A skills taxonomy is a classified list: it names skills and sorts them into categories and subcategories. A skills ontology goes further. It defines how skills relate to each other, to roles, to proficiency levels and to evidence, so software can reason about them. A taxonomy tells you where a skill lives. An ontology tells you what the skill connects to and what it makes possible.
That is the short answer. The longer one matters if you are an HR leader deciding what to build, buy or ask a vendor about. This guide explains both structures, shows a worked example, and says when a plain taxonomy is enough.
TL;DR
- A taxonomy is a hierarchy of skill names and categories. It gives your organization a shared vocabulary.
- An ontology adds typed relationships, definitions, proficiency, context and evidence. It lets systems match people to roles and learning.
- You need a taxonomy first. An ontology is only as good as the inventory of skills and proficiency data behind it.
- Leadership is a good test case: the word means different things for a supervisor, a nonprofit director and a veteran entering civilian management.
- Start small: pick a taxonomy, define levels and evidence, then add relationships where decisions depend on them.
Table of Contents
- What a skills taxonomy is
- What a skills ontology is
- Skills taxonomy vs ontology side by side
- A worked example: "leadership" in four settings
- Starting points you don't have to invent
- Why taxonomies stall without data
- How to choose: a decision guide
- Perspective: skills data is only useful if people practice
- Connecting the model to actual leadership learning
- Common mistakes to avoid
- FAQ
- Sources
What a skills taxonomy is
A taxonomy is a classification scheme. A development bank report on occupational and skills structures describes the purpose as giving a common language and set of standards for understanding and comparing skills. In its simplest form, it names things and puts them in groups.
The National Academies review of the US occupational database O*NET shows what this looks like in practice. Its prototype skills taxonomy groups skills under headings such as process skills (active listening, active learning, critical thinking) and cross-functional skills (problem solving and others).
A taxonomy typically gives you:
- A controlled vocabulary, so "people management" and "managing people" do not become two skills.
- A parent-child hierarchy, such as Leadership, then Coaching, then Giving feedback.
- A way to report: how many people are tagged to a category, and which categories have training attached.
What it does not give you is meaning beyond the tree. A taxonomy can say that feedback sits under communication. It cannot say that someone strong in feedback is close to ready for a team lead role, or that the skill matters differently in a classroom than on a warehouse floor.
What a skills ontology is
An ontology is a model of a domain: the things in it, the properties they have and the relationships between them. The University of Toronto's work on an ontology for skill and competency management frames the problem as understanding the skills and competencies of an organization's people well enough to use them. The point is use, not just labeling.
In a skills ontology, relationships carry meaning. Common ones include:
- Is part of / is a type of: the hierarchy a taxonomy already has.
- Requires: a role requires a skill at a given level.
- Is related to or often co-occurs with: two skills that tend to appear together.
- Is evidenced by: a certification, an assessment result, a project or an observed behavior.
- Leads to: a skill that opens a path to another role.
The taxonomy is still there, as the visible spine. Researchers working on ontology-based skills management make the same point: the hierarchy is only the part users see. A production ontology also needs identifiers, definitions, proficiency and context, evidence, links to roles and learning, and governance. Without that, matching breaks when employees and job descriptions use different wording.
Skills taxonomy vs ontology side by side
| Question | Taxonomy | Ontology |
|---|---|---|
| What is it? | A hierarchy of skill names | A network of skills, roles, levels and evidence |
| Relationships | Mostly parent and child | Many typed relationships |
| Handles synonyms and context | Weakly | Designed to |
| Supports matching and gap analysis | Limited, mostly by tag | Yes, if the data exists |
| Effort to build | Weeks to months | Ongoing, needs governance |
| Best for | Shared language, reporting, course tagging | Internal mobility, workforce planning, personalized learning |
| Main risk | Becomes a static labeling exercise | Becomes a diagram nobody feeds with data |
Most articles ranking for this topic draw the same line: a taxonomy is a list, an ontology is a living network. That is a fair summary. The practical question is how much of the network your decisions need.
A worked example: "leadership" in four settings
Take the single word "leadership". In a taxonomy it is a node, perhaps with children like delegation, coaching and decision making.
Now ask what it means for four people:
- A first-line supervisor in a distribution center. Leadership means running a shift, giving clear direction and handling conflict fast.
- A faculty member advising a student organization. It means facilitation and mentoring without authority.
- A nonprofit program director with a volunteer team. It means motivating people who are not paid and cannot be reassigned.
- A veteran moving into civilian management. The leadership experience is real, but the vocabulary, the authority structure and the expectations around influence differ.
The Toronto research makes this point directly: a useful model links a capability to the activity, the setting, the level and the observable evidence, rather than treating "communication" or "leadership" as one thing everywhere. A taxonomy flattens these four people into one tag. An ontology can keep them apart and still show what they share.
It also shows why leadership shouldn't be walled off from other skills. The World Economic Forum's Future of Jobs Report 2025 lists leadership and social influence among employers' core skills, next to analytical thinking, resilience, creative thinking and technological literacy. It also reports that, compared with 2023, leadership and social influence rose 22 percentage points in how often employers identify it as core. A model that connects leadership to those neighboring skills reflects how employers actually describe the need.
Starting points you don't have to invent
You rarely need to build a taxonomy from scratch. In the US, the O*NET content model from the Department of Labor's occupational information program is a solid baseline. It organizes information about jobs and workers into a structured system covering worker characteristics, skills, knowledge, work activities, work context and occupations. It even includes a leadership orientation characteristic.
The World Economic Forum's report classifies its survey results using its own global skills taxonomy, which is useful if you want to compare your skills with what employers say they need.
A reasonable approach is to borrow a public framework, then adapt the wording to your own roles. Public frameworks give you coverage. Your own edits give you fit.
