When performance is below expectation, the immediate question is often what people need to learn. Sometimes that is the right question. Sometimes it is not.
The problem may be that people lack the necessary skills or knowledge. It may be that they have those skills but cannot yet combine and apply them effectively in the situations the work requires. Or they may already be capable, while slow systems, unclear processes, workload, management or other conditions prevent that capability from becoming performance.
Those are different problems, even when the business symptom looks similar.
If we collapse them into a generic “skills gap”, the intervention can quickly become more training. But the right response depends on where the constraint actually sits.
That is why the distinction between skills, capability and performance matters in practice. Organisations do not need more precise terminology for its own sake. Managers, L&D teams and executives need to make better decisions about the problem they are actually trying to solve.
What skills can tell us, and what they cannot
Skills give organisations something valuable: a more precise way of understanding what people know or are able to do.
A well-designed skills model can help connect people to learning, roles, opportunities and workforce needs. It can support recruitment, development, mobility and workforce planning by giving the organisation a clearer language for describing the abilities that matter.
The limitation is not the usefulness of skills. It is what we conclude from them.
Knowing that someone possesses, or has demonstrated, a particular skill tells us something important. But it does not automatically tell us whether they can draw on that skill alongside the knowledge, behaviours and judgement required to deal with the situations the work actually presents.
This distinction appears in capability and competence literature, where broader capability is often associated with integrated application, judgement and the ability to respond effectively when conditions vary or become less familiar.
So the question is how far skills evidence allows us to go. A collection of individual skills may tell us a great deal about the resources a person has available. On its own, it does not establish that those resources can be brought together and applied effectively in real work.
That is where capability becomes necessary as a separate concept.
What capability adds beyond individual skills
Capability goes further than possessing a set of individual skills.
It involves being able to draw on relevant skills, knowledge and behavioural repertoires, combine them appropriately, and apply them effectively in the situations the work presents. That integration matters because real work rarely asks people to use one skill in isolation.
Judgement plays an important role in that process. A person may know what to do in principle and have the required skills, but still need to decide what matters in a particular situation, which knowledge applies, which response is appropriate, when to adapt, when to escalate and what trade-offs are acceptable. In that sense, judgement is better understood as part of the mechanism that brings the other elements together than as simply another item in a capability checklist.
Capability also implies more than a single successful demonstration. A person may perform well once under familiar or controlled conditions without being able to repeat that performance reliably or adapt when the situation changes. Broader capability therefore involves some degree of consistency and adaptability.
This is where context matters. The capability required for a role depends partly on the complexity, uncertainty, variation, autonomy and consequence involved in the work. Someone may be capable in a tightly defined situation but not yet ready to perform effectively where information is incomplete, conditions vary or greater judgement is required.
Capability literature in professional practice has long associated broader capability with judgement, adaptability and the ability to respond in unfamiliar or changing situations. Our working interpretation builds on that distinction: capability is not simply what someone knows or can demonstrate once, but what they can bring together and apply reliably and adaptively across the range of situations the work reasonably requires.
That range still has limits. Capability does not mean being able to perform successfully under every conceivable condition. It remains bounded by the demands of the role and the work itself.
Capability is not the same as performance
Capability, however, is still not the same as performance.
The context discussed above helps define what a person needs to be capable of handling. Performance is shaped by something broader: what happens when that capability meets the conditions of the work.
Those conditions include systems, tools, processes, workload, management, incentives, access to information, role clarity and the smaller forms of organisational friction that can make good performance easier or harder.
A slow system, unclear process, unrealistic workload or conflicting incentive sits outside the individual’s capability, even though it may have a significant effect on performance.
A capable person can still underperform in a poor environment. The reverse can also happen. Strong procedures, close supervision, automation or extensive support can sometimes allow someone with more limited capability to perform acceptably under favourable conditions.
Performance-improvement approaches have long treated performance as something shaped by both the performer and the wider system of work. COM-B makes a similar distinction by separating capability from opportunity and motivation. The frameworks differ in purpose, but both support the same practical boundary: what a person is able to do and the conditions that influence what they actually do should not be treated as the same thing.
Performance can therefore provide valuable evidence of capability without being identical to it. Repeated effective performance across different but relevant situations strengthens the case that capability is present because it shows reliability and adaptation.
Even then, the conditions still matter.
If we want to understand whether a performance problem is really a capability problem, we need to know not only what happened, but also the environment in which it happened.
The same performance problem can require a completely different intervention
This distinction becomes more useful when we apply it to an actual performance problem.
Consider two organisations with the same weak customer-service metric.
