WeAreHuman@Work | #047 | Employable Enough to Leave: 65% want AI skills, 52% stopped waiting

New here? This is row three of ten. Start with #044 and the People Expectations Framework™, both free.

Last week, row two ended on a promise: if you cannot offer job security, offer employability, and say so out loud. This week, the row that holds organisations to it.

Employability is the most transactional promise in the deal: will my market value stay current while I work for you? Two questions sit inside it. Do my skills stay relevant? And if this job went, could I find the next one?

Most organisations answer the first question with a learning catalogue and refuse the second one outright. The reason is always the same sentence, said in budget meetings with a straight face: if we train them, they'll leave. Branson answered it in 2014, in the line at the top of this letter. Its first half is row three. Its second half belongs to the rows still to come.

So here is my position, before the evidence: keeping people employable enough to leave is the price of the deal. When you stop paying it, they don't leave. They start paying it themselves, on their own time, and they stop counting you in.

I learned the price early. At Deloitte, every one of us had to log at least 40 hours of training a year. It wasn't a perk, and nobody called it development: it was a compliance requirement, meant to show our knowledge was up to date. The firm sold our judgement, so it could not afford to let that judgement go stale. Nobody in that building asked whether the 40 hours would help us leave. Of course they would. That was never the point.

Expectation first, experience second, and then the question every training budget hides: who is it for?

— Tanguy

Row three, read twice: what people expect an employer to do for their market value, and what they get. Then the third Expectation Gap, a self-audit and five metrics.

Here is the figure for your next people review, and five things it tells you.

Half your people have already taken the job off your hands. Randstad surveyed 27,062 workers in 35 markets. AI is the training people want most, and 65% want their employer to invest more in their AI skills. Then the turn: 52% are already seeking opportunities to future-proof their skills independently, outside formal employer programmes. That is calculative withdrawal in row three: people do not complain; they reprice. The course goes on the personal card, the certification goes on the CV, and the employer quietly drops out of the calculation. In Switzerland, 63% believe AI at work will mainly benefit companies, not them (47% globally).

Most people are being used, not renewed. 44% trained in six months, flat. Over half with no recent training at all. Adam Smith saw the risk in 1776, in the same book that praised the division of labour: the worker whose whole life is spent on a few simple operations "naturally loses, therefore, the habit of such exertion, and generally becomes as stupid and ignorant as it is possible for a human creature to become." The language is of its time. The mechanism is not. A job that uses a skill without renewing it is spending someone's market value, and nobody books the depreciation.

Employability is rationed by rank, and by age. The higher you sit in the organisation, the more of it you get. Last week it was line of sight; this week it is skills. The chart below shows the staircase: 33 points from the executive floor to the front line. Workers over 55 are the least likely to say their employer invests in their skills (12% strongly agree). And people notice: in Mercer's survey, 72% see unequal access to AI tools and training, and 35% would consider leaving if they felt disadvantaged by it. The organisations most worried that training makes people leave are the ones losing people over who gets trained.

Yes, trained people leave. That is the deal with working. Here is the strongest case for the sceptics, and it is real. In EY's data, quit intent sits at 21% among employees with fewer than 4 hours of AI training and climbs to 45% among those with 81 hours or more. Grant all of that; it still doesn't make training the problem. Employability is supposed to let people leave. Whether they want to stay is decided in other rows. ADP shows the cost of getting row three wrong: among workers who strongly agree their employer invests in them, 53% are fully engaged. Where that investment is missing, only 12% are fully engaged. The choice is not between trained people who might leave and untrained people who stay. It is between people who leave fit and people who stay stale.

Name the price and pay it openly. Last week's promise was Rosabeth Moss Kanter's employability security: the confidence of being worth hiring elsewhere. Keeping it costs three things most CFOs cut first in a downturn: a written floor of learning time, the way Deloitte wrote down its 40 hours; skills that travel, with credentials people can carry; and access published by level and age, so the front line and the over-55s do not fund their own future. None of this is a learning catalogue. A catalogue nobody has the time to open is a library, not a promise. Call it what it is in the employment deal, out loud: we will keep you employable enough to leave. Then let the other nine rows give people their reasons to stay.

Keeping people employable enough to leave is the price of the deal. The reason to stay is another row.

Your turn. When did your employer last pay for a skill you could take to your next one? Hit reply; I read every one. And if you know a leadership team about to freeze the training budget "because people leave", forward this before the decision is made.

Three questions for your next leadership meeting. Yes or no.

Three "no" answers mean your people are already paying for row three themselves. The five metrics below tell you how much.

No new inventions: five established measures, read together rather than one by one.

🕰️ Row four: Flexibility — Sunday 25 October

Employability asks whether you could leave. Flexibility asks on what terms you stay: when, where and how much you work. Next week reads the fourth Expectation Gap, the distance between the policy on the intranet and the permission in the team.

Ten rows, ten Sundays. In December, subscribers receive the complete Expectation Gap: all ten rows in one workbook, with the questions and five metrics for every row. Stay for the set.

— Tanguy

Employable enough to leave — the row-three finding: keeping people's market value current is the price of the employment deal, not a retention risk. Other rows decide whether they stay. This edition's frame.

Do-it-yourself employability — what people do when the employer stops keeping its side: they fund, schedule and certify their own skills, outside the organisation. Randstad's 52%.

Calculative withdrawal — the response to a broken transactional promise: people reprice their effort and their loyalty rather than protest. In row three, it looks like a personal card course.

The Expectation Gap (row three) — the distance between "to be kept marketable outside, not only useful inside" and what people judge they are getting, read two ways: do skills stay current, and is the next job findable? This edition's measure.

Flexibility — control over when, where and how much I work, written into the terms of the deal. Read next Sunday as the gap between policy and permission. Row four.

The Expectation Gap sits inside the People Centricity Methodology™: Empathise, Strategise, Realise, Analyse. It is Empathise made measurable, the first two links of the People Centricity Formula™, read one row at a time.

Branson, R. [@richardbranson]. (2014, March 27). Train people well enough so they can leave, treat them well enough so they don't want to [Post]. X. https://x.com/richardbranson/status/449220072176107520

Dulac, T. (2026, September). People Expectations Framework™. PeopleCentriX. https://www.peoplecentrix.eu/people-expectations-framework

EY. (2025). Work reimagined survey 2025. Ernst & Young Global. https://www.ey.com/en_gl/insights/workforce/work-reimagined-survey

Kanter, R. M. (1989). When giants learn to dance: Mastering the challenges of strategy, management, and careers in the 1990s. Simon & Schuster.

ManpowerGroup. (2026). Global talent barometer 2026: Report and key findings. ManpowerGroup. https://www.manpowergroup.com/en/insights/january-2026-global-talent-barometer

Mercer. (2026). Global talent trends 2026: Driving exponential performance: Solving the human-machine equation. Mercer.

Randstad. (2026). Workmonitor 2026: The great workforce adaptation. Randstad N.V. https://www.randstad.com/workmonitor/

Richardson, N., Hayes, M., & Northup, J. (2026). Today at work, 2026 issue 1: People at work. ADP Research.

Rothwell, A., & Arnold, J. (2007). Self-perceived employability: Development and validation of a scale. Personnel Review, 36(1), 23–41. https://doi.org/10.1108/00483480710716704

Smith, A. (1776). An inquiry into the nature and causes of the wealth of nations. W. Strahan and T. Cadell.