Executive Summary
The World Economic Forum's Future of Jobs Report 2025 estimates that 39%2 of workers' existing skill sets will be transformed or become outdated over 2025–2030. This article provides a complete, actionable 5-year framework: including quarterly audit templates, the 70-20-10 learning model updated for the AI age, and investment frameworks for employers.
39%of workers' existing skill sets expected to be transformed or become outdated, 2025–2030
The one disruption figure its publisher actually prints
World Economic Forum, Future of Jobs Report 2025. An earlier version printed a four-step timetable (23%, 44%, 62%, 72%) under the same attribution; the report prints no such ladder and the rows were withdrawn on 2026-09-12. Read 39% as a five-year horizon, not a countdown.
Source 2 World Economic Forum
The Disruption Timeline: What Is Coming
The World Economic Forum's Future of Jobs Report 2025 prints one horizon figure rather than a year-by-year timetable: employers expect 39% of workers' existing skill sets to be transformed or become outdated between 2025 and 2030. Read it as a sequencing instruction, what changes first in your own role, not as a countdown.
The Updated 70-20-10 Model for the AI Age
The traditional 70-20-10 learning model (70% on-the-job, 20% social learning, 10% formal training) was developed by the Center for Creative Leadership in the 1980s. Four decades later, it needs an update for the AI era:
The New 40-25-20-15 Model
| Component | Old | New | Why |
|---|---|---|---|
| On-the-job learning | 70% | 40% | Job tasks are changing faster; on-the-job alone creates "recency bias" |
| AI-augmented practice | : | 25% | AI tutors, simulated scenarios, accelerated feedback loops |
| Social/peer learning | 20% | 20% | Human connection and mentorship remain irreplaceable |
| Formal training | 10% | 15% | Structured learning is more critical when foundational knowledge shifts |
The 5-Year Strategy Template
Year 1: Assessment and Foundation
Quarter 1: Complete a comprehensive skills audit. Use the OECD Skills for Jobs database, LinkedIn Skills Assessments, and your employer's competency framework. Map every skill you possess against the surplus/deficit data from the OECD Skills Outlook 2025.
Quarter 2: Define your Core/Adjacent/Aspirational allocation (see our Skills Architecture article). Set SMART goals for each category. Identify your 3 highest-value adjacent skills.
Quarter 3: Begin formal training in your #1 adjacent skill. Simultaneously, start integrating AI tools into your daily work (25% allocation).
Quarter 4: First quarterly review. Adjust based on market signals. Create your first "intersection project": work that combines your core + new adjacent skill.
Year 2: Deepening and Differentiating
Focus: Move from competent to proficient in your first adjacent skill while beginning your second. Year 2 is where most plans stall: the novelty has gone and the payoff has not yet arrived. Build accountability structures: study groups, public commitments, mentorship.
Year 3: The Compound Year
This is where compound stacking produces visible results. Your skill combinations should be generating unique value: projects others cannot do, perspectives others do not have. Target: 2 complete adjacent skills producing combination premiums. Begin your Aspirational skill exploration.
Year 4: Strategic Positioning
Shift from skill acquisition to skill signaling. Publish, present, mentor, contribute. Build your reputation around your unique skill architecture. Signalling is worth doing, and this desk prices it at nothing: no dated edition of any published survey we could open measures inbound opportunities against profile signalling, so no figure stands here. The withdrawn number is listed in the revision history at the end of this dossier.
Year 5: Architecture Review and Reset
Complete a full architecture review. What was "adjacent" in Year 1 should be "core" by Year 5. What was "aspirational" should be "adjacent." Set a new 5-year cycle with updated market intelligence.
The Quarterly Skill Audit Framework
Every 90 days, answer these 7 questions:
- Which of my core skills has the market valued more or less since last quarter? (Check OECD Skills for Jobs, LinkedIn job postings)
- Have I deepened my core by at least 10% this quarter? (Measured by projects completed, feedback received, assessments passed)
- What is the current salary premium for my target adjacent skill? (Check Glassdoor, LinkedIn Salary Insights)
- Have I created at least one intersection project this quarter?
- What new technology or methodology has emerged that affects my core?
- Am I visible? (Posts, talks, publications, open-source contributions)
- What will I stop doing this quarter? (Skill pruning is as important as skill adding)
🏢 For Employers: L&D Investment Framework
This desk structures L&D investment against a clear hierarchy. The tiering below is our framework, not an OECD recommendation:
- Tier 1: Critical Skills (50% of budget): Skills where market shortages directly impact revenue. AI, cybersecurity, and domain-specific technical skills.
- Tier 2: Enabling Skills (30% of budget): Skills that amplify Tier 1 effectiveness. Leadership, communication, project management.
- Tier 3: Future-Horizon Skills (20% of budget): Skills with long-term strategic value but no immediate ROI. Quantum computing, synthetic biology, space economy.
No document this desk could open prices this allocation against an even one, so the dossier prints no multiple. Treat the tiering as a rule for ordering attention and test it against your own measured outcomes.
Key metric: Track "time to competency", the average time from the start of training to productive use of the skill. No OECD edition publishes an average for it, so measure your own baseline and compare each cohort against that rather than against a borrowed benchmark.
Set that horizon against what Europe records people doing. Eurostat reports that 47% of adults aged 25 to 64 took part in education or training in the previous 12 months in 2022, and 64.2% in informal learning over the same period. Source(s): Eurostat, Adult learning statistics (Statistics Explained, data extracted May 2026). The scarce thing is not activity. It is order.
