Interview guides · Independent Software

Independent Software Data Scientist Interview Questions & Guide 2026

Causal inference drills for Independent Software data scientist prep: diff-in-diff, matching, instruments, and the limits of observed data.

The 60-second brief

As a Data Scientist at Independent Software, you are a pivotal bridge between complex, high-stakes data and actionable national security intelligence. You will not be working in a silo; instead, you will collaborate directly with mission teams to tackle challenges that require both rigorous statistical modeling and a deep understanding of unique domain datasets, such as SIGINT, RF, and PROFORMA data. Your work directly influences operational efficiency and mission outcomes by transforming incomplete or unstructured data into clear, evidence-based insights.

Coding rounds for this role are usually data-manipulation shaped rather than data-structure shaped: group-bys, joins, time windows, ranking within a partition. Confirm the format before spending a week on graph traversal.

Focus areas

Power experiments for heavy-tailed account revenue
Analyse at the account grain, cluster errors
Separate contracted seats from actively used seats

Overview

This role is defined by the intersection of advanced technical mastery and mission-oriented problem solving. Whether you are automating analytical workflows in Jupyter Notebooks, optimizing data pipelines, or developing predictive models to detect anomalies, your contributions are immediate and tangible. Independent Software values inquisitive professionals who can translate abstract mission requirements into precise technical specifications, ensuring that the results you generate are both robust and understandable to non-technical stakeholders.

The Independent Software interview process

Independent Software candidates report 3 rounds · ≈ 3-5 weeks. The stages below are what candidates describe, not a published process.

Initial Screening

A screening call is a matching exercise run by someone who will not evaluate your statistics. They are checking that the work described on your resume is work you personally did, and that its scope matches the level the role is written for. Logistics get settled in the same half hour so nobody spends an interviewer's afternoon on a mismatch. The answer that fails is the one narrated in the plural. If every sentence is 'we built' and 'the team decided', there is nothing specific to write down about you. Name the piece that was yours, the decision you made inside it, and what changed after.

What to demonstrate

Whether the ownership implied by your resume survives one round of follow-up about who actually did which part
Whether your described scope (data size, stakeholders, what shipped) matches the seniority the role is written at
Whether timeline, location and compensation expectations make the rest of the loop worth scheduling

How to prepare

Rewrite your top three resume bullets in the first person singular, each with the decision you made and what moved afterwards, then say them out loud once so the 'we' does not return under pressure
Attach one number to each project: the baseline, the change, and the window it was measured over. Where impact was never measured, say that plainly rather than inventing a figure
Settle your compensation range before the call and give it as a range with a reason behind it, such as current total comp or a competing timeline, instead of deflecting the question twice

Technical Rounds

This round decides whether someone can hand you a schema and a question and trust the number that comes back. Correctness under a clock is the bar, not clever syntax. The habit that separates strong from weak answers is checking the grain: after every join, know how many rows you expect and whether the count moved. Most wrong answers in this format are not wrong logic, they are a fan-out from a key that turned out not to be unique, or a filter applied before an aggregate when it belonged after. Say what you expect before you run it.

What to demonstrate

Whether your row counts survive each join, and whether you notice on your own when they do not
Deliberate handling of rows that fail to match, including whether the question needs an inner join or a left join with the non-matches kept and counted
Whether NULLs are treated on purpose, given that a NULL compares equal to nothing and that COUNT of a column skips it
Reaching a defensible answer inside the window instead of a refined one after it

How to prepare

Take a two-table schema, write a join that fans out on purpose, then fix it by collapsing the many-side to one row per key before joining. Repeat until the fix is reflex rather than recall.
Write a funnel as one query and print the distinct user count at each stage, then confirm each stage is a subset of the one above it rather than assuming it
Do a few timed runs in a plain text box with no autocomplete and no formatter, since assessment editors often have neither

Behavioral Assessments

Most of the weight in this round sits on the disagreement questions. Data work routinely produces an answer someone senior did not want, and the interviewer is trying to learn what you do in that hour. Both failure modes are common: folding as soon as a director pushes back, and treating the pushback as ignorance to be corrected with a better chart. A strong answer usually contains a specific thing the other person knew that you did not, and describes how you found out whether it changed the conclusion.

