Chapter 1 · Data Analyst
What a Data Analyst actually does
~7 min read
The title covers a wide range of realities. At one company it means building dashboards; at another it means writing SQL all day to answer product questions; at a third it sits close to a junior data scientist running experiments. What unites every version is a single responsibility: converting raw data into decisions a business can trust.
1.1 The three things you are really paid for#
- Accuracy. A wrong number is worse than no number, because people act on it. A large share of the craft is defensive work: making sure the answer is correct before anyone sees it.
- Relevance. Answering the question that was actually asked, or the better question hiding behind it, rather than the one that was easiest to query.
- Clarity. A correct, relevant answer that nobody understands changes nothing. Communication is not a soft skill here. It is the deliverable.
1.2 Where the analyst sits#
Analysts are the translation layer between groups that rarely speak the same language: the business thinks in goals and money, engineering thinks in systems and data, and leadership thinks in risk and decisions. You will spend as much time in conversations as in queries, and the value you add often comes from reframing the request.
| Stakeholder | What they ask for | What they actually need |
|---|---|---|
| Product manager | How many users clicked this? | Whether the feature is worth keeping |
| Marketing | Pull last month's signups. | Which channels to fund next |
| Finance | Revenue by region, please. | Where the forecast is at risk |
| Leadership | Is the business healthy? | The few numbers that predict next quarter |
1.3 The role family and where you can grow#
| Role | Core focus | What is added versus analyst |
|---|---|---|
| Data / business analyst | SQL, dashboards, stakeholder analysis | This guide |
| Analytics engineer | dbt, modeling, pipelines | Software engineering, modeling at scale |
| Data engineer | Ingestion, warehousing, orchestration | Systems and infrastructure depth |
| Data scientist | ML, statistical modeling | Programming plus advanced statistics |
1.4 What good looks like at each level#
| Level | What they are trusted with |
|---|---|
| Junior | Answers well-specified questions accurately; reliable SQL and dashboards |
| Mid | Owns a reporting area; catches problems unprompted; improves data quality |
| Senior | Sets the metric and data strategy; is trusted in ambiguity; mentors others |
1.5 The tools you will actually use#
Job posts list long tool stacks, but the daily reality centers on a few. Learn these deeply rather than collecting shallow exposure to many.
| Tool | Role in the job | Priority |
|---|---|---|
| SQL | Extracting and shaping data from the warehouse | Essential, learn first |
| Excel / Sheets | Quick analysis, reconciliation, stakeholder-facing work | Essential |
| BI (Power BI, Tableau, Looker) | Dashboards and self-serve reporting | High |
| Python or R | Heavier wrangling, statistics, automation | Useful, role-dependent |
1.6 A week in the life#
| Day | Representative work | Skill in play |
|---|---|---|
| Monday | Refresh weekly metrics; investigate an anomaly | Monitoring, diagnostics |
| Tuesday | Pull and analyze data for a launch readout | SQL, analysis |
| Wednesday | Build or fix a dashboard for a stakeholder | BI tooling |
| Thursday | Deep-dive a business question end to end | The full workflow |
| Friday | Document a pipeline quirk; partner with data eng | Governance, collaboration |
Get the next chapter and weekly interview tips by email
One short email per week. Skim in a minute. Unsubscribe anytime.