What is the difference between a data analyst and a data engineer?

We get this question more often than any other, and usually not as a question but as a job advert: wanted, a data analyst who also builds pipelines. That is not one role. That is two.

What does a data analyst do?

A data analyst turns raw numbers into answers somebody can act on. The work starts with the question behind the question: not “produce last month’s report”, but “what decision do you want to take, and what do you need to know for it”.

Strong analysts are therefore as much a conversation partner as a calculator. They set up analyses, explain trends and bring the people involved along to the outcome. That last part is not a detail: an analysis nobody retells changes nothing.

What does a data engineer do?

A data engineer builds the foundation: pipelines, platforms and integrations every other role leans on. They unlock sources, connect them, and set up maintenance and monitoring so the data is still there tomorrow.

Without that role, every dashboard stays manual work. Data that is moved by hand goes wrong at some point, and usually at a moment when nobody is looking.

The difference, in one sentence

The analyst uses the data; the engineer makes sure it is there by itself every day.

That is also why the two roles do not replace each other. An analyst without reliable supply spends the day cleaning instead of thinking. An engineer without an analyst builds a tidy pipeline nobody asks a question of.

Why job adverts still ask for both

Because from the outside it looks like one field. Both roles work with the same data, use partly the same words and often sit in the same team. The difference is in what they do all day, and that rarely makes it into the advert.

The result is predictable: the vacancy stays open for months, because the people who fit one half drop out on the other. Put both in one vacancy and you are looking for two people at once.

Which one do you need first?

It depends on where things stall, and that is usually clear within ten minutes:

• Are the same files combined by hand every month, or does a dashboard stop because one source changed? Then you need an engineer.
• Is the data there, but nobody knows exactly what a number means or which decision follows from it? Then you need an analyst.

And if you can only hire one: start on the side where it hurts now. The other side can be filled temporarily, and that is a good deal cheaper than searching for someone who masters both trades at the same level.