What to demonstrate

Whether you can state the other side's argument accurately before you explain why you disagreed
What you treated as evidence during the disagreement, such as a rerun under their assumption or a holdout check, rather than persuasion technique
Whether you distinguish being overruled from being wrong, and can give an example of each

How to prepare

Write out one disagreement where you turned out to be wrong, and say what in the data misled you. Candidates prepare the story where they were right, and the follow-up asks for the other one.
For your main disagreement story, be ready to say what result would have made you drop your position. If no such result exists, you were not arguing from the data.
Practise stating the opposing position out loud in one sentence the stakeholder would accept, then continue the story.

Where candidates lose points

Randomising an experiment at the user level when users share an account

Two problems fire at once. Colleagues in one workspace see each other's work and talk to each other, so a treated user changes the behaviour of a control user in the same account, which violates the no-interference assumption and biases the estimate toward zero. Separately, outcomes within an account are strongly correlated, so the effective sample size is roughly n / (1 + (m - 1) * rho) for m users per account and intra-class correlation rho, not n. With rho around 0.3 and twenty users per account that is a design effect near 6.7, meaning a user-level confidence interval is about two and a half times narrower than it should be and results cross significance thresholds on noise alone. Randomise the account and cluster the standard errors.

Reporting a mean over accounts when account revenue is heavy-tailed

When a small number of accounts hold most of the revenue, the sample mean is dominated by whichever of them happens to be in the sample, and the sample variance keeps growing as more data arrives instead of stabilising. In that regime the usual central-limit-based confidence interval understates uncertainty, and a single renewal or a single large account's batch job can flip the sign of a measured effect. The fixes are to pre-register a winsorisation or capping rule before looking at the outcome, to report account counts crossing a threshold alongside the revenue figure, or to define the estimand on a bounded transform. Choosing the cap after seeing the result is a separate and worse problem, because the cap then encodes the answer.

Stopping an experiment the moment it crosses significance

Fix the sample size or duration before launch, or use a method built for continuous monitoring such as a sequential test, always-valid confidence intervals, or group-sequential boundaries. Repeatedly checking a fixed-horizon p-value against 0.05 pushes the real false-positive rate well above 5 percent.

Optimising accuracy on a heavily imbalanced target

State the base rate first, then choose the metric from the relative cost of a false positive against a false negative: precision and recall at the operating threshold, PR-AUC, or expected cost. At a 1 percent positive rate, predicting the majority class for everyone scores 99 percent accuracy and is worthless.

Questions & practice

How do you validate the performance of a model when the ground truth is difficult to obtain?

Medium · machine learning and modelling

Approach

Check what information would not exist at prediction time, and exclude it.
Frame the prediction: the label, the moment of prediction, and the action it triggers.
Pick an evaluation metric that matches the cost of each error type, not a default.

Follow-up

How would you choose the decision threshold, and who owns that choice?
Where could label leakage enter this setup?

Simulate how the renewal calendar distorts monthly churn rates

Medium · simulation · churn · numpy · cohorts

Simulate 1,200 accounts on annual contracts. Draw each account's renewal month from a deliberately lumpy calendar: 30% renew in January and the remaining 70% are spread evenly over the other eleven months. At each renewal an account churns with probability 0.18, independent of month; survivors renew and come back twelve months later. Run 24 simulated months. For each month compute two rates: churned accounts over all live accounts, and churned accounts over accounts whose term ended that month. Report the mean and the month-to-month standard deviation of each series, and state which one belongs in an executive summary.

Approach

Build the panel with numpy state arrays rather than a per-account loop: a next_renewal_month vector, an alive boolean vector, and a loop over the 24 months only. Looping over 24 months is fine; looping over 1,200 accounts inside it is what makes the simulation too slow to iterate on.
Maintain the live set honestly. An account that churned in month m must leave the denominator from m+1 onward and can never be renewal-eligible again; if it stays, the naive rate drifts downward for reasons that have nothing to do with churn and the calendar effect gets buried.
Compute both series over the same months and compare dispersion, not only level. The eligible-base rate should sit near 0.18 with binomial noise scaled by that month's renewal count; the naive rate spikes in January and collapses in thin months.
Quantify the gap instead of describing it: the ratio of the two means is roughly the reciprocal of the average monthly renewal-eligible fraction, and the naive series' standard deviation is driven by the signing calendar rather than by customer behaviour.
Check against the closed form before trusting the output. With the live set maintained correctly the eligible-base rate is an unbiased estimator of 0.18 in every month, so a systematic offset means the bookkeeping is wrong, not that the simulation found something.