How this dossier sequences a five-year plan
- Name the horizon, not the deadline: 39% of existing skill sets transformed or outdated over 2025 to 2030 is a price on change, and it dates nothing individually.
- Take inventory of proof, not of skills: for each layer you hold, write down the artefact a stranger could inspect. Layers with no artefact are claims.
- Apply the evidence test: the next skill worth adding is one a task available to you now can demonstrate. If no such task exists, the skill is not next.
- Apply the decay test: keep only what remains useful if the tool changes. A method survives a tool release; a tool tutorial does not.
- Apply the scarcity test: prefer the combination your market cannot easily assemble from two ordinary profiles.
- Review quarterly against outcomes you measured yourself, since no publisher prints an average time to competency you could benchmark against.
Read the order rather than the list. This is the desk’s own sequencing rule, stated so it can be argued with; the only borrowed figure in it is the horizon.
Sources
- World Economic Forum (2025). Future of Jobs Report 2025. Geneva.
- OECD (2025). Skills Outlook 2025. OECD Publishing, Paris.
- Coursera (2024). Global Skills Report 2024.
- OECD (2024). Employment Outlook 2024. OECD Publishing, Paris.
📚 Skills Mastery Series: Continue Reading
This article is part of CareerOn's 5-part flagship series on skills strategy, backed by OECD, WEF, and LinkedIn data.
Sequencing beats collecting
Most five-year skill plans fail for a structural reason: they are lists, not sequences. A list assumes skills are independent and additive. In practice they compound, and the order determines the return. Foundations (statistical reasoning, writing that survives scrutiny, one domain understood deeply enough to spot when a number is wrong) raise the yield of everything learned afterwards. Tool skills learned first decay before the foundation arrives to make them useful.
A defensible sequence therefore front-loads durable capability and back-loads specificity. Year one and two: one analytical foundation and one domain, deep enough to produce original work rather than summaries. Year three: the tooling that makes that work faster and shareable. Year four: the systems layer: how your output is used by others, what breaks, and what it costs. Year five: leverage: teaching, standard-setting, or building something others depend on. The point is not the calendar precision; it is that each stage is chosen because the previous one made it worth more.
Three tests to apply before adding anything
- The evidence test. At the end of the learning, what artefact exists that a sceptical stranger could evaluate? If the answer is a certificate, the return is weak. If it is a documented analysis, a shipped system, or a measured improvement, the return is real.
- The decay test. How much of this is still valuable in five years? Vendor-specific interfaces decay fastest; measurement discipline, causal reasoning, negotiation and clear writing decay slowest. A portfolio weighted entirely toward fast-decaying skills requires permanent re-investment simply to stand still.
- The scarcity test. Is the constraint in your target market the supply of this skill, or the supply of people who can combine it with something else? Increasingly it is the latter, which means the marginal hour is often better spent on the second, unfashionable half of a pair than on deepening the fashionable half.
Two disciplines make the plan survive contact with reality. First, review quarterly against evidence produced rather than hours consumed; a plan with no artefact after three months is not behind, it is wrong. Second, budget for abandonment: expect to drop roughly one commitment a year as the market moves, and treat that as the plan working rather than failing. A five-year strategy is not a promise about the future; it is a mechanism for making each year's learning worth more than it would have been alone, and that mechanism only works if it is sequenced, evidenced and revised.
What the European statistics actually constrain
A five-year plan is a bet on scarcity, so it is worth checking whether scarcity is where the plan assumes it is. Eurostat put the EU job vacancy rate at 2.0% in the second quarter of 2026, down from 2.1% in the first quarter and from 2.1% a year earlier (Eurostat, Job vacancy statistics, 15 September 2026). Aggregate hiring pressure is easing, not building. A strategy that rests on the phrase "everyone is hiring" is resting on a number that is going the other way.
The fashionable half of the market shows the same shape. ICT specialists were 5.0% of EU employment in 2025 against 3.5% in 2015, growing 4.8% a year while total employment grew 0.9% a year, but annual growth fell to 2.6% between 2024 and 2025 (Eurostat, ICT specialists in employment, 2 June 2026). The category is still expanding and still decelerating. That is the profile of a market where entry is getting more crowded rather than less, which is an argument for pairing a technical skill with a domain rather than deepening it alone.
The credential half is already saturated. 85.7% of people aged 20 to 24 in the EU held at least an upper secondary qualification in 2025, and tertiary attainment among men aged 25 to 34 was 39.3% (Eurostat, Educational attainment statistics, 27 May 2026). Meanwhile the EU employment rate for tertiary graduates reached 86.8% in 2025 against 83.2% in 2016, while the over-qualification rate stood at 21.4% (Eurostat, Employment annual statistics, 17 April 2026). Read together, those two figures say something a course catalogue never will: a qualification reliably buys work, and roughly one worker in five is doing work below the level of the qualification they hold. The scarce thing is not the certificate. It is the evidence that you can do the work the certificate implies.
That is why the sequence in this dossier front-loads artefacts rather than enrolments. The statistics do not tell you which skill to learn. They tell you that the market clearing your plan has to survive is a market with easing vacancies, a decelerating technical premium and a saturated credential baseline, and that in such a market the only durable advantage is work a sceptical stranger can inspect.