaches (simple historical average revenue per user * average lifetime, discounted cash flows, predictive models), and explain pros/cons of each approach for a non-contractual e-commerce business versus a subscription service.

EasyTechnical

104 practiced

Explain customer acquisition cost (CAC) and customer lifetime value (LTV). Describe how you would compute each for Netflix, what assumptions are commonly made, which time windows to use, and why the LTV to CAC ratio matters for strategic decisions like marketing spend and content investment.

Sample Answer

Customer Acquisition Cost (CAC): average spend to acquire one paying customer. For Netflix compute CAC = (marketing + sales + promotional + onboarding costs) over a period ÷ new paid subscribers acquired in that period. Include free-trial conversion costs and attribution window (e.g., 30–90 days after an impression).

Customer Lifetime Value (LTV): discounted gross profit you expect from a customer over their relationship. For Netflix a practical formula:LTV = sum_{t=1..T} (ARPU_t × gross_margin_rate × retention_prob(t) / (1+discount_rate)^t)Where ARPU_t is monthly revenue per user (after discounts), retention_prob(t) comes from cohort survival curves, gross_margin_rate excludes content amortization if you treat content as investment (or include it per-user if allocating content costs), and T is horizon (e.g., 36–60 months or lifetime until survival ~0).

Common assumptions:- Stable churn/retention by cohort (or modelable decay)- Constant ARPU (or known price-path)- Constant gross margin (or modeled by geography/device)- Appropriate attribution window for CAC- Choice whether to include content costs in per-user gross margin or treat content as a capital investment

Time windows:- Use monthly granularity for subscription business; report annualized LTV as LTV_monthly × 12.- Cohort analysis across 12–36 months is typical; for longer-lived subscribers extend to 60 months or use parametric survival (Weibull/Exponential) to extrapolate.- CAC payback period: months until cumulative gross contribution equals CAC (important short-term cash metric).

Why LTV:CAC ratio matters:- Unit economics health: common rule-of-thumb LTV:CAC ≈ 3:1 (>=3 good, 1–3 needs optimization, <1 unsustainable).- Informs marketing budget: how much you can spend to acquire customers profitably and how quickly you recover spend (cash flow).- Guides content investment: if LTV supports higher acquisition spend, you can invest more in content; conversely, deteriorating retention (lower LTV) argues for product/retention fixes before growth-heavy content spend.- Strategic trade-offs: high CAC with long payback may be acceptable if balance sheet supports it and LTV is strong; sensitivity analysis (vary churn, ARPU, margin) quantifies risk.

As a data scientist, I’d compute cohort retention curves, estimate ARPU and margin by cohort/geography, model parametric churn for extrapolation, run DCF with multiple discount rates, and produce LTV:CAC and CAC payback dashboards with sensitivity scenarios to guide marketing and content strategy.

Unlock Full Question Bank

Get access to hundreds of Growth Metrics & Unit Economics interview questions and detailed answers.

Join thousands of developers preparing for their dream job.