The data analyst's role, and the roles around it
What a data analyst actually does day to day, and how the role differs from data scientist, data engineer and IT business analyst.
This lesson builds on: What data analytics is — and the four questions it answers.
- Describe the core responsibilities of a data analyst
- Distinguish the analyst from the data scientist, data engineer and IT business analyst
- Decide which role should own a given task
A day in the work of an analyst
A data analyst answers business questions with data. In practice, a typical week mixes:
- Requests — "Why did conversion drop in region West?", "Can I get revenue by product family for the board pack?"
- Data work — writing SQL, joining tables, checking that totals reconcile with Finance, cleaning messy inputs.
- Analysis — comparing periods and segments, finding drivers, quantifying effects.
- Communication — dashboards, short memos, a slide for a meeting, a conversation that changes someone's plan.
- Maintenance — keeping recurring reports and dashboards correct when source systems change.
Roughly, analysts spend far more time on understanding the question and preparing data than on fancy methods. That is normal and is where professional quality is decided.
The neighbouring roles
| Role | Core question | Main tools | Typical deliverable |
|---|---|---|---|
| Data analyst | What happened, why, and what should we do? | SQL, Excel, Power BI/Tableau, Python | Analysis, dashboard, recommendation |
| Data scientist | What can we predict or optimise with statistical models? | Python/R, ML libraries, experimentation | Model, experiment design, algorithm |
| Data engineer | How do we move, store and serve data reliably at scale? | SQL, Python, orchestration, cloud warehouses | Pipelines, data models, platforms |
| IT business analyst | What should a system do to meet business needs? | Process maps, requirements, user stories | Requirements, specifications, acceptance criteria |
Analyst vs data scientist
The boundary is fuzzy and varies between companies. A useful rule: the analyst explains and informs decisions mostly with descriptive and diagnostic methods (and simple predictive ones); the data scientist builds predictive or optimisation models as products. An analyst who forecasts next quarter's demand with a seasonal model is doing predictive analytics — the title does not matter, the rigour does.
Analyst vs data engineer
The engineer makes data available and reliable: ingestion from source systems, scheduling, data warehouse tables, monitoring. The analyst uses that data to answer questions. When an analyst finds the same cleaning step repeated in ten reports, that logic probably belongs upstream, owned by engineering.
Analyst vs IT business analyst
Despite the similar name, the IT business analyst focuses on systems and processes: gathering requirements for a new CRM, writing user stories, defining how a feature should behave. Data analysts and IT business analysts often collaborate — for example, when a new system must capture a field that analysts need for reporting.
Worked example: who owns the task?
| Task | Best owner | Why |
|---|---|---|
| "The nightly sales load failed and the dashboard is empty." | Data engineer | Pipeline reliability |
| "Which customer segments respond best to cashback?" | Data analyst | Diagnostic analysis of business data |
| "Build a real-time recommendation model for the website." | Data scientist (with engineering) | Production ML system |
| "The new CRM needs a 'lead source' field — define its allowed values." | IT business analyst, with input from the data analyst | System requirement that affects reporting |
A professional analyst does not silently patch broken pipelines inside a dashboard. Raise the issue with the owner, apply a documented temporary fix if you must, and make sure it is removed later.
Common mistakes
Small companies may expect one person to do analysis, engineering and modelling. Judge a job by its responsibilities, not its title.
Charts are outputs. The value is in correct numbers, the right question and a decision that improves.
Practice
The nightly job that loads sales data into the warehouse has failed for three days. Whose responsibility is fixing the pipeline?
A new CRM system is being configured. Which task is most typical for an IT business analyst?
Summary
- The analyst answers business questions with data and communicates the answer so that it changes decisions.
- Data scientists build predictive/optimisation models; data engineers build reliable data infrastructure; IT business analysts specify systems and processes.
- Boundaries vary by organisation — rigour and clarity of ownership matter more than titles.