The skills-based organization is mostly theatre

There is a phrase that has, over the past three years, migrated from consulting decks into earnings calls, talent strategy slides and, eventually, LinkedIn bios: the skills-based organization. It promises a company that hires, develops and deploys people not by job title but by capability. It is, on paper, the most rational way to organise modern work.

In practice, it is mostly theatre.

What people are actually buying

When a Fortune 500 announces it has “moved to a skills-based model,” what has usually happened is one of three things. A skills taxonomy has been purchased from a vendor and bolted to the existing HRIS. A small pilot in a single function — usually engineering or data — has been run for a quarter. Or a job architecture project has been quietly relabelled. None of these is the thing. The thing is operational: it is the ability to staff a project, set a salary band, or trigger a development plan from a skill signal rather than a title.

That capability requires more than a taxonomy. It requires managers who can assess skills consistently, an internal labour market that actually moves people, and a compensation system that does not collapse the moment seniority and title decouple. Almost no one has the third.

Why the rhetoric outpaces the practice

The structural reason is simple. Titles are how power, pay and identity are anchored in the modern firm. A skills-based model threatens all three at once, and asks middle managers — the people who would have to do the daily work of skill assessment — to police a system that erodes their own authority. Predictably, they don’t.

The technological reason is that “AI will infer everyone’s skills from their work” has been a year away for five years. Inference is fine for broad clusters. It is unreliable for the granularity that compensation decisions demand. So the human assessment work remains, and the human assessment work is what no one wants to do.

What’s left after the theatre

There is a smaller, more honest version of this idea that does work. Several companies have built genuine internal talent marketplaces — Schneider Electric and Unilever being the most cited — where employees opt in, projects are posted, and skills are matched at the margin. These do not replace the org chart. They sit alongside it. They move perhaps 5–15% of capacity. That is real, and that is meaningful, and it is also nothing like the rhetoric.

The mistake is treating the skills-based organisation as a transformation. It is, at best, an overlay. The companies that admit this get something useful out of it. The ones that don’t get a taxonomy, a consultant invoice, and a slide they can show the board.