Why Durable Skills Aren’t Showing Up

Ask any company what it wants from its people and you get the same short list: communication, collaboration, judgment, adaptability, the ability to solve a problem no one has seen before. It is in the job posts. It is on the values slide. And then most organizations do almost nothing to systematically build those skills in the people they already hired.

That gap, the one you might grumble about in candidates, is the same one your own onboarding and talent development quietly reproduce. Which points to something worth sitting with: this is not a knowledge problem. It is a design problem.

The Demand is Real & it is Climbing

Start with what the market is actually paying for. America Succeeds’ Durable by Design analysis, drawn from nearly 76 million job postings, finds that 76% now request at least one durable skill, up from 64% in just four years. Nearly half ask for three or more. Eight of the ten most-requested skills in the labor market are durable ones, and the top five are requested roughly four times more often than the top five technical skills. Communication alone shows up on more than 26 million postings.

This is not an American quirk. The World Economic Forum’s Future of Jobs 2025 lands in the same place globally, naming analytical thinking the top core skill, with resilience, leadership, and creative thinking close behind. Different datasets, different continents, same conclusion.

A quick note on language, because it matters more than it looks. The fastest way to lose this argument is to call these skills “soft.” Soft makes them sound optional. There is nothing optional about the thing that gets you promoted, or the thing that gets you let go. They are durable because they outlast any single tool, title, or tech stack.

AI is Not Lowering Their Value but Repricing Them Upward

AI is very good at absorbing the routine: retrieval, first drafts, pattern-matching, the predictable middle of the work. What it does not absorb is the harder human part. The judgment about what to do with that draft. The conversation that actually moves a team. The call you make when the situation is ambiguous.

Think about your own best day at work this year. It probably was not the day you executed a process flawlessly. It was the day you read a room correctly, made a hard call with incomplete information, or got people who disagreed to move in the same direction. That work does not automate. And as AI takes the rest, its value does not disappear. It concentrates.

The data now backs this precisely. PwC’s 2026 Global AI Jobs Barometer, built on more than a billion job ads across 27 countries, finds that the skills demanded in the most AI-exposed roles are changing more than twice as fast as in the least exposed ones, with rising demand for human capabilities like judgment, creativity, and leadership. More striking: the most AI-exposed entry-level roles are now 7 times more likely to require traditionally senior skills like leadership, because AI is quietly handling the junior tasks those jobs used to be built around. The bar for a new hire moved up, not down.

If you still think the safe bet is technical depth, look closer. McKinsey’s 2026 analysis of automation exposure across roughly 10,500 skills puts software development and SQL among the more exposed skills, above empathy, leadership, and collaboration. Read that twice. The technical bet that looks safe is quietly getting riskier, and the human bet that looks soft is quietly getting safer.

So Why isn’t Anyone Building Them Well?

Because these skills do not behave like content. You cannot lecture someone into being a strong collaborator. They are developed, not delivered. And that is exactly where the design failure shows up. Three patterns repeat almost everywhere:

• Named but never taught on purpose: More than 20 states have written down the graduate they want, yet America Succeeds’ Beyond the Portrait 50-state policy scan finds most still lack the conditions to translate that vision into practice. A portrait on the wall does not change what happens on a Tuesday morning.
• Invisible to the person building them: Learners and employees develop these skills through real work, then cannot name what they built, so they cannot improve it, track it, or prove it.
• Treated as an add-on: The skill gets bolted on as a workshop or a career-center visit instead of woven through the actual work, where it would actually take hold.

Put plainly: we have gotten very good at declaring the destination and bad at building the road to it. The good news, and the reason none of this requires a new budget line, is that most of the development is already happening. It is simply unnamed. (America Succeeds’ PK-16 developmental progressions and early career performance rubric exist precisely to give organizations a shared, trackable language for skills that already show up in real work.)

Where the Pipeline Actually Breaks

Now picture the employer’s side of the table. A stack of candidates, all saying the same things about themselves, none of it carrying any real information, because everyone says it and nothing backs it up.

Consider the student who spent a semester holding together a chaotic group project. They learned more about collaboration and conflict resolution than any lecture could teach. Then, at the end, they wrote “group project” on a resume as if it were a chore they survived. The work happened. We just never gave them a clean way to prove it.

That is the break. And as hiring shifts away from the degree as a stand-in toward demonstrated skills, a weak signal costs more every year. The most valuable thing you can hand the person on the other side of the table is evidence they can actually read.

The Tension Worth Holding

So the situation is this: demand for durable skills is high and rising. AI is making them more valuable, not less, and making the gap more expensive by the year. And most of our systems, in schools and inside companies alike, still are not designed to build these skills, make them visible, or prove them.

That is the uncomfortable part. Here is the hopeful one: a set of very different schools have already cracked this, and what they do translates directly into how any organization builds and hires talent. That is where we go next.