Worked solution (30 min)

rng = np.random.default_rng(0); p = [0.30] + [0.70/11]*11; next_renewal = rng.choice(12, size=1200, p=p); alive = np.ones(1200, bool)
For m in range(24): eligible = alive & (next_renewal == m); churn = eligible & (rng.random(1200) < 0.18); record churn.sum(), eligible.sum(), alive.sum() at month start; alive &= ~churn; next_renewal[eligible & ~churn] += 12
naive = churned / live_at_start; eligible_rate = churned / eligible, left as NaN where eligible == 0.
Report naive.mean(), naive.std(ddof=1), np.nanmean(eligible_rate), np.nanstd(eligible_rate, ddof=1) and the ratio of the two means.

Expected result: The eligible-base rate averages close to 0.18 (roughly 0.17 to 0.19 at this sample size) with no January spike. The naive rate averages near 0.015, about one twelfth of it, and its January values run several times a typical month, so its month-to-month standard deviation is a large fraction of its own mean.

Verify it yourself

The 24 monthly churn counts sum to 1200 minus the final alive count.
Switching the calendar to uniform (p = 1/12 everywhere) collapses the naive series' dispersion while leaving the eligible-base mean unchanged.
Setting the churn probability to 0 makes both series identically zero and leaves 1,200 accounts alive.
Quadrupling the account count roughly halves the standard deviation of the eligible-base rate without moving its mean.

Follow-up

Compounded over twelve months the naive rate lands close to the true annual churn. Does that rescue it?
How would you report churn in a month where only nine accounts were renewal-eligible?
Eighteen-month terms are now being sold alongside annual ones. What breaks?

Audit daily usage rows for grain and arithmetic violations

Easy · data-quality · pandas · grain

You are handed fct_usage_daily as a pandas DataFrame with account_id, workspace_id, sku_code, usage_date, billable_quantity, included_quantity_applied, overage_quantity, list_amount_cents, discount_amount_cents, net_amount_cents, cogs_cents, is_restated, first_written_at and restated_at. The declared grain is one row per (account_id, workspace_id, sku_code, usage_date). Write audit(df) returning a DataFrame with one row per failing check: check name, failing row count, and one example key. Cover at minimum grain duplication, negative quantities or amounts, the identity net = list - discount, billable = included + overage, and rows where is_restated is true but restated_at is null.

Approach

Check the grain before anything else with df.duplicated(subset=key, keep=False), and count rows rather than groups so a key appearing twice contributes 2 — if the grain is broken every arithmetic count below it is uninterpretable.
Express each invariant as a boolean Series over the whole frame. The cent columns are integers and compare exactly, so use !=; the numeric(18,6) quantity columns need np.isclose with atol=1e-6 because included + overage is a decimal sum.
Handle null as its own failure mode. Comparisons against NaN return False, so a check written as rows_that_pass = (a == b - c) silently files every null-amount row wherever the negation happens to land; build each check as violations = ~condition | column.isna().
Collect the checks as a list of (name, mask) pairs and assemble the output in one pass, so adding a check is one line and every check reports in the same shape.
Order the output with structural failures (grain, null keys) above arithmetic failures, and report zero-count checks too — a check that silently disappears when it passes is indistinguishable from a check that was never run.

Follow-up

Which of these should block a dashboard refresh and which should only warn?
Rows with is_restated = true legitimately change value after first write. How do you make yesterday's audit result reproducible?
How would you extend this to catch a partition that is missing entirely rather than wrong?

How would you aggregate data across multiple sources to support a real-time monitoring dashboard?

Medium · sql and data manipulation

Approach

State the window function and its partition and ordering out loud before writing it.
Compute rates by summing numerator and denominator separately, never by averaging rates.
Handle the rows that do not match: a LEFT JOIN with a NULL check is usually the question.

Follow-up

What breaks if events arrive late or out of order?
How would you verify this result without re-running the same query?

How would you use SQL window functions to identify trends in a time-series dataset?

