SHIFT.The demand-for-work indexEdition 1 · August 2026
← Insights
Foundations·Jul 15, 2026·4 min read

What we mean by 'demand for work'

Every number on SHIFT counts one thing — a job advertisement. What that measures, what it stands in for, and what it deliberately leaves out.

"Demand for work" sounds like a plain phrase, but it hides at least four different things. It could mean how many people a field employs, how many vacancies sit open right now, how much those jobs pay, or how many employers are actively looking to hire. Each is a real number, and each moves for its own reasons. SHIFT measures exactly one of them, on purpose — and the whole index depends on being clear about which.

SHIFT counts advertised demand: the volume of job advertisements employers post for a role, in a market, over time. When a company decides it wants to hire a registered nurse and publishes an advertisement to find one, that is the event we count. Nothing before it, nothing after it — the act of advertising itself.

Why an advertisement

A job advertisement is a rare thing in labour data: it is public, it is dated, and it means the same thing this year as it did last year. An employer does not post one unless it currently wants to pay someone to do that work. That makes the advertisement a clean, repeatable signal of intent — an employer, at a moment in time, choosing to seek a particular kind of worker. Count those choices, and compare a kind of work against the whole market it sits in, and you can watch the intent to hire rise and fall without ever needing to know a single salary or headcount.

Each advertisement is a vote that this kind of work is worth paying for — now, in this market. SHIFT counts the votes.

What an advertisement is not

An advertisement is a request, not a result. It tells you an employer went looking; it does not tell you they hired, how long the role stayed open, or what they eventually paid. So SHIFT is not a count of jobs filled, not a measure of how many people work in a field, and not a salary tracker. It is also not a measure of how many people want the job — that is the supply of workers, the other side of the market entirely, and we don't touch it. Demand, here, means only the pull from employers.

The blind spots we accept

Some hiring never becomes an advertisement. Roles filled by word of mouth, internal promotions, informal arrangements, and work brokered on gig platforms all sit outside what we can count. We don't estimate around those gaps or pretend to see them — we measure the advertised layer of the labour market, and we say plainly that it is a layer, not the whole. Where advertising practice itself changes — a sector moving on- or off-line, a shift in how job titles are written — we treat that as a limit on interpretation, not a hidden fact to smooth over.

So when SHIFT says demand for a role is rising, it means something narrow and exact: employers in that market are advertising to hire it more than they used to. Narrow — but public, datable, and the same measurement in both years we counted. That is the trade we make everywhere on this site: a smaller claim you can trust over a larger one you can't.

More readings
The demand-for-work index

Read the shape, not the noise.

Every kind of work counted the same way, in nine languages, and compared against the market it sits in — the direction and pace of change, honestly.