SHIFT.The demand-for-work indexEdition 1 · August 2026
← ALL 23 KINDS OF WORK
The demand for your work · Edition 1 · August 2026

Data analysts & entry

Classified AI-exposed

We counted job postings whose title contains one of these terms — not a category we assigned, the literal searches that produced every number on this page:

business analystdata analystdata entrydata scientist

Plus the equivalents in eight other languages. The list of searches is fixed before the counting. How we measure.

Is this work disappearing?

Over the two years to June 2026, advertised demand for this work fell — in the world taken as a whole.

The numbers are in the document. How much it moved, how much of the market it holds, and where it stands against the other kinds of work we track.
Direction is robust. This figure is tested against five alternative ways of combining markets, fixed in advance. All five put this work on the same side. See the method.

Your country matters more than your job

We can see this work in 21 countries with enough postings to be honest about. They do not look alike. In some, this work is pulling ahead of everything around it. In others, it is falling behind.

And in 8 of them, employers advertised fewer of these jobs in the latest twelve months than two years earlier. Fewer jobs, not a smaller share. The worldwide answer above hides that completely.

Which country is which is in the document. Each of the 21 named, and what happened to this work there.

What you get for €19

Everything we know about this work, and nothing else — as long as the work needs it to be. In plain language, with every number shown.

Get the document

Later editions, quarterly updates and up to three kinds of work are the annual pass, €49 — and you can move up to it in the first 30 days for the difference.

How to read this

Postings are not jobs

One posting can be one vacancy or fifty, and some roles advertise more than others. This measures hiring activity, not headcount.

It is a group, not your desk

These terms cover a group of related titles. Two people with different jobs can match the same group.

It does not say why

We show where demand went. Whether AI, interest rates, offshoring or demography moved it is beyond what postings can prove.

Counts flatter everything

The postings we can see grew ×2.20 partly because our provider indexed more. That is why every comparison here is a share or a ratio.

Method, sources and the admission rule: Methodology