Medium · sql and data manipulation

Approach

Handle the rows that do not match: a LEFT JOIN with a NULL check is usually the question.
Compute rates by summing numerator and denominator separately, never by averaging rates.
Say which table is the grain you start from, and join outward from it.

Follow-up

How would you verify this result without re-running the same query?
How does the query change if the join becomes one-to-many?

Seven-day activation rate by signup cohort week

Easy · funnel · cohorts · censoring

dim_account carries account_id, created_at, is_internal and is_current; fct_api_request carries account_id, request_at, http_status, api_key_id and traffic_class. Build a signup cohort by the ISO week of created_at over accounts with is_internal = false. An account counts as activated when it issues a request with http_status < 400, a non-null api_key_id and traffic_class <> 'synthetic_monitor' within 168 hours of its own created_at. Return cohort accounts, activated accounts and the rate per week, and exclude any week that has not yet fully elapsed its 168-hour window.

Approach

Collapse dim_account to one row per account_id before anything else. It is a type 2 dimension, so a plan or status change gives the same account several rows; filtering to is_current = true is the cheapest correct choice here because created_at does not change across versions.
Express the window as interval arithmetic on the timestamptz column: request_at >= created_at AND request_at < created_at + interval '168 hours'. A date-difference of 7 days is a different and wrong condition for accounts created mid-day.
Test activation with EXISTS rather than a join to MIN(request_at). EXISTS short-circuits, keeps the cohort at one row per account, and cannot fan out.
Aggregate by date_trunc('week', created_at AT TIME ZONE 'UTC'), counting accounts and activated accounts, and divide as a ratio of counts.
Drop unreportable weeks: the last account in a cohort week is created just under week_start + 7 days, so the week is only complete once now() >= week_start + interval '14 days'. Without that filter the newest week always looks like a regression.

Worked solution (20 min)

Write the cohort CTE: SELECT account_id, created_at FROM dim_account WHERE is_current AND NOT is_internal AND created_at >= <window start>.
Add the activation predicate as a correlated EXISTS over fct_api_request on account_id with the four conditions: http_status < 400, api_key_id IS NOT NULL, traffic_class <> 'synthetic_monitor', and the 168-hour bracket.
Group by date_trunc('week', created_at AT TIME ZONE 'UTC'); select count() AS cohort_accounts, count() FILTER (WHERE activated) AS activated_accounts, and the ratio cast to numeric.
Add HAVING or an outer WHERE that keeps only weeks where week_start + interval '14 days' <= now().
Spot-check one account that activated on hour 167 and one that activated on hour 169 to confirm the boundary is exclusive at the top.

Expected result: One row per ISO cohort week with cohort_accounts, activated_accounts and activation_rate between 0 and 1; the newest row is at least 14 days old, and accounts with no rows at all in fct_api_request appear in the denominator with activated = 0 rather than disappearing.

Verify it yourself

Summing cohort_accounts across weeks equals the count of non-internal current-version accounts created in the same window.
Re-running the query a day later leaves every previously reported week byte-identical; any movement means the censoring filter is wrong.
activated_accounts is never greater than cohort_accounts, and the count of distinct account_id in the activated set matches a separate COUNT DISTINCT over the fact table.

Follow-up

The median time-to-first-successful-call is more informative than a fixed-window rate. Why can you not compute it from this query, and what estimator does it need?
How would you separate accounts that never called from accounts that called and got only 4xx responses, and which of those is a product problem?

How would you design a set of metrics to measure the success of a new analytical tool?

Medium · product sense and metrics

Approach

Fix the population and the time window before naming any metric.
Restate the decision this analysis has to support, and who acts on the answer.
Name one primary metric, then the guardrail that stops it being gamed.

Follow-up

What would you do if the primary metric and the guardrail moved in opposite directions?
How would you detect that the metric is being gamed rather than genuinely improving?

How do you balance the need for high-precision models with the need for low-latency, real-time insights?

Medium · product sense and metrics

Approach

Fix the population and the time window before naming any metric.
Decompose the metric into the rates that drive it, and say which one you would check first.
Name one primary metric, then the guardrail that stops it being gamed.

Follow-up

How would you detect that the metric is being gamed rather than genuinely improving?
Which segment would you cut first, and what would that rule out?

Describe a situation where you had to pivot your approach due to shifting mission priorities.

