SQL and Data Manipulation
SQL is the backbone of the technical screen. You will write queries with correct joins, filters, aggregations, and often window functions to calculate metrics. Strong performance means accurate results, efficient logic, and narration of trade-offs (e.g., deduping, NULL handling, performance).
Be ready to go over:
- Joins and filtering – Multi-table joins, WHERE vs. HAVING, handling missing keys.
- Aggregations and windows – Rolling metrics, ranking, deduplication, sessionization.
- Data cleaning – String ops, date truncation, type casts, outlier handling.
- Advanced concepts (less common) – CTE chains for complex logic, performance hints, pivoting/unpivoting.
Example questions or scenarios:
- “Write a query to compute weekly active users per product and identify the top 3 products per account each week.”
- “From events (view, click, purchase), compute conversion rates with a 24-hour window per user.”
- “Find the first time a user performed action X after onboarding; handle users with multiple onboarding events.”
- “Identify A/B experiment users with multiple exposures and exclude them from analysis.”
Product Analytics and Experimentation
Experimentation is central to decision-making at Atlassian. Interviews assess whether you can define success metrics, design robust tests, and interpret results with guardrails and diagnostics. Strong candidates show fluency with power, MDE, SRM checks, novelty/winner’s curse, and incremental value.
Be ready to go over:
- Metric design – Primary/secondary metrics, guardrails (e.g., latency, support load), anti-goals.
- Test design – Randomization, bucketing, power and MDE, ramp strategies.
- Diagnostics – Sample Ratio Mismatch (SRM), CUPED/bucketing, variance reduction, segment stability.
- Advanced concepts (less common) – Heterogeneous treatment effects, sequential testing, CUPAC, AA tests at scale.
Example questions or scenarios:
- “Design an A/B test to improve Jira onboarding completion; define success and guardrails.”
- “You observe significance on day 7 but SRM is flagged—what’s your next step?”
- “A test improves clicks but increases support tickets. How do you advise the PM?”
- “Power analysis for a 1% MDE with historical variance X—how many users and for how long?”
Statistics and Analytical Reasoning
Expect light-to-moderate statistics aligned with product decisions. You will be asked to reason about distributions, confidence intervals, p-values, bias, and trade-offs between parametric and non-parametric methods. Strong answers prioritize correct interpretation and practical implications.
Be ready to go over:
- Inference basics – Confidence intervals, hypothesis testing, Type I/II errors.
- Bias and variance – Confounding, selection bias, regression-to-the-mean.
- Effect size and uncertainty – Practical significance vs. statistical significance.
- Advanced concepts (less common) – Bootstrapping, variance reduction (CUPED), Bayesian A/B.
Example questions or scenarios:
- “Explain p-value vs. confidence interval to a PM deciding whether to ship.”
- “Your metric distribution is skewed—what do you do?”
- “Two cohorts differ pre-test; how do you de-bias?”
- “When would you choose a non-parametric test for median time-to-value?”
Case Studies and Business Metrics
Most candidates face a product/business case alongside SQL. You will structure ambiguous prompts, translate them into metrics and hypotheses, and outline the analysis. Strong performance combines product sense with measurable recommendations and awareness of risks and data quality.
Be ready to go over:
- Metric frameworks – Acquisition, activation, engagement, retention, monetization.
- Instrumentation – Event taxonomy, logging quality, identity stitching.
- Decision trade-offs – Short-term activation vs. long-term retention, revenue vs. user experience.
- Advanced concepts (less common) – Causal inference for non-randomized changes, difference-in-differences.
Example questions or scenarios:
- “Jira DAU is flat, but trial starts are up. Diagnose and propose next steps.”
- “Define leading indicators of team activation in Confluence and how you’d validate them.”
- “A new recommendation panel increases clicks but not retention—what now?”
- “Design an analysis to quantify the impact of a pricing page change on conversions and churn.”
Behavioral and Atlassian Values
You will be evaluated on how you collaborate, communicate, and make decisions aligned with Atlassian values. Expect broad prompts; your job is to bring specific, outcome-focused stories. Strong answers show ownership, learning, and customer-centric decisions.
Be ready to go over:
- Working with PM/Eng – Disagree-and-commit, influencing metric definitions, prioritization.
- Open communication – Transparent trade-offs, clear documentation, async updates.
- Customer focus – Choosing the safer path when impact on users is uncertain.
- Advanced concepts (less common) – Navigating long timelines and skill-pool hiring; resetting expectations.
Example questions or scenarios:
- “Tell me about a time you influenced a roadmap without authority.”
- “Describe a situation where an experiment contradicted stakeholder intuition. What did you do?”
- “Share a time you made a call that protected customers at the cost of a metric.”
- “When a case was outside your domain, how did you get to a sound recommendation?”
