Unit Economics and Scaling Questions

Covers measuring and modelling the economics of acquiring and servicing customers and how those economics change as a business grows. Candidates should be able to calculate Customer Lifetime Value for cohorts using retention, spend per period, and margin assumptions; compute payback period and contribution margin per customer; and compare Customer Lifetime Value across acquisition channels and customer segments. Understand the relationship between Customer Lifetime Value and Customer Acquisition Cost and how that ratio informs sustainable growth. Expand analysis to unit economics beyond customers to units of product or transaction level, identifying fixed and variable cost drivers, per unit gross margin, and break even points. Reason about scale effects including economies and diseconomies of scale, what operational components break or become bottlenecks at higher volume, and how unit costs change with automation, capacity constraints, supplier pricing, fraud and support load. Be prepared to build simple spreadsheet models and run sensitivity and scenario analyses, propose operational and pricing levers to improve unit economics, and design experiments and metrics to track improvements over time.

HardTechnical

19 practiced

Write a Python function (pseudocode acceptable) to Monte Carlo simulate LTV for a cohort given distributions for monthly churn rate, ARPU, and payment processing fees. Describe inputs, outputs, and how you'd present simulation results (e.g., percentiles) to stakeholders.

Sample Answer

Approach: run N Monte Carlo cohort simulations; each simulation samples per-month churn, ARPU, and fee rates from given distributions, then simulates customer retention and cashflows month-by-month to compute cohort LTV (discounted or undiscounted). Aggregate simulation results to produce percentiles, mean, CI, and sensitivity.

python

import numpy as np

def simulate_cohort_ltv(
    cohort_size: int,
    months: int,
    n_sims: int,
    churn_sampler,      # function(size) -> array of churn rates (prob per month)
    arpu_sampler,       # function(size) -> array of ARPU per month
    fee_sampler,        # function(size) -> array of fee rates (fraction of ARPU)
    discount_rate=0.0   # monthly discount rate (e.g., 0.01 for ~12% annual)
):
    """
    Returns: dict with per-simulation LTVs and summary stats.
    """
    ltvs = np.zeros(n_sims)
    for i in range(n_sims):
        # sample monthly parameters (can be scalars or arrays of length `months`)
        churn = churn_sampler(months)       # shape (months,)
        arpu = arpu_sampler(months)         # shape (months,)
        fee = fee_sampler(months)           # shape (months,)
        surviving = 1.0
        ltv = 0.0
        for t in range(months):
            # expected paying customers this month = surviving * cohort_size
            revenue_per_customer = arpu[t] * (1 - fee[t])
            cash = surviving * revenue_per_customer
            discount = (1 + discount_rate) ** (t)
            ltv += cash / discount
            # update survival for next month (geometric retention)
            surviving *= (1 - churn[t])
        ltvs[i] = ltv  # total cohort LTV in currency units
    return {
        "ltvs": ltvs,
        "mean": float(np.mean(ltvs)),
        "median": float(np.median(ltvs)),
        "percentiles": {p: float(np.percentile(ltvs, p)) for p in [5,10,25,50,75,90,95]},
        "std": float(np.std(ltvs))
    }

Key points:- Inputs: cohort_size, months horizon, n_sims, three samplers (they can draw from parametric distributions: beta for churn, normal/lognormal for ARPU, beta for fee), discount_rate.- Output: per-simulation LTV array plus summary stats and percentiles.- Complexity: O(n_sims * months). Memory: O(n_sims).- Edge cases: ensure churn in [0,1], ARPU >=0, fee in [0,1]; handle zero churn or extreme churn; vectorize for performance.Presentation to stakeholders:- Show median LTV and 90% CI (5th–95th percentile) with a histogram and a cumulative density plot.- Table: mean, median, std, percentiles.- Waterfall: expected first-month revenue vs. attrition over time.- Sensitivity: tornado chart showing how +/- changes in mean churn, ARPU, fee affect median LTV.- Recommendations: highlight probabilities of LTV exceeding CAC thresholds and scenarios needing action.

EasyTechnical

35 practiced

List and explain three common fixed costs and three common variable costs you would include when calculating unit economics for a physical product. Then describe briefly how you would allocate shared overhead (e.g., HQ rent, product R&D) to units for reporting purposes.

HardTechnical

33 practiced

A BI dashboard shows LTV increasing but churn rate is creeping up in a specific region. As the BI analyst, how would you reconcile these conflicting signals and what additional metrics or decomposition would you present to stakeholders?

Sample Answer

Situation: The dashboard shows LTV rising while churn in one region is increasing — an apparent contradiction that could mislead stakeholders.

Approach / Task: I would reconcile these signals by decomposing LTV and churn by cohorts, segments, and revenue drivers, verify data accuracy, and present a clear set of metrics and visualizations that explain how both can be true.

Actions I would take and what I'd present:- Data quality checks: confirm definitions (churn window, LTV formula), verify regional mapping and time alignment (revenue lag vs churn measurement).- Cohort LTV vs cohort churn: show LTV by acquisition month cohort alongside their monthly retention curves — rising LTV overall can be driven by older cohorts while recent cohorts exhibit higher churn.- Revenue decomposition: break LTV into ARPU (avg revenue per user), purchase frequency, average order value, and customer lifetime. Show which component is growing (e.g., higher spend by power users) despite rising attrition among casual users.- Segment analysis: churn by persona, product SKU, channel, price tier, and device. Highlight if high-value customers are sticking while low-value volumes drop (or vice versa).- Time-lagged analysis: compare current churn (recent months) to LTV calculated over longer windows; annotate the dashboard to show that LTV is backward-looking.- Leading indicators: present product usage, engagement metrics (DAU/MAU, session length), support tickets, NPS/CSAT, promo/discount exposure — to explain cause of churn spike.- Sensitivity/what-if: show how a sustained churn increase by X% in the region would impact projected LTV and revenue.

