Search for “most in-demand skills” and you get dozens of near-identical lists, with no source and no date. This article tries the opposite. It starts from three public, checkable sources (the World Economic Forum’s Future of Jobs Report, the Excelsior survey run by Unioncamere-ANPAL for the Italian labour market, and Eurostat’s digital skills indicators) and pulls out what’s actually worth knowing. Where the data stops and my reading begins is marked in the text.
Where the numbers come from
The Future of Jobs Report is the most quoted source in the world on this topic, and it’s worth knowing how it’s built before trusting it: it rests on a survey of more than a thousand employers, together employing over 14 million people across 55 economies and 22 industries. The key thing to understand is that it doesn’t measure the labour market: it measures employers’ expectations about the next five years. Excellent as a compass for direction, poor as a source of precise forecasts, and anyone quoting it like a census is overselling it.
For the Italian market the closest source is the Excelsior Information System run by Unioncamere and ANPAL, which surveys Italian firms’ hiring plans monthly and, more interestingly, how hard they find it to fill them.
The one figure that matters most
If I had to keep a single number from the 2025 edition of the WEF report, it would be this: employers expect around 39% of the key skills their roles require to change by 2030. Not that 39% of jobs disappear, but that the skills needed to do them shift.
That distinction changes the right strategy entirely. If the risk were the role vanishing, the rational response would be to switch careers. If the role stays but its content changes, the rational response is to stay put and update what you can do. The second is by far the more likely scenario for most people.
The same report estimates that roughly 59% of workers will need training by 2030 and, in a less quoted and more uncomfortable figure, that a significant share of them won’t get it. Translated: individual initiative matters, because waiting for your employer to train you is a bet that statistically loses.
The fastest-growing skills
The report separates the most common skills from the fastest-growing ones. They’re two different lists, and mixing them up is why you find contradictory rankings everywhere.
Fastest-growing (where demand is accelerating):
| Skill | Why it’s accelerating |
|---|---|
| AI and big data | From technical specialism to a cross-cutting requirement, including non-technical roles |
| Networks and cybersecurity | The attack surface grows with every digitised process |
| Technological literacy | Using, evaluating and choosing digital tools rather than just enduring them |
| Creative thinking | The thing automation replicates worst |
| Resilience, flexibility, agility | A direct consequence of how fast roles change |
| Curiosity and lifelong learning | It becomes the meta-skill once 39% of the rest changes |
| Leadership and social influence | Coordinating people stays hard to automate |
Most in demand right now, by contrast, is led by analytical thinking, which the report identifies as a core skill for the large majority of the employers surveyed: the ability to break a problem down, weigh the available information and reach a defensible decision.
The interesting part is how little the two lists overlap. Technical skills grow faster, but cognitive and interpersonal skills are still what people get selected on. Bet only on the first and you end up technically good and professionally stuck; bet only on the second and you stay likeable and replaceable.
What the mismatch actually looks like
Here the picture shifts, and it’s worth looking at separately, because global reports describe a market that isn’t yours.
The most telling figure from recent Excelsior surveys is the difficulty of filling roles: the share of planned hires for which firms say they struggle to find suitable candidates sits consistently around 45-50%. Almost one hire in two. The stated causes split in two: no candidates at all, and candidates who show up without the required skills.
There’s a practical consequence most people underrate: the problem isn’t only a shortage of openings, it’s the mismatch. Positions stay unfilled while people are looking for work. Anyone who lands exactly where the mismatch is has far more negotiating leverage than average. That’s also why asking for a raise works much better from a profile that’s hard to replace.
On the digital side, the EU’s Digital Decade target is for 80% of adults to have at least basic digital skills by 2030. Eurostat data shows Italy still below the European average on that indicator. Bad news for the country, good news for the individual: the level of digital skill that makes you competitive is lower than you think.
The three skills I’d bet on this year
From here on this is my own reading, flagged as such. Crossing the three sources, if I had to choose where to put a year’s worth of learning time, I’d pick these three in this order.
1. Practical use of generative AI in your own job
Not “understanding AI” in the abstract: being able to use it to do faster what you already do. It has the best ratio of learning time to immediate return, and unlike almost everything else it doesn’t require changing roles to be useful. The part that genuinely makes the difference is writing precise instructions and checking the output, not knowing twenty different tools: I’ve written a guide on writing effective prompts and one on using ChatGPT at work.
One thing enthusiastic courses tend to skip: the value is added by whoever checks. Generative tools produce plausible wrong statements with exactly the same confidence as correct ones, and the responsibility for what you deliver stays yours.
2. Reading and presenting data
You don’t need to become a data analyst. You need to read a spreadsheet without being fooled by it, build a chart that doesn’t lie, and back a claim with a number someone else can check. It’s the skill that turns an opinion into a proposal, and it’s surprisingly rare even in senior roles. The concrete starting point is mastering a handful of functions: I’ve collected them in Excel at work: 10 functions that save you hours.
3. Basic security hygiene
It shows up among the fastest-growing skills in the WEF report, and in daily practice it means something far more mundane than the name suggests: recognising a phishing attempt, managing passwords in a non-suicidal way, understanding what can and cannot go into a cloud tool. Italy’s National Cybersecurity Agency and ENISA publish free, reliable material for non-specialists. The basics anyone working with a computer needs are in cybersecurity: the basics everyone needs.
The soft skills holding all of it up
A point technical reports often leave implicit: interpersonal skills aren’t a consolation prize for people who can’t code. The WEF core-skills list consistently includes analytical thinking, resilience, leadership and the ability to learn, and in the Excelsior surveys Italian firms name flexibility, teamwork and problem solving among their most frequent requirements.
The practical reason is simple: a technical skill gets you through screening, a soft skill determines what you’re trusted with afterwards. Which ones actually count and how to demonstrate them is the subject of the most in-demand soft skills.
Turning this into something doable
The problem with articles like this one is that they end with a list of ten things to learn, which is the best way to learn none of them. The formulation that works is the opposite: one skill at a time, with a deadline and a checkable output.
A method you can apply in a month, with observable goals instead of good intentions, is the one described in how to learn a new skill in 30 days. And if you need to work out which skill first, the most useful criterion isn’t which is most in demand overall, but which is most in demand where you’re already credible: the same skill is worth far more when it adds to what you already do, instead of starting from zero in a field where you’re the newest arrival.
Frequently asked questions
Are technical skills becoming more important than interpersonal ones? No, they’re growing faster, which is a different thing. The WEF report shows acceleration on technology skills and persistence of cognitive and interpersonal skills among core requirements. The first change the list, the second remain the filter.
Do I need a degree or a certificate for these skills to count? It depends on the role, but for digital and AI skills the market looks far more at what you can produce than at which certificate you hold. A concrete, checkable output weighs more than an attendance certificate.
How long do these forecasts hold? The WEF report has a 2030 horizon and is updated periodically: the figures cited here should be re-read when the next edition comes out. That’s why you find the last-updated date at the top of this article.
If 39% of skills change, does what I know today become useless? No: it means roughly six skills in ten stay valid. The point of continuous learning isn’t starting over, it’s adding the missing piece to what already works.
Sources
- World Economic Forum, Future of Jobs Report 2025. 2030 horizon, survey of over 1,000 employers across 55 economies and 22 industries.
- Unioncamere and ANPAL, Excelsior Information System. Survey of Italian firms’ hiring plans and difficulty filling roles.
- Eurostat, Digital economy and society. Indicators on basic digital skills and Digital Decade targets.
- Italian National Cybersecurity Agency and ENISA. Free introductory material on basic information security.