Medium · product sense and metrics

Approach

State what result would change your recommendation, so the answer is falsifiable.
Fix the population and the time window before naming any metric.
Restate the decision this analysis has to support, and who acts on the answer.

Follow-up

Which segment would you cut first, and what would that rule out?
How would you detect that the metric is being gamed rather than genuinely improving?

If a critical mission metric suddenly drops, what is your systematic process for diagnosing the root cause?

Medium · product sense and metrics

Approach

Fix the population and the time window before naming any metric.
Decompose the metric into the rates that drive it, and say which one you would check first.
State what result would change your recommendation, so the answer is falsifiable.

Follow-up

What would you do if the primary metric and the guardrail moved in opposite directions?
How would you detect that the metric is being gamed rather than genuinely improving?

What are the most common experimentation pitfalls you encounter when running A/B tests?

Medium · a/b testing and experimentation

Approach

Name the guardrails that would stop a launch even on a positive primary result.
State the primary metric and the minimum effect worth shipping, then size the test.
Decide the analysis before seeing data, including how long it runs and when you look.

Follow-up

What would you do if you could not randomise at all?
How would you handle interference between treated and control units?

How do you determine the required sample size to reach statistical significance for a low-volume feature?

Medium · a/b testing and experimentation

Approach

Name the guardrails that would stop a launch even on a positive primary result.
Name the randomisation unit first; it decides the variance and what the test can detect.
Decide the analysis before seeing data, including how long it runs and when you look.

Follow-up

What would you conclude if the result is positive but the test is underpowered?
How would you handle interference between treated and control units?

Triage a sample ratio mismatch before reading the result

Easy · srm · assignment · instrumentation

An account-randomised test assigned arms 50/50 by hashing account_id. Exposure logging recorded 4,200 accounts: 1,987 in treatment and 2,213 in control. A chi-square goodness-of-fit test against the intended 2,100/2,100 split gives chi-square(1) = 12.2, p about 0.0005. The analyst reports the primary metric up 6% in treatment and asks to ship. Explain what the imbalance implies about that 6%, rank the causes you would investigate first given this domain's data, and name for each the single query against fct_api_request or dim_account that confirms or eliminates it.

Approach

State the rule before diagnosing anything: a sample ratio mismatch below a 0.001 alarm threshold invalidates the readout. Whatever mechanism removed 113 accounts from one arm almost certainly removed them non-randomly, which makes it a selection effect on the outcome. The 6% is not to be adjusted, caveated or shipped; it is unusable until the mechanism is named.
Verify the test is testing the right thing. chi-square = (1987 - 2100)^2 / 2100 + (2213 - 2100)^2 / 2100 = 6.08 + 6.08 = 12.16 on one degree of freedom. Confirm the denominator is the intended assignment count rather than the observed total, and that you are testing assignment rather than an analysis population that has already been filtered.
Rank causes by how this domain actually breaks rather than by textbook order: is_internal accounts filtered after assignment instead of before; exposure logged on a code path that returns 5xx more often in one arm, silently dropping those accounts; assignment recorded at first request but exposure at a later event, so accounts that churned in between appear in only one arm; dim_account type-2 versioning giving one account_id two surrogate keys and two hash inputs; deployment_model = 'self_hosted' accounts that never emit exposure telemetry at all.
Attach one discriminating query to each: count arms before the is_internal filter; compare the http_status >= 500 rate by arm on the exposure endpoint in fct_api_request; compare the distribution of created_at and churned_at by arm in dim_account; count distinct account_sk per account_id inside the experiment window; cross-tabulate arm against deployment_model.
Compare counts at three fixed stages, assignment, first exposure, and analysis population, and localise the divergence to one of them. The stage where the arms first separate names the subsystem, and everything downstream of it is a symptom.
Report the outcome as abort, fix, rerun, and be explicit about what survives: the variance estimate for re-powering, the instrumentation fix, and nothing whatsoever about the effect size.

Worked solution (20 min)

Recompute the statistic: two terms of 113^2 / 2100 = 6.08, total 12.16, p about 0.0005 on one degree of freedom, below the 0.001 alarm line.
Pull arm counts at assignment, at first exposure, and in the analysis population, and find the first stage at which they diverge.
At that stage, cross-tabulate arm against is_internal, deployment_model, and the 5xx rate on the exposure endpoint.
Write the abort note naming the mechanism, the corrected code path, and the rerun date, with the effect estimate explicitly withheld.

