2026-08-04 · 9 min read

What Are Skills? Types, Skill Sets, and How Organizations Map Them

Josh Friedman

Josh Friedman

Ask a plant manager what skills their maintenance crew has and you'll get a list of people. Ask the HR system and you'll get a list of job titles. Ask the LMS and you'll get a list of courses completed. None of those is a list of skills. This is the quiet problem underneath most workforce initiatives: everyone uses the word, and almost nobody in the building is working from the same definition.

That matters more than it sounds. Recruiting, training budgets, succession plans, and staffing decisions all rest on an implicit model of what people can do. When the model is fuzzy, every decision built on it is fuzzy too. So before anything else, it's worth getting precise about the base unit.

What Is a Skill?

A skill is a demonstrable ability to perform a specific task or activity to a defined standard. Three parts of that sentence carry the weight. Demonstrable means it can be observed or tested, not just claimed. Specific means it names an activity narrow enough to assess: "configure a VLAN on a Cisco switch," not "networking." Defined standard means there is a level of performance the skill can be measured against, which is what makes proficiency possible. A skill is distinct from knowledge (knowing how a VLAN works), from an ability (the underlying capacity to reason about networks), and from a competency (the bundle of skills, knowledge, and behaviors a role requires). Skills are acquired through practice, they can be assessed independently of the person's job title, and many of them decay without use. That last point is the one most systems ignore, and it is why "trained" and "currently skilled" are not the same thing.

Everything else in workforce planning is built on top of this unit. Get it right and the rest of the model becomes tractable.

Skills, Competencies, Knowledge, and Abilities

These four words get used interchangeably, and the confusion has real costs when they end up as columns in the same spreadsheet.

Knowledge is what someone understands. A nurse knows the pharmacology of anticoagulants. Knowledge is testable with questions and it does not by itself produce a result.

Ability is the underlying capacity to do something. Spatial reasoning, working memory, manual dexterity. Abilities are relatively stable, hard to train quickly, and mostly the domain of selection rather than development.

Skill is the applied, practiced execution. The nurse titrating a heparin drip correctly at the bedside is exercising a skill. It draws on knowledge and ability but is neither.

Competency is the bundle a role requires: the skills, the knowledge, and the observable behaviors that together produce acceptable performance in a job. "Medication management" is a competency. It contains a dozen skills.

The practical rule: skills are what you assess and develop. Competencies are what you attach to roles. If you build the framework the other way around, you end up assessing things too broad to score and developing things too vague to train. The full distinction, with examples, is in skills vs. competencies.

The Types of Skills

Taxonomies of skill types vary, but four distinctions do most of the work in practice.

Hard skills and technical skills

Hard skills are teachable, measurable, and usually specific to a domain or tool. Welding to a code standard. Writing SQL. Reading a P&ID diagram. Operating a forklift. Technical skills are the subset tied to a particular technology, method, or system. They are the easiest skills to define, the easiest to assess with a test or observation, and the ones most likely to have a formal certification behind them.

They are also the skills most likely to go stale. A technical skill on a system that changes every 18 months has an 18-month half-life whether the training record says so or not.

Soft skills and behavioral skills

Soft skills are the interpersonal and cognitive skills that shape how work gets done: communication, prioritization, conflict handling, coaching. The term undersells them. In most roles they determine whether the hard skills produce value. A brilliant engineer who cannot explain a root cause to a non-engineer produces less than a competent one who can.

Behavioral skills are the observable, situation-specific version of soft skills, and the version worth assessing. "Communication" is not assessable. "Delivers a clear shift handover covering open issues, risks, and owners in under five minutes" is. The move from soft skill to behavioral skill is mostly a matter of writing down what good looks like.

Transferable skills

Transferable skills move with the person across roles, functions, and industries. Structured problem solving. Data literacy. Project coordination. Stakeholder management. They are the skills that make internal mobility possible, and the ones most organizations fail to record because no single job description owns them.

An organization that tracks transferable skills can see that the customer support lead who ran the CRM migration has 60% of what the junior product manager role needs. One that tracks only job titles sees a support lead.

Durable skills and perishable skills

This is the distinction most frameworks miss entirely. Durable skills hold their value across years and technology cycles: analytical reasoning, writing, negotiation. Perishable skills degrade without practice or become obsolete as tools change: a specific software version, a procedure that gets revised, a regulatory process that changes annually.