Result / Recommendation: I’d deliver an executive dashboard with cohort retention charts, component decomposition, and recommended actions (targeted retention campaigns for affected segments, pricing/UX experiments, monitoring cadence). This clarifies that rising LTV can coexist with localized churn and provides actionable next steps to mitigate risk.

MediumTechnical

17 practiced

A sudden 12% drop in 3-month LTV is observed for a recent acquisition cohort. List a prioritized troubleshooting plan (5 steps) you would take as a BI analyst, including SQL checks, cohort comparisons, and stakeholder questions.

Sample Answer

1) Validate metric calculation (sanity & SQL checks) - Confirm LTV definition (revenue window, refunds, currency, user dedup). Re-run core query for cohort aggregated LTV and count. Example:

sql

SELECT cohort_date, SUM(revenue)/COUNT(DISTINCT user_id) AS ltv_3m
FROM events
WHERE event_date BETWEEN cohort_date AND cohort_date + INTERVAL '90 days'
GROUP BY cohort_date;

- Check joins, timezones, attribution logic, and if new ETL changed revenue tables.

2) Verify data quality and volume shifts - Check cohort size, new-user counts, event ingestion rates, and missing-day gaps. Compare raw rows ingested vs prior cohorts. Ask: "Any recent pipeline changes or backfills?"

3) Cohort comparisons and segmentation - Break down by acquisition channel, country, device, and promo. If drop concentrated, isolate channel-specific issues. Query sample:

sql

SELECT channel, AVG(ltv_3m) FROM cohort_metrics WHERE cohort_date = 'YYYY-MM' GROUP BY channel;

4) Product/marketing changes & external factors - Ask stakeholders: new pricing, offers, cancellation policy, marketing creatives, or macro events. Correlate dates of changes with cohort date. Check A/B tests overlapping cohort.

5) Monitor, escalate, and recommend actions - If data issue: roll back fix, reconcile numbers, alert ops. If real behavioral shift: recommend hypotheses for experiments (e.g., re-engagement emails), prioritize channels with worst drop, and set short-term KPIs. Provide a summary dashboard widget showing before/after and top contributing segments for stakeholders.

MediumTechnical

19 practiced

Discuss pros/cons of last-touch versus multi-touch attribution when computing channel-level CAC and LTV. Propose a practical approach for a mid-size company with limited instrumentation and explain how to implement it in analytics pipelines.

Sample Answer

Situation: You need channel-level CAC and LTV that stakeholders trust, but instrumentation is limited and resources are constrained.

Pros/Cons- Last-touch - Pros: Simple, deterministic, easy to compute from conversion event → channel; low engineering overhead; aligns with many marketing reports. - Cons: Over-credits the final touch, ignores upper-funnel influence; biases toward retargeting/paid search; distorts CAC and LTV decisions for channels that assist early.- Multi-touch (heuristic or algorithmic) - Pros: More accurate reflection of the customer journey; distributes credit to awareness and nurturing channels; better for optimizing full-funnel spend. - Cons: Requires better instrumentation (user-level IDs, cross-device tracking), more complex pipelines, assumptions in model (rules or attribution algorithm) can be contested.

Practical approach for a mid-size company1. Adopt a hybrid, pragmatic attribution: default to last-touch for reporting speed but provide a multi-touch “assisted” view using a simple, explainable rule (e.g., position-based: 40% first touch, 40% last touch, 20% evenly split across middle touches) and time-decay as an optional variant.2. Instrumentation priorities: ensure persistent user_id (login or cookie + hashing), consistent UTM capture on entry, and store touch timestamp + channel/source/medium.3. Reporting cadence: use last-touch in real-time dashboards; run nightly batch multi-touch attribution to produce adjusted CAC/LTV for weekly/monthly strategic reports.

Implementation in analytics pipelines- Data collection: capture touch events table (user_id, touch_id, timestamp, channel, campaign, utm_params, session_id).- Sessionize & stitch: backfill/clean user identifiers; deduplicate bots; join touches to conversion/payment events via user_id and time window (e.g., 90 days).- Attribution job (SQL/ETL): - Build touch sequences per user → apply rule-based weights to each touch. - Aggregate weighted cost: channel_CAC = sum(weight * cost_at_touch) / conversions; LTV by assigning conversion/revenue credit proportionally.Example pseudocode SQL steps: 1) window over user_id order by timestamp to enumerate touches; 2) assign weight per position or time-decay factor; 3) sum(weight * revenue) by channel; divide by conversions to get CAC.- Validation & monitoring: compare last-touch vs multi-touch weekly; monitor shifts > X% and investigate; sample check with manual funnel analysis.- Governance: document model, assumptions, and refresh cadence; store both metrics (last-touch and multi-touch) in BI layers so stakeholders can choose.

Key trade-offs and communication- Be transparent: show both metrics and explain biases. Use last-touch for tactical campaign reporting, multi-touch for strategic budget allocation. Iterate attribution model as instrumentation improves (move to probabilistic or algorithmic models when user linking and event coverage are robust).

Unlock Full Question Bank

Get access to hundreds of Unit Economics and Scaling interview questions and detailed answers.

Join thousands of developers preparing for their dream job.