Expected result: chi-square(1) = 12.16, p about 0.0005, so a 0.001 alarm threshold fires and the readout is abandoned. The divergence is localised to a single pipeline stage by comparing arm counts at assignment, at exposure and at analysis, and the most likely causes in this domain are post-assignment internal-account filtering or exposure logging that fails asymmetrically on 5xx responses.

Verify it yourself

A 113-account shortfall on 2,100 expected is a 5.4% deficit in one arm, which under a sound hash is roughly a 1-in-2,000 event. That is what the 0.001 threshold is calibrated to catch.
Counts at the assignment stage should already be balanced if the bucketing is correct. If they are not, the bug is in the hash or the salt, not in the logging.
Any fix that restores balance by filtering on a variable measured after exposure has conditioned on a post-treatment variable and reintroduces the bias it appears to remove.

Follow-up

The imbalance disappears once you restrict to accounts with at least one successful request. Does that fix the experiment or confirm the bug?
What alarm threshold would you set for this check, and why is 0.05 the wrong one for a diagnostic you run on every experiment every day?
How would you detect a mismatch confined to one segment while the overall split looks clean?

Monthly logo churn triples with no change in satisfaction

Hard · renewal-eligible base · censoring · timestamp semantics

Monthly logo churn, computed as churned accounts divided by all paying accounts, tripled last month. Contracts are annual. From fct_subscription_period (account_id, term_start_date, term_end_date, amendment_type, auto_renew, booked_at, superseded_by_id, is_current) and dim_account (account_id, churned_at, account_status, employee_band, acquisition_channel), rebuild churn on the renewal-eligible base, separate calendar effects from customer behaviour, and state whether retention actually changed. Note that churned_at is sometimes set when the record was updated rather than at term end.

Approach

Rebuild the denominator as accounts whose term_end_date falls in the month. An account eleven months from renewal sits in the current denominator while being structurally incapable of entering the numerator, so on annual contracts the published rate understates the truth by roughly the reciprocal of the annual renewal fraction and moves with the signing calendar.
Plot the renewal-eligible base by month across two years. A signing surge twelve months earlier reproduces itself as an eligibility surge now, and a rate whose denominator ignores that tracks the sales calendar rather than customer sentiment.
Date each churn by term_end_date, never by churned_at or an update timestamp. Inspect the distribution of churned_at minus term_end_date: a backlog cleared in one batch appears as a mass at a single date and shifts losses into whichever month the operations team did its paperwork.
Apply the 45-day grace for late renewal paperwork so the most recent month and a half is marked not reportable, rather than printing a number that will rise once the paperwork lands.
Compare corrected gross logo retention against its own trailing distribution, and if a real change survives, cut it by employee_band, acquisition_channel and plan_tier before proposing any cause.

Follow-up

Some contracts in the window have not reached their renewal date. When does this require a survival estimator rather than a simple rate, and which one would you use?
How do you report churn to an audience that wants a monthly number when the underlying event is annual and lumpy?

Suggested practice checklist

For a candidate whose interviews will centre on A/B testing, metric movement and causal claims. Design comes before arithmetic, arithmetic before analysis, and the week ends by rehearsing the readout rather than the derivation.

Day 1 — Design one test end to end on paper

Take a single feature change and write the full design: randomization unit, the exact point of exposure, the primary metric with its grain, guardrails, allocation, planned duration, and the decision rule committed before any data exists.
Write why the randomization unit must sit at or above the level where treatment can spill over, and give one case where user-level randomization is still contaminated (shared accounts or devices, or two participants in the same marketplace).
State in advance what you will do if the primary metric is flat while a secondary metric is significant.

Deliverable: A one-page test design with a decision rule written before launch.

Day 2 — Power arithmetic until it is automatic

Compute required sample size per arm for a binary metric with the normal approximation, n is approximately 2 times (z for alpha/2 plus z for power) squared times p(1 minus p) divided by delta squared, for baselines of 2, 10 and 40 percent at a 5 percent relative lift, and note that for a fixed relative lift the requirement falls as the baseline rises because delta grows proportionally with p.
Redo the calculation for a continuous metric using variance in place of p(1 minus p), and show why a heavy-tailed quantity such as revenue per user needs either far more traffic or a capped version with a stated cap.
Convert one of the results into weeks given a weekly eligible traffic figure, then list the two honest ways to shorten it (accept a larger detectable effect, or reduce variance) and write why quietly lowering the power target is a decision to miss more real wins, not a speedup.

