The career promise nobody can leave over

Randstad’s 2026 employer brand research asked more than 166,000 people and 6,400 companies, across 34 markets, what makes them choose one employer over another. The top five came back in an order most of us could have guessed: salary and benefits, work-life balance, job security, work atmosphere. The fifth is the news. Career progression has joined the list, in Randstad’s words “reflecting talent’s growing focus on long-term growth and employability.”

It is worth pausing on what kind of promise that is. Salary is delivered on payday. Work-life balance is delivered this week, or it isn’t. Job security is delivered by an absence of bad news. Work atmosphere is delivered every day, in the room. Career progression is the only one of the five that is delivered later, by people who may not be there when it comes due, under conditions nobody can see yet. It is the longest-dated promise an employer can make, and it has just become one of the five things people choose you for.

In the same year, employers described their own intentions in the opposite tense. Deloitte put the question to more than 9,000 business and HR leaders across 89 countries: seven in ten said their primary competitive strategy for the next three years is to be fast and nimble. One of the three shifts the report names is the move from static plans to dynamic orchestration, which is a considered way of saying that fewer things will be settled in advance, and that settled things will be re-settled more often.

I don’t read that as hypocrisy. They are two rational answers to the same uncertainty, and they happen to point in opposite directions. The market has started choosing employers on the strength of a commitment about the future, in the year employers decided to hold fewer of them.

Then there is the question of who delivers it. Not the careers page. A career promise is delivered by a manager, in a series of ordinary conversations that mostly go unwitnessed and rarely get written down. Gallup has manager engagement at 22% for 2025, down nine points since 2022, with the steepest single-year fall between 2024 and 2025, from 27% to 22%. Engagement overall fell to 20%, its lowest since 2020. So the promise arriving at the top of the list is the one with the longest payback and the thinnest layer left to carry it.

And then the market froze. LinkedIn’s Economic Graph, which counts hiring on its own platform rather than in the economy, has the hiring rate down 10.7% year on year in the UK, 6.3% in India and 4.8% in the United States. Gallup finds 52% of people worldwide saying it is a good time to find a job: recovered from the trough, still short of 2019's 55%. In the UK specifically, the CIPD warned in August that hiring is stuck in a low-hire, low-fire cycle, and the trade press reported the next day that vacancies had fallen to their lowest level since 2014. Those last two are UK figures, and I would rather say so than let them stand in for a continent.

Here is where the three lines meet, and it is not where most of the commentary is looking. Almost no employer measures whether it keeps a career promise. It infers it, from attrition. Good people are not leaving, so the thing must be working. That inference was never strong, but it was cheap and roughly directional, and a great many employer brand dashboards rest on it.

A frozen market destroys it. Regretted attrition in a market where nobody can move is not evidence that you kept the promise. It is evidence that the door is shut. The failure mode is not that the signal disappears. It is that the signal keeps arriving, monthly, in the same format as last year, and it looks better than last year. A broken promise and a kept one now produce identical numbers, and the identical number is the flattering one.

The measurement that would settle it is the one almost nobody publishes. LinkedIn will tell you the hiring rate market by market; no equivalent public index tracks whether people are moving inside the companies that stopped hiring outside them. I don’t think that is an oversight in the data industry. It reflects what employers themselves bother to collect.

So here is the test I would apply to any progression promise before it goes near a careers page. Name where it is delivered: which manager, which conversation, which quarter. Then name the number that would tell you it had stopped, and it cannot be a number that depends on someone quitting. Two candidates, neither of them difficult: the share of open roles filled from inside, and the share of people who had a progression conversation that changed something within two quarters. Both are collectable this month. Both are unflattering, which is the actual reason they are not on the dashboard.

The freeze will end. When it does, the promise gets tested by everybody at once, by people who have been waiting a year or more to test it, and the bill arrives as a quarter of resignations that look sudden and are not. The employers who come through that will be the ones who spent the frozen year measuring something other than whether people stayed. A promise nobody can leave over is not a promise being kept. It is a promise that has not been tested yet.