Chapter 11 · Data Analyst
Landing the role
~8 min read
Everything so far builds toward this. Analyst interviews test four things: SQL, analytical and case thinking, communication, and behavioral fit. Here is how they are structured and a realistic plan to get ready.
11.1 The four rounds#
| Round | What is tested | How to prepare |
|---|---|---|
| SQL screen | Writing correct queries live | Drill joins, aggregation, and windows out loud |
| Case / analytical | Structuring an ambiguous problem | Practice the metric-dropped diagnostic |
| Metrics / business sense | Defining and reasoning about metrics | Learn KPIs for the company's domain |
| Behavioral | Communication, ownership, fit | Prepare STAR stories with real impact |
11.2 The diagnostic framework for case questions#
"Sales dropped 20 percent last week, what happened" is the archetypal case. Interviewers want structure, not a lucky guess. A reliable approach:
- Clarify. Time range, which metric exactly, and is the drop real or a tracking bug?
- Segment. Break by channel, platform, geography, and new versus returning to isolate where the drop lives.
- Internal versus external. Did we ship something, change a price, or break a flow? Or is it seasonality, a holiday, a competitor, an outage?
- Hypothesis and confirmation. Name the most likely cause and the exact query you would run to confirm it.
11.3 An eight-week plan#
| Weeks | Focus | Concrete goal |
|---|---|---|
| 1 to 2 | SQL fluency | Fifty or more problems; joins, aggregation, windows automatic |
| 3 | Warehousing and data quality | Explain star schema, ELT, and cleaning clearly |
| 4 | Statistics and experiments | Explain p-values and A/B basics plainly |
| 5 | Cases and business sense | Run the diagnostic framework on ten prompts |
| 6 | Portfolio project | One end-to-end analysis with a written story |
| 7 to 8 | Mocks and behavioral | Five mock interviews; six polished STAR stories |
11.4 The portfolio project that gets callbacks#
One genuinely good end-to-end project beats ten certificates. Take a real dataset, ask a real question, and produce a short written analysis: the question, your approach, the cleaning you did, the finding, a clear chart, and a recommendation. It demonstrates SQL, analysis, visualization, and communication, which is the entire stack of this guide, in one artifact a reviewer can skim in two minutes.
11.5 Common mistakes#
- Jumping to a query before understanding the question. Clarify first; interviewers hire the structure, not the reflex.
- Reciting a metric with no denominator. Precision about the base signals experience.
- Never checking whether the data is correct. Raising instrumentation and reconciliation marks you as someone who has shipped real work.
- Going silent while thinking. Narrate your reasoning; silence reads as being stuck.
- Ignoring the domain. Showing awareness of the industry's data and rules sets you apart from purely technical candidates.
Get the next chapter and weekly interview tips by email
One short email per week. Skim in a minute. Unsubscribe anytime.