Deliverable: A small script or sheet that maps baseline, minimum detectable effect, alpha and power to sample size and weeks, cross-checked against a published calculator.

Day 3 — Variance and the unit-of-analysis problem

Take a ratio metric whose denominator is not the randomization unit (clicks per session, randomized by user) and compute the standard error twice, once naively at session level and once by the delta method or a user-level bootstrap, then record how much the naive version understates it.
Implement CUPED on simulated data: choose a pre-period covariate X measured before assignment, estimate theta as Cov(Y, X) divided by Var(X), and analyse Y minus theta times (X minus its mean) in place of Y. Confirm the variance of the adjusted outcome equals the raw variance multiplied by one minus the squared correlation between Y and X, so a correlation of 0.45 removes about 20 percent of the variance and not 80.
Now run that simulation a few hundred times and confirm the adjusted effect estimate is unbiased for the same effect rather than numerically identical to the raw one. Within any single run the two differ, sometimes by a large fraction of the true effect, because the two arms' pre-period covariate means never coincide exactly in a finite sample; they agree in expectation, which is the property that matters and the one to state out loud.

Deliverable: A notebook showing the adjusted estimator with a measurably smaller variance than the raw one, plus a repeated-simulation table showing the two estimators agreeing on average while differing run by run.

Day 4 — Validity threats you can actually test for

Run a sample ratio mismatch check as a chi-square goodness-of-fit test against the intended allocation, and write the three causes you would chase first (assignment logged before exposure, an arm-specific redirect or load failure, bot filtering applied asymmetrically).
Simulate peeking: generate A/A data, test daily at alpha 0.05 across 14 looks, record the inflated false positive rate, then apply an alpha-spending boundary or commit to a fixed horizon and confirm the rate returns to nominal.
Write how you would separate a novelty effect from a durable lift using the treatment effect plotted against days since first exposure, and what shape would change your recommendation.

Deliverable: One table showing the peeking false positive rate before and after correction, plus a written SRM triage list.

Day 5 — When randomization is not available

Write the identifying assumption for difference-in-differences (parallel trends in the absence of treatment), then plot pre-period trends for two candidate control groups and justify rejecting one of them.
Design a switchback test for a change where user-level randomization would leak across participants, choosing a time-block length against the carryover you expect and saying how you would detect carryover in the data.
List what an interrupted time series or a synthetic control buys you and the one thing neither can rule out: an unobserved shock that coincides with the launch.

Deliverable: A one-page memo recommending a single quasi-experimental design and naming its weakest assumption explicitly.

Day 6 — The readout query

Write the assignment-to-exposure join that returns exactly one row per unit per experiment, and handle units appearing in both arms by excluding and counting them rather than silently keeping one.
Compute the per-arm metric, its variance and the relative lift with a confidence interval in SQL, then reproduce the identical numbers in a notebook as a cross-check.
Add a segment breakdown and write the sentence that keeps it from being p-hacking: segments declared in advance, everything else reported as exploratory and corrected for multiplicity.

Deliverable: A single query that outputs the full readout table, matched to a notebook recomputation.

Day 7 — Present it to someone who will not read the appendix

Give a 10-minute readout of a real or simulated experiment in the order decision, number, uncertainty, caveat.
Have your listener ask "can we ship it" in the case where the primary is flat and a guardrail moved, and answer with a recommendation rather than a request for more data.
Rewrite your opening line so the recommendation lands before any methodology.

Deliverable: A one-page readout whose first line is the recommendation.

Bring numbers, including the ones you are unsure of

An answer without a quantity is hard to interrogate, so interviewers keep probing until they find one. Come with the baseline, the change, the window it was measured over, and how confident you were. If the effect never got measured, say so and say what you would have measured. Fabricated precision is worse than an honest gap.

How would you explain the concept of p-values to a non-technical stakeholder?

Medium · behavioural and stakeholder questions

Approach

Name the disagreement or constraint, and how you resolved it with evidence.
State the situation in two sentences and spend the rest on your reasoning.
Pick a story where you drove the decision, not one where you observed it.