Why taxonomies stall without data
The most common failure is not choosing the wrong structure. It is stopping at the structure.
A study of a skills management system for strategic workforce planning found that the payoff depends on an inventory: employee profiles, proficiency or evidence, the skills each role requires and a gap analysis. Without those, a taxonomy stays a labeling exercise instead of a workforce-development system.
Research on skills matching in talent management also describes firms shifting focus from jobs to skills and, in several cases, still being in the process of establishing a skills taxonomy. In other words, many organizations are early on this path, and that is normal.
Four questions tell you where you stand:
- Do we have one agreed list of skill names, with definitions?
- Do we know, for a meaningful share of employees, which skills they have and at what level?
- Do we know what our key roles require?
- Can we link a gap to something people can actually do, such as a lesson, a practice exercise or a stretch assignment?
If the answer to the first is no, build the taxonomy. If the first three are yes and the fourth is no, you are ready to add relationships and learning links, which is the ontology work.
How to choose: a decision guide
A taxonomy is enough when you mainly need consistent tagging, reporting by category, and a clean way to organize a course catalog. It is also the right first step for small teams.
Move toward an ontology when you want internal mobility recommendations, role-to-person matching, skill adjacency ("who is closest to this role?"), or personalization that adapts as people grow.
Be careful when a vendor promises an ontology but cannot show you the definitions, the proficiency scale, how evidence is captured or who governs changes. Ask to see the model, not only the interface.
A simple build path:
- Adopt or adapt a public taxonomy, such as O*NET, for your core roles.
- Define three to five proficiency levels with observable behaviors for each.
- Capture evidence: self-assessments, manager input, completed practice, project outcomes.
- Map required skills to your ten most important roles.
- Add relationships only where a decision will use them.
- Assign an owner for changes, and review the model on a set schedule.
Perspective: skills data is only useful if people practice
We build a leadership development platform, so we see this debate from the learner's side. A skills model, however elegant, does nothing until someone changes what they do on Monday morning.
Two things follow from that. First, leadership skills are behaviors, and behaviors are built by repetition. That is why we focus on short microlessons and practical exercises rather than long courses. Second, we think leadership development belongs to every employee, not only managers. A taxonomy that tags only people with "manager" in their title misses most of the people who lead projects, peers and volunteers.
So when we look at a skills framework, we ask: does it describe the person in context, and does it connect to something they can practice this week? A structure that answers yes to both is worth maintaining. One that answers no is a spreadsheet.
— Drew
Connecting the model to actual leadership learning
Once skills are defined, the next problem is learning that fits each person. Organizations that serve very different groups feel this most. A university, a nonprofit and a corporate HR team all need leadership content, but they define success differently.
This is where we put our work: Leaderly AI is a mobile-first platform that delivers personalized microlessons, behavioral habit exercises, self-assessments (including DiSC, EQ and MBTI), customizable learning journeys and analytics dashboards. HR teams and educators can customize content and scale it to everyone in the organization. If you are mapping leadership skills and want a way to turn gaps into daily practice, see Leaderly AI, or the pages for businesses, higher education and nonprofits.
If your skills model includes interpersonal capabilities, our articles on direct communication, the power of influence in leadership and people centricity show how we break those broad labels into behaviors people can practice.
For teams thinking about leadership alongside broader cognitive and technology skills, a digital leadership curriculum is one example of how learning content can combine those strands.
Common mistakes to avoid
- Building a huge taxonomy first. Thousands of skills nobody tags is worse than 150 well-defined ones.
- Treating the ontology as an IT project. Definitions and proficiency levels need HR and business owners, not just engineers.
- Ignoring wording. If employees say "coaching" and job posts say "mentoring", you need synonyms captured in the model.
- Leaving out evidence. A self-reported tag with no date or proof ages quickly.
- Skipping governance. Skills change. Without a named owner and a review rhythm, the model drifts.
- Isolating leadership. Employers describe a blend of cognitive, interpersonal and technology skills. Keep leadership in the same model.
FAQ
What is a skills ontology?
A skills ontology is a structured model that defines skills and the relationships between them and to roles, proficiency levels, learning and evidence. It includes a taxonomy as its hierarchy but adds meaning, such as which skills are related or which lead to a role. That lets software match people to roles and suggest development.
What is the difference between ontology and taxonomy?
A taxonomy classifies items into a hierarchy of categories. An ontology describes a domain with many kinds of relationships and attributes, not only parent and child. In HR terms, a taxonomy names and groups skills, while an ontology also connects them to people, roles, levels and evidence.
Do I need an ontology, or is a taxonomy enough?
If you need shared language, reporting and course tagging, a taxonomy is enough. If you want internal mobility, role matching or adaptive learning, you will need ontology features. Most organizations should build the taxonomy first and add relationships where decisions depend on them.
What are the 10 most important job skills?
The World Economic Forum's 2025 employer survey lists analytical thinking; resilience, flexibility and agility; leadership and social influence; creative thinking; motivation and self-awareness; technological literacy; empathy and active listening; curiosity and lifelong learning; talent management; and service orientation. Treat it as a global benchmark, then compare it with the roles in your own organization.
Sources
- The Future of Jobs Report 2025, Skills outlook (World Economic Forum)
- The ONET Content Model (ONET Resource Center)
- A Database for a Changing Economy: Review of O*NET (National Academies)
- An Ontology for Skill and Competency Management (University of Toronto)
- Ontology-Based Skills Management: Goals, Opportunities and Challenges
- Skills Management System as a Tool for Strategic Workforce Planning
- A Skills Taxonomy for LAC: Lessons Learned and a Roadmap (IDB)
- A skills-matching perspective on talent management (Human Resource Management)