In the first, employees struggle to diagnose and resolve more complex customer issues. They may know the process and have some of the required skills, but they cannot yet apply those skills reliably when the situation becomes less familiar or demands more judgement.
That points to a genuine capability problem.
The intervention should therefore focus on developing what people are able to do. That may involve targeted learning, realistic practice, coaching, feedback, job aids, guided experience and opportunities to work through progressively more demanding situations. Another course may form part of the response, but capability development is broader than course completion.
Now consider the second organisation.
Its employees are already capable of handling complex customer issues, but the systems are slow, approval limits are too restrictive and customers are repeatedly transferred between teams. The employees know what to do and can do it, but the work environment prevents that capability from becoming effective performance.
The intervention looks very different.
The organisation may need to redesign processes, change decision rights, improve systems, simplify handoffs, revise escalation rules or address workload and management practices. Some learning may still be required to support those changes, but it is not the solution to the underlying problem.
The business symptom may look similar in both organisations, but the constraint may sit in capability, in the environment around the work, or in some combination of the two. Similar outcomes should not automatically trigger similar interventions.
Skills data can create visibility without proving workforce readiness
The same distinction matters when organisations look at skills data at scale.
A skills inventory can give leadership useful visibility. It may show, for example, that hundreds of employees are associated with skills such as data analysis, customer experience, project management or people leadership. That information can support workforce planning, development, hiring and succession decisions.
But it still has limits.
Knowing that people are associated with particular skills does not automatically tell us whether they can combine those skills with the knowledge and judgement the work requires. It does not tell us whether they can apply them reliably across the situations the strategy depends on, whether the work environment enables that capability to become performance, or whether that performance contributes to the outcome the organisation actually needs.
The question for leadership is therefore not only:
What skills do we have?
It is also:
What can our people actually do with them under the conditions our strategy requires?
Skills data can create a sense of workforce readiness that is stronger than the evidence supports. A well-constructed skills model can improve visibility considerably, but visibility is not proof of execution capability.
For workforce planning, development investment, hiring, succession and technology decisions, that difference matters. Skills data is valuable when it helps organisations ask better questions about the workforce. It becomes less useful when it is expected to answer questions about capability and performance that it was never designed to answer on its own.
Related does not mean linear
The distinctions developed through the article can be brought together in a broader interpretive lens:
Learning → Skills → Capability → Performance → Business Impact
This is a way of keeping the relationships already established visible without treating them as a linear pipeline.
Learning can contribute to skill development. Skills can contribute to capability. Capability can support performance. Performance can contribute to business outcomes. None of those relationships is guaranteed, and none of the stages automatically proves the next.
There are also feedback loops. Performance can expose capability gaps. Work itself can develop capability. Business outcomes can change what the organisation needs people to be capable of, which may in turn reshape learning priorities.
This kind of distinction is already familiar in L&D. Frameworks such as Kirkpatrick and LTEM separate learning from workplace application and downstream results. Their value here is adjacent rather than foundational.
The purpose of the lens is to keep these questions separate enough that organisations do not claim more from their evidence than it can support.
Conclusion
The value of these distinctions is what they allow the organisation to do differently.
A skills gap, a capability gap and a performance-environment problem are not interchangeable. If they are treated as though they are, the diagnosis can be wrong and the investment can follow in the wrong direction.
The practical requirement is straightforward: understand where the problem actually sits before deciding what to do about it.
Even when the problem is genuinely one of capability, the answer is not automatically another course. Developing capability may require learning, but also practice, coaching, feedback, support and experience in the conditions where the work needs to be done.
References
Fraser, S.W. and Greenhalgh, T. (2001). Coping with complexity: educating for capability. BMJ, 323(7316), 799–803. doi:10.1136/bmj.323.7316.799.
BMJ article
Lester, S. (2014). Professional standards, competence and capability. Higher Education, Skills and Work-Based Learning, 4(1), 31–43. doi:10.1108/HESWBL-04-2013-0005.
Emerald article
Michie, S., van Stralen, M.M. and West, R. (2011). The behaviour change wheel: A new method for characterising and designing behaviour change interventions. Implementation Science, 6, 42. doi:10.1186/1748-5908-6-42.
Implementation Science article
Rummler, G.A. and Brache, A.P. (2012). Improving Performance: How to Manage the White Space on the Organization Chart. 3rd ed. San Francisco: Jossey-Bass.
Publisher record at Wiley
Kirkpatrick Partners. The Kirkpatrick Model.
Official Kirkpatrick Model overview
Thalheimer, W. (2024). The Learning-Transfer Evaluation Model (LTEM), Version 13. Work-Learning Research.
Official LTEM page and current model
