Published 9 September 2026 · Updated 10 October 2026 · By Marie Castelli, Partner, Mentivis
Skills obsolescence: how do you know if your team is at risk?
Skills obsolescence is the loss of value of a skill when the technologies, methods or expectations of a job change. According to the World Economic Forum's Future of Jobs Report 2025, employers expect 39% of workers' existing skills to be transformed or become outdated between 2025 and 2030. A team is at risk when its key skills are routine, tied to a specific tool and rarely refreshed.
Pillar: Skills measurement: methods, tools and metrics
What is skills obsolescence?
A skill becomes obsolete when it no longer allows someone to perform the job at the expected level. Three mechanisms lead there:
- Automation. A task is taken over, wholly or partly, by software or AI.
- A change of tool or method. The skill was tied to software, a standard or a process that has been replaced.
- Changing expectations. The job now requires other skills (analysis, customer relations, supervising AI).
People sometimes talk about the "half-life" of skills, meaning the time after which half of what was learned loses its value. It is a useful image, not a measurement: there is no official statistic on the lifespan of a skill. In particular, the claim that the lifespan of a skill has fallen "from 30 years to 5 years according to the World Economic Forum" does not appear in the summary of the Future of Jobs Report 2025, and we could not find its source.
What does the data say?
Three public sources frame the topic:
| Source | Date | What it measures | Result |
|---|---|---|---|
| World Economic Forum, Future of Jobs Report 2025 | 7 January 2025 | Share of workers' skills that will be transformed or become outdated, according to employers | 39% over 2025-2030, compared with 44% in the 2023 edition and 57% in the 2020 edition |
| OECD, Employment Outlook 2023 | 11 July 2023 | Share of jobs in occupations at high risk of automation (all technologies, including AI) | 27% of jobs in OECD countries |
| IBM Institute for Business Value, Augmented work for an automated, AI-driven world | 11 August 2023 | Share of the workforce needing to reskill because of AI and automation, according to the executives surveyed | 40% over the following three years |
These figures are employer estimates or projections. They give an order of magnitude and a trend, not a forecast for a given team.
The World Economic Forum links the fall in the indicator between 2023 and 2025 to the growth of training: 50% of workers had completed training, reskilling or upskilling measures, compared with 41% in the 2023 edition. The same report ranks AI and big data, followed by networks and cybersecurity and technology literacy, as the fastest-growing skills.
Which signals show that a team is at risk?
Three criteria help assess a skill's exposure:
- The degree of routine. Skills that consist of applying a rule to structured data (data entry, checks against fixed criteria, producing standard documents) are the most exposed to automation.
- Dependence on a tool. A skill defined by mastery of a specific piece of software is exposed to every change in that tool. A skill defined by a type of problem to solve, whatever the tool, lasts longer.
- The share of the work AI can produce. If generative AI produces in a few minutes a result comparable to an employee's, the value of that skill falls, while the value of supervising and checking rises.
Two internal signals add to these: no training on the skill for several years, and the skill being concentrated in one or two people.
How to run the diagnosis: the proficiency / exposure matrix
A simple exercise is to list the team's 20 to 30 key skills (those whose absence would halt the activity) and score them on two axes:
- current proficiency in the team, from 1 to 5;
- exposure to obsolescence within three years, from 1 to 5, based on the three criteria above.
| Low exposure | High exposure | |
|---|---|---|
| High proficiency | Skills to maintain | Skills to evolve or redeploy |
| Low proficiency | Skills to develop as needed | Training plan priorities |
To build the list and the levels, see Skills matrix: definition, template and how to use it and Skills mapping.
What should you do with skills at risk?
- Revise the skills framework. Redefine skills around the problems to be solved rather than tools (see Skills framework).
- Train people to supervise AI. Checking, correcting and completing what an AI tool produces is a skill in itself (see Which AI skills to train first).
- Organise internal moves. Employees who master an exposed skill know the business: they are often the best candidates for related skills that are growing.
- Review the diagnosis every year, when preparing the skills development plan.