How can companies forecast future skills needs?

A three-year hiring plan can tell a company that it expects to add 500 people. It cannot tell the company whether the skills attached to those 500 roles will still be the right ones three years from now.

That has become a bigger planning problem as AI changes the work inside existing jobs. ManpowerGroup’s 2026 Talent Shortage Survey found that AI model and application development and AI literacy are now among the hardest capabilities for employers to find globally. At the same time, communication, collaboration and adaptability remain highly sought after.

Companies therefore need to forecast more than headcount. They need a view of which skills the business will require, which ones it already has and where shortages are likely to appear.

What is skills forecasting?

Skills forecasting estimates the capabilities a business will need over a defined period and compares that demand with the skills likely to be available internally and externally.

This requires more than projecting today’s job vacancies forward.

Take a company planning to introduce AI across its customer operations. Its workforce plan might show that overall headcount will remain broadly stable. The skills picture could look very different. Some tasks may be automated, existing employees may need AI literacy or stronger analytical skills, and entirely new specialist capabilities may be required.

Randstad argues that this is one of the limitations of traditional headcount forecasting and recommends breaking work into tasks and skills to understand future workforce requirements more accurately.

How do companies identify the skills they will need?

Start with the work the business expects to do.

Product launches, technology investments, expansion plans, automation and changes to operating models all create signals about future demand. Workforce teams can translate those plans into the tasks and capabilities required to deliver them.

External labour-market data adds another layer. Job-posting trends, competitor hiring, salary movements and the availability of particular skills can show whether demand is increasing faster than supply.

The forecast should also be revisited. A skills plan produced once a year can become outdated quickly when technology or business priorities change.

How do companies know which skills they already have?

This is often where workforce gap analysis becomes difficult.

Job titles are a poor proxy for capability. Two employees with the same title may have very different technical knowledge, project experience or proficiency.

Organizations can build a better skills inventory using employee profiles, assessments, learning records, project history and internal mobility data.

Many companies are still building this foundation. Mercer’s 2025/2026 Skills Snapshot found that 38% of organizations maintain a single enterprise-wide skills library, while 55% map skills directly to jobs.

Without a reliable view of existing skills, a company can mistake an internal visibility problem for an external talent shortage.

How can talent analytics help forecast skills gaps?

Talent analytics can connect signals that would otherwise sit separately.

A skill might be appearing more frequently in permanent requisitions while contingent demand for the same capability is also rising. Employees with that skill may be leaving at a higher rate, while internal training takes longer than expected to produce the required proficiency.

Seen together, those signals give workforce teams something they can act on before vacancies become urgent.

This is where workforce demand forecasting becomes more useful than simply estimating future hiring volumes.

What happens after a future skills gap is identified?

Not every forecasted shortage should become a recruitment target.

A company may already employ people who can move into the work. Others may be able to develop the required capability before demand peaks. Some specialist skills may be needed only for a project and make more sense through contingent talent.

For genuinely scarce capabilities, external hiring may need to begin much earlier.

Mercer’s 2026 research shows how significant this is becoming: 65% of executives expect between 11% and 30% of their workforce to be redeployed or reskilled because of AI over the next two years.

A useful skills forecast gives enterprise workforce planning enough lead time to make those choices.

Finding out today that a capability will be scarce next year gives HR, talent acquisition and business leaders options. Finding out when the requisition lands gives them a hiring problem.