Data engineer vs data analyst: roles, tools, day-to-day work, and which path fits your team

Part of our topic guide on Data Engineering.

Data engineers build the systems that make data usable; data analysts use those systems to answer business questions. Both write SQL, both think in tables, both care about correct data — but the work splits at the boundary between infrastructure (engineers, upstream) and decisions (analysts, downstream).

The short version: if your dashboards are wrong, late, or missing, you have a data engineering problem. If your dashboards are fine but no one is acting on them, you have a data analyst problem. On a small team one person sometimes does both — but the work, the tooling, and the day-to-day rhythm are different enough that it matters which role you hire first.

Side-by-side at a glance

For current UK salary ranges, ITJobsWatch and the ONS Labour Force Survey are the closest things to live data — both update faster than any benchmark we could quote here, and the gap between engineer and analyst salaries shifts with hiring cycles.

What the work actually looks like

A data engineer's week is mostly building or maintaining data pipelines — the code and infrastructure that moves raw data from operational systems (CRM, billing, product database, third-party APIs) into the warehouse in a clean shape. They also own data quality, schema design, access control, and the on-call rota when a pipeline breaks at 3am.

A data analyst's week is mostly turning that curated data into answers. They sit closer to the business — interviewing stakeholders about what they're trying to decide, building or updating dashboards, writing analysis documents, and presenting findings. Where an engineer's success looks like "the data is right, fresh, and trusted", an analyst's success looks like "the team made a better call because of this analysis".

Where the role boundary blurs

On smaller data teams (5 people or fewer) one person typically wears both hats and the title is usually "data analyst" until the pipeline workload forces a split. That works until the team needs more than one warehouse, a formal data contract with engineering, or a real testing discipline — at which point the engineering work becomes a full-time job and a dedicated hire pays for itself within months.

The analytics engineer title sits between the two. Analytics engineers own the transform layer — taking raw warehouse tables and modelling them into clean, well-named, trusted datasets for analysts to query. In the UK the role is still uncommon outside scale-ups; most teams fold the work into either the engineer or the analyst job depending on where the heavier need is.

Which one should you hire first?

For a team starting from zero, the order depends on what's missing:

• No warehouse or trusted data yet — hire an engineer first. There is nothing for an analyst to analyse, and an analyst left to wrangle CSVs will produce inconsistent answers and burn out within a quarter.
• Warehouse exists, dashboards are wrong — hire an engineer to fix pipeline reliability. Hiring another analyst on top of broken data multiplies the wrong answers.
• Data is reliable but no one acts on it — hire an analyst. Add business context, framing, and the storytelling craft that turns numbers into decisions.
• Both look broken — hire the engineer first, because the analyst's output depends on the engineer's foundations. A skilled analyst on broken data still produces a broken answer.

Training apprentice data professionals across UK employers since 2022, we've seen the wrong order cost teams roughly a year of compensating — typically with a senior analyst pulled into junior engineering work that neither role is set up to ship cleanly.

Career path between the two

Crossing from analyst to engineer is a well-trodden path — analyst SQL and warehouse intuition transfers directly. The bridge is usually two skills: production code discipline (version control, testing, CI/CD applied to data) and infrastructure literacy (orchestration, cloud warehouses, schema evolution). Our broader writeup on what data engineering is walks through the modern stack engineers learn.

Crossing the other direction — engineer to analyst — is straightforward when the engineer has strong stakeholder instincts and wants to be closer to the business decision. The hard part is usually the communication craft, not the SQL.

Apprenticeships into either role

Both paths run through the UK Apprenticeship Levy:

• Data engineer — Level 5 Data Engineering apprenticeship, 18 months, 6 hours/week study time alongside a real engineering job. Apprentices ship production pipelines from week one.
• Data analyst — Level 4 Data Analytics apprenticeship, 14 months, same time commitment, ends in a portfolio of real business analyses the apprentice has shipped at their employer.

For non-levy-paying employers or self-funders, the Skills Bootcamp in Data Analytics is the fastest route into an analyst role; the engineering equivalent runs through the apprenticeship.