The ROI of Learning: Unlocking the Value in Your Talent
Learning creates value only when organisations can connect investment to proficiency, proficiency to performance, and performance to measurable business outcomes.
Organisations are entering a period in which some skills are becoming increasingly valuable while others are declining in relevance. The World Economic Forum estimates that 39% of workers’ core skills will change by 2030, while 63% of employers identify skills gaps as a major barrier to business transformation. These shifts are occurring amid rapid digital transformation, requiring organisations to reconsider their workforce-planning assumptions. The optimal response is not simply to increase training volumes, thereby placing further strain on the workforce. The key is to identify learning gaps and focus on the development of the critical skills that predict future performance.
The central question is not how much learning an organisation delivers, but whether that learning improves proficiency, changes workplace performance and creates economic value.
For many organisations, learning is still evaluated through activity: enrolments, attendance, completion rates, hours delivered and learner satisfaction. These measures describe participation rather than return. A programme may be well received yet fail to improve capability, or may develop capabilities that have little bearing on future performance. It may increase knowledge without changing behaviour, or change behaviour without producing a material business outcome. The return on learning can therefore be established only by tracing a credible chain from investment to proficiency, from proficiency to performance, and from performance to organisational value.
Start with what predicts performance
The first requirement is precision. Learning should begin with job profiles and a task-based understanding of what successful performance requires. The first step is to identify the competencies, skills, knowledge and mindset that are highly correlated with performance in a particular role. This provides a stronger basis for development than generic competency lists or broad course catalogues.
A competency evaluation can then establish each learner’s starting point relative to the role and an appropriate peer group. Instead of assigning the same curriculum to everyone, the organisation can identify individual development gaps and construct an individualised learning journey. This concentrates investment where it is most likely to influence performance and avoids spending on material the learner has already mastered.
The wider model links an individual’s profile to a recommendation engine, learning-journey design, curated and newly created content, coaching and mentoring. AI can also be leveraged to connect learning with organisational knowledge by ingesting policies, procedures, practices, deliverables and input from subject-matter experts. Learning is therefore treated as part of an integrated capability system rather than isolated learning events.
Measure proficiency, not completion
The second requirement is to determine whether learning has produced a meaningful improvement in proficiency. Completion does not prove competence. A learner may remember a concept without being able to integrate it into work, demonstrate it consistently or apply it under pressure. Summative evaluations should therefore test capability against the role’s predictive requirements rather than merely testing recall. The resulting proficiency gain becomes the first indicator of learning effectiveness: how much capability improved and what each percentage point of improvement cost.
Connect capability to workplace performance
The third requirement is attribution. A proficiency gain creates value only when it contributes to improved performance. Research on training transfer shows that learner characteristics, intervention design and the work environment all influence whether new capability is applied.
A holistic view of performance reflects this complexity by measuring the execution of individual tasks, contributions to colleagues’ performance, leadership impact and team outcomes, while recognising the influence of culture, institutional systems and the broader operating environment on performance. Evidence can be drawn from self-assessment, 180-degree and 360-degree feedback, manager review, analysis of work outputs and relevant key performance indicators.
This approach avoids attributing every movement in performance to training alone. It can reveal whether improved capability is being enabled or constrained by leadership, processes, incentives, technology or culture. It also supports measurement from individual performance through to team, divisional and organisational outcomes.
Calculate the return on investment
The final calculation begins with the full cost of learning. This includes direct costs such as content, platforms and facilitators; indirect costs such as travel and employee time; and opportunity costs arising from productive effort redirected toward learning. These costs should be compared with proficiency gain, performance gain and the financial value attributable to that improvement.
The measurement chain can therefore be expressed as: cost per learner; cost per percentage point of proficiency gained; performance improvement associated with proficiency gain; cost per percentage point of performance gained; and ultimately, economic value generated per dollar invested. Depending on the intervention, benefits may include higher productivity, increased sales, improved quality, fewer errors, lower operational risk, reduced attrition or faster time to competence.
We found that up to 87% of corporate learning adds no significant value because it focuses neither on the predictors of performance nor on individual learning gaps.
Artificial intelligence (AI) makes this model practical at scale. It can support competency mapping, evaluations, recommendation models, individualised learning journeys, content curation and creation, coaching, and integrated knowledge management. Continuous performance feedback can then refine both the learning strategy and the calculation of realised benefits.
The question is no longer how much learning was delivered, but how much organisational value it created.
Learning should not be treated as a catalogue of courses or a discretionary employee benefit. It is an investment in productive capability. Its value is unlocked when organisations identify what predicts performance, develop it deliberately, measure proficiency gains, observe transfer into work and quantify the resulting business benefit.
The value of AI is therefore not that it enables organisations to deliver more learning. It enables them to direct learning toward the capabilities that matter, individualise development at scale and measure whether improved proficiency translates into performance and economic value. The future of corporate learning will be defined not by participation, but by demonstrable contribution to workforce capability and organisational performance.
By Dr. Juan Swartz Ph.D. CA, Chief AI Officer at AI Unlox
Sources and model references
• World Economic Forum, The Future of Jobs Report 2025: Skills Outlook.
• World Economic Forum, The Future of Jobs Report 2025: Workforce Strategies.
Dr. Juan Swartz Ph.D. CA is the Chief AI Officer at AI Unlox, an AI-powered learning and talent development platform. AI Unlox joined us at the 7th HR & People Development Summit in Dubai and the 26th People Development Summit, both in 2026 – come and see them at one of our upcoming Summits to explore these models in action.