Teaching employees entirely new skills for a different role within the organization, typically triggered by automation, restructuring, or the creation of new positions.
Key Takeaways
Reskilling happens when someone's current role is disappearing or shrinking, and the organization trains them for a completely different position. A bank teller learning to become a digital customer experience specialist is reskilling. A factory assembler learning robotics programming is reskilling. A print journalist retraining as a video content producer is reskilling. The role changes. The department might change. The core skill set changes. This isn't about incremental improvement. It's a career pivot supported by the employer. Why would a company invest $10,000 to $25,000 to reskill someone when they could just hire a person who already has the skills? Three reasons. First, the employee already knows the company: its culture, processes, customers, and politics. That institutional knowledge takes years to rebuild with an external hire. Second, reskilling signals to the entire workforce that the company takes care of its people. That message drives engagement and retention across the board. Third, in tight labor markets, the people with emerging skills are expensive and hard to find. Growing your own talent is often the only realistic option at scale.
The choice between upskilling and reskilling depends on whether the employee's current role is evolving or disappearing.
| Factor | Upskilling | Reskilling |
|---|---|---|
| The role | Stays the same but requires new skills | Is eliminated, automated, or fundamentally changed |
| Learning scope | Adjacent or advanced skills within the same domain | Entirely new skill domain |
| Time commitment | 2 weeks to 6 months | 3 months to 18 months |
| Investment per person | $500 to $3,000 | $5,000 to $25,000+ |
| Risk to employee | Low: familiar territory with incremental growth | Higher: steep learning curve, new identity |
| Best trigger | Technology upgrade, new tools, process changes | Automation, restructuring, role obsolescence |
| Example scenario | Accountant learning Power BI for reporting | Accountant retraining as data analyst |
| Success rate | High: 80%+ when aligned to role needs | Moderate: 60 to 70% reach full proficiency in new role |
Reskilling isn't charity. It's a financial and strategic decision with quantifiable returns.
Laying off an employee costs 50 to 200% of their annual salary when you factor in severance, unemployment insurance, recruiter fees, onboarding time, and lost productivity during the transition. Reskilling that same employee typically costs $5,000 to $25,000 and preserves years of accumulated organizational knowledge. Amazon's Career Choice program, which pays up to 95% of tuition for employees pursuing in-demand fields, costs the company far less per head than its average cost-per-hire of $7,000+ for technical roles.
Companies that can reskill their workforce internally have more control over their talent pipeline. They don't depend on unpredictable labor markets. When Walmart needed more truck drivers, it reskilled existing associates rather than competing in a market where driver salaries had spiked 20% in a single year. JPMorgan Chase reskilled thousands of back-office employees into tech roles, creating a pipeline that would have taken 2 to 3 years to build through external hiring.
Mass layoffs make headlines and destroy employer brand. Reskilling programs make different headlines. When AT&T announced its $1 billion Future Ready initiative to reskill 100,000 employees, it became a case study in responsible workforce management. Employees across the company reported higher engagement scores because they saw a tangible commitment to their future. Glassdoor reviews and LinkedIn employer branding both benefit from visible reskilling programs.
Reskilling is harder than upskilling because you're asking someone to fundamentally change their professional identity. The process needs structure, support, and patience.
Start with workforce planning data. Which roles are shrinking? Which are growing? Map the supply-demand dynamics over a 2 to 5 year horizon. Collaborate with department heads and industry analysts to identify which roles automation, AI, or market shifts will impact. Simultaneously identify which new or expanding roles could absorb reskilled employees. The closer the cultural or contextual overlap between the old and new role, the higher the success rate.
Not every employee in an at-risk role is a good candidate for reskilling. Assess both aptitude (cognitive ability, learning agility, relevant foundational skills) and motivation (willingness to change, openness to challenge, career aspirations). Use a combination of psychometric assessments, skills tests, and candid conversations. Never force reskilling on someone who doesn't want it. A reluctant learner won't succeed, and the investment will be wasted.
Create a structured curriculum that takes the employee from their current skill level to full competency in the new role. Break it into phases: foundational knowledge (4 to 8 weeks), applied skills building (8 to 16 weeks), supervised practice in the new role (4 to 12 weeks), and independent performance with coaching (ongoing). Each phase should have clear assessments and go/no-go decision points. Assign a mentor from the target function who can provide context, answer questions, and help the employee build relationships in their new team.
Reskilling is stressful. Employees are leaving behind expertise they've built over years and starting from scratch. Provide psychological support: regular check-ins, peer cohorts (so reskilling employees don't feel isolated), access to coaching, and explicit permission to struggle. Adjust performance expectations during the transition. An accountant who's learning to be a data analyst shouldn't be held to the same output standards as a tenured data analyst during their first six months.
These large-scale examples show what reskilling looks like in practice and what results companies have achieved.
Amazon pre-pays 95% of tuition for employees pursuing education in high-demand fields like healthcare, IT, and transportation. Since 2012, over 200,000 employees have participated. The program costs Amazon approximately $1.2 billion total. Employees can reskill into roles both inside and outside Amazon, which the company views as a long-term employer brand investment. Participating employees show 30% lower turnover than non-participants, saving significant replacement costs.
Facing the shift from hardware to software-defined networking, AT&T invested $1 billion to reskill 100,000 employees. They partnered with Udacity, Coursera, and Georgia Tech to create nanodegree programs in data science, cybersecurity, and cloud computing. Within four years, reskilled employees filled 50% of the company's tech-management roles that previously would have been hired externally. Internal time-to-fill dropped by 40%.
Walmart's program pays 100% of college tuition and books for its 1.5 million US associates. Participating employees can earn degrees and certificates in business, technology, healthcare, and other fields. Since launch, over 120,000 associates have enrolled. Walmart reports that participants are promoted at 2x the rate of non-participants and are 3x more likely to be retained after three years.
Reskilling programs have a failure rate of 30 to 40% when poorly designed. These are the most common pitfalls.
Data points that quantify the reskilling imperative facing organizations worldwide.
AI is the largest reskilling trigger since the Industrial Revolution. Organizations must decide now which roles will change and how to prepare their people.
Goldman Sachs estimates that 300 million full-time jobs globally could be affected by generative AI. Roles with high exposure include data entry, basic customer service, routine document processing, scheduling, simple code generation, and financial reconciliation. But "affected" doesn't always mean "eliminated." Many roles will transform rather than disappear, requiring workers to learn how to work alongside AI tools rather than be replaced by them.
AI creates demand for roles that didn't exist five years ago: AI trainers, prompt engineers, AI ethics reviewers, human-AI workflow designers, and data quality specialists. It also amplifies demand for distinctly human skills: complex problem-solving, emotional intelligence, creative strategy, and stakeholder management. Workers reskilling from AI-affected roles into these areas bring valuable domain expertise. A customer service agent who reskills into AI chatbot training brings years of knowledge about what customers actually ask.
Don't wait until roles are already automated to start reskilling. Begin by auditing every role for AI impact: classify each as "no change," "augmented" (AI assists the human), or "transformed" (role fundamentally changes). For augmented roles, upskilling is sufficient. For transformed roles, start reskilling immediately. Create cohort-based programs rather than sending individuals through self-paced courses. Cohorts provide peer support, accountability, and a shared sense of purpose during an unsettling transition.