Catalog
8 branches of analytics, product and AI interview questions
Explore eight skill tracks, called branches, filterable by subtopic, company, tool and difficulty. SQL questions run against sample tables, choice and numeric answers are checked instantly, and written cases are scored against evaluation criteria.
Filters (1)
Branches
Access
Free onlyQuestion type
Type to find a company, then pick it from the list below.
Tool
48 questions · page 1 of 2
What a 95% interval actually says
Distinguish the frequentist claim from the common misreading.
Quick check · Intervals · 6 min · Medium
Statistics & Inference · Original Uplift question
Ordering an executive readout
Choose the right first line for a results email.
Quick check · Writeup · 5 min · Easy
Analytics Case & Communication · Original Uplift question
Why peeking inflates false positives
Explain the statistical cost of checking results daily.
Quick check · Readout · 5 min · Easy
Experimentation & A/B Testing · Original Uplift question
Reading a flattening retention curve
Interpret a curve that flattens above zero.
Quick check · Retention · Amplitude · 6 min · Medium
Product Analytics & Metrics · Original Uplift question
Choosing the fact table grain
Pick the right grain for an order-line fact table.
Quick check · Modeling · dbt · 6 min · Easy
SQL & Data Modeling · Original Uplift question
Ranking four features with RICE
Order a backlog using reach, impact, confidence and effort.
Quick check · Prioritization · 7 min · Easy
Product Management · Original Uplift question
Why last-click flatters brand search
Identify the bias in last-click attribution.
Quick check · Attribution · 5 min · Easy
Digital & Marketing Analytics · Original Uplift question
Catching agent regressions before release
Prevent shipped updates from breaking tasks that used to work.
Quick check · Evaluation · 6 min · Easy
AI Agents & Automation · Original Uplift question
Offline vs online evaluation
You're shipping an AI support agent.
Quick check · Evaluation · 6 min · Medium
AI Agents & Automation · Company-context scenario
Which attribution model over-credits early touchpoints
A customer sees a Facebook ad (first touch), later clicks a Google search ad, then converts after an email.
Quick check · Attribution · 6 min · Medium
Digital & Marketing Analytics · Company-context scenario
What the Central Limit Theorem lets you do
Why does the Central Limit Theorem matter for A/B testing, where the underlying metric (e.g.
Quick check · Foundations · 6 min · Easy
Statistics & Inference · Company-context scenario
Big week-1 lift that fades to zero
A redesigned button shows +8% engagement in week 1, but by week 4 treatment and control are indistinguishable.
Quick check · Readout · 5 min · Easy
Experimentation & A/B Testing · Company-context scenario
Type I vs Type II error trade-off
A medical screening test is being tuned.
Quick check · Intervals · 5 min · Easy
Statistics & Inference · Company-context scenario
Measuring groundedness in a RAG system
ProYour retrieval-augmented (RAG) assistant sometimes states facts not supported by the retrieved documents.
Quick check · Product metrics · 6 min · Medium
AI Agents & Automation · Company-context scenario
Spotting the vanity metric
A growth team celebrates that 'total registered users crossed 10 million.' Why might a PM push back on using this as the headline success metric?
Quick check · Metrics · 5 min · Easy
Product Management · Company-context scenario
MMM vs multi-touch attribution
ProWith third-party cookies deprecating, a CMO asks why the team is investing in Marketing Mix Modeling (MMM) instead of relying on multi-touch...
Quick check · Attribution · 6 min · Medium
Digital & Marketing Analytics · Company-context scenario
What goes in the first line of a readout
You're writing a results email to executives about an experiment.
Quick check · Writeup · 5 min · Easy
Analytics Case & Communication · Company-context scenario
The returning-planes problem
In WWII, analysts studied bullet holes on planes that RETURNED from missions and proposed armoring the most-hit areas.
Quick check · Bias · 5 min · Easy
Statistics & Inference · Company-context scenario
Bias-variance trade-off
ProA model has near-perfect training accuracy but poor test accuracy.
Quick check · Modeling intuition · 6 min · Medium
Statistics & Inference · Company-context scenario
Wins in every segment, loses overall
ProYour treatment beats control in BOTH the mobile segment and the desktop segment, yet loses when you pool all users together.
Quick check · Readout · 6 min · Medium
Experimentation & A/B Testing · Company-context scenario
Which success metric reflects the user's experience
An agent solves 70% of tasks on the first attempt and 90% within three attempts.
Quick check · Product metrics · 6 min · Easy
AI Agents & Automation · Company-context scenario
A 50/50 split that came out 52/48
You designed a 50/50 experiment.
Quick check · Data quality · 6 min · Medium
Experimentation & A/B Testing · Company-context scenario
When the mean misleads
You report 'average revenue per user = $50' but leadership is confused because almost no one spends near $50.
Quick check · Descriptive stats · 5 min · Easy
Statistics & Inference · Company-context scenario
What CUPED buys you
ProYour experiments are underpowered and you can't get more users.
Quick check · Design · 7 min · Hard
Experimentation & A/B Testing · Company-context scenario