Follow-up

How did you know the outcome was caused by your change?
What did you decide not to do, and why?

Announce a metric fix that cuts the headline number

Medium · metric definitions · stakeholders · communication

Weekly active organisations, the count on the company dashboard, has never excluded rows where dim_account.is_internal is true, and it counts traffic with traffic_class in synthetic_monitor and load_test. Correcting both reduces that count by 11 percent and removes most of the growth reported over two quarters. The figure appears in a board deck and in two teams' quarterly goals, one written on the count and one on the weekly active organisation ratio, whose denominator is accounts whose account_status was in ('trial','free','active_paid') through the week. Decide the order in which you tell people, what the dashboard shows during the transition, and what you propose happens to goals already set against the old definition.

Approach

The interviewer is probing whether you can land a correction as an operational change with a plan attached, rather than as an announcement other people then have to clean up after.
Quantify each exclusion separately before telling anyone: internal accounts, synthetic monitors, load tests. Three known quantities are a discussion; one alarming total is an argument.
Be precise about which side of the metric each exclusion touches, because one team's goal is on a count and the other's is on a ratio. The traffic-class filters remove requests, so they shrink the numerator only. Dropping internal accounts removes them from the ratio's denominator as well, since internal accounts carry ordinary account_status values and therefore sit in that denominator. Internal accounts are active in almost every week while the real base is not, so the numerator loses a larger share than the denominator and the ratio falls by less than the count does. Compute both and say which one the 11 percent is before anybody assumes.
Check whether the trend changes, not only the level. A constant 11 percent shift is a rebasing and nothing more. A shift that widens over time means the reported growth was partly internal or synthetic, which makes the existing goals unachievable as written and changes what you are asking teams to do.
Sequence the disclosure: the metric owner and the two teams whose goals move first and privately, then the board channel with a written bridge, then the dashboard. The dashboard is last because a number that changes without explanation is read as instability rather than as a fix.
Run both series for one reporting period with the bridge visible, restate history rather than letting the series break at a date, and set the date the old series is removed.
Propose the goal treatment yourself: rebase each target by the shift measured on the metric that target is written against, rather than leaving each team to negotiate individually, which is where corrections of this kind usually die.

Follow-up

One team's quarterly goal is now unreachable. Rebase the target or let it miss, and what does each choice teach the organisation?
How would this have been caught when the metric was first defined?
What else on that dashboard shares this failure mode, and how would you find out this week?

Scope an open-ended request to predict account churn

Medium · scoping · churn modelling · operating point

A customer success director asks for a list of accounts about to churn. You know only that the team has six people and that contracts are annual. Available data is fct_subscription_period, fct_usage_daily, fct_api_request, fct_support_ticket and dim_account. Before writing any code, produce the questions you need answered, a proposed definition of about to churn, and the shape of the artefact you would hand back, including the operating point that turns a score into a decision.

Approach

The interviewer is probing whether you convert a vague request into a decision with a capacity constraint attached. A candidate who starts talking about model families has already failed the exercise.
Pin the event and the horizon first. Churn is only possible at term_end_date, so the population is accounts renewing in the next 60 to 90 days, not the whole base. Ask explicitly whether contraction and downgrade count as churn or only full non-renewal, because the three have different base rates and different interventions.
Pin the action and the capacity. Six people times a realistic number of meaningful interventions per week gives k, and k is what the list is ranked to. Evaluate on precision at k rather than a global AUC over accounts that will never be contacted.
Audit leakage before choosing features. Every feature needs a timestamp proving it existed before the prediction date. A downgrade amendment, a churn reason code, and a ticket opened after the renewal conversation started are all leaks that will make the offline number look excellent and the live list useless.
Ask for the counterfactual now rather than later. Coverage is assigned deliberately, so without a held-out slice agreed at the start the intervention can never be evaluated, and you will be asked for its impact in nine months regardless.
Propose the smallest artefact that closes the loop: a weekly ranked list sized to capacity with two or three inspectable reasons per row, plus a stated policy for accounts below the line.

Follow-up

The director insists all accounts are in scope, not only those renewing soon. How do you answer without simply refusing?
Historical non-renewals number about 30 a year. At what point do you tell them a model is the wrong tool and a rules list is better?
Which candidate features would you drop purely because you cannot date them?