The distinction changes how you manage them. Durable skills need to be developed once and refreshed rarely. Perishable skills need currency tracking, recertification triggers, and a decay model. Treating both the same way, as a checkbox that stays checked forever, is how a workforce ends up with a qualified-on-paper crew that cannot actually perform. There is a longer treatment in durable vs. perishable skill sets.

What Is a Skill Set?

A skill set is the combination of skills a person holds or a role requires, grouped so that it describes a coherent capability. Individuals have skill sets. Roles have required skill sets. The gap between the two is where every development plan starts.

A useful skill set is specific enough to assess and complete enough to describe the work. Three worked examples show what that looks like.

Data center technician

SkillTypeTypical required level (0–5)
Rack and cable installation to site standardHard / technical4
Power distribution fault isolationHard / technical3
Cooling system monitoring and alarm responseHard / technical3
Change ticket documentationBehavioral3
Escalation communication under incident conditionsBehavioral3
Vendor certification maintenance (e.g., CDCP-type credentials)PerishableCurrent

Half of this skill set is perishable. A technician who was excellent on the previous generation of power distribution units and has not touched the new ones is not a level 3 today, regardless of what a two-year-old record says.

Project manager

SkillTypeTypical required level (0–5)
Scope definition and change controlTransferable4
Schedule development and critical path analysisHard4
Risk identification and mitigation planningTransferable3
Stakeholder expectation managementBehavioral4
Budget tracking and variance reportingHard3
Facilitation of decision meetingsBehavioral3

Almost nothing here is perishable, and most of it is transferable. That is why project managers move across industries more easily than most roles, and why a skills-based system can spot a strong internal candidate in a function the job description never mentioned.

Registered nurse, acute care

SkillTypeTypical required level (0–5)
Patient assessment and early deterioration recognitionHard / clinical4
Medication administration and reconciliationHard / clinical4
Electronic health record documentationTechnical3
Handover communication (SBAR or equivalent)Behavioral4
De-escalation with patients and familiesBehavioral3
Unit-specific equipment operationPerishableCurrent

Clinical skill sets mix everything: durable clinical judgment, perishable equipment competence, behavioral communication, and formal licensure. This is why healthcare has invested more in competency architecture than most industries. The cost of a stale record is measured in patient outcomes.

How Skills Are Measured

Naming a skill is the first half. The second half is a scale. Without one, "has SQL" means anything from "wrote a SELECT statement once" to "tunes production query plans."

Proficiency levels solve this by attaching a defined standard to each level of a skill, usually on a 0–5 or similar scale, with a behavioral description of what each level looks like. Level 2 in root-cause analysis might read "identifies probable causes for routine failures with guidance." Level 4 might read "leads structured analysis on novel failures and validates corrective actions." The difference between the two is assessable, and the gap between where someone is and where the role needs them is a number rather than an impression.

The scale is what turns a list of skills into a skills matrix, a gap analysis, or a succession plan. There is a full guide to building one in proficiency levels: scales, examples, and how to define them.

How Organizations Map Skills

Individual skills, grouped into skill sets, measured on proficiency scales. The remaining question is how an organization holds all of it in one place. The answer is a skills library, sometimes called a skills taxonomy or skills framework.

What a skills library is

A skills library is the organization's controlled vocabulary of skills: each skill named once, defined once, given a scale once, and then referenced everywhere it matters. Roles point to skills in the library. Assessments score people against skills in the library. Learning content is tagged with the skills in the library it develops. Because there is one definition, a gap on the maintenance team and a gap on the operations team mean the same thing and can be compared.

Without a library, every manager writes their own version of "troubleshooting," every job description invents a new phrasing, and the organization ends up with 400 skills that are really 80.

Why fixed industry taxonomies fail

There is a tempting shortcut: license a pre-built taxonomy with 30,000 skills already defined and load it in. It looks impressive in a demo. It rarely survives contact with the organization.

The reason is that no two organizations define competence the same way, and the definitions that stick are written in the organization's own language, mapped to its own roles and standards. A generic "Electrical Troubleshooting" skill with a generic description does not tell a technician at a specific plant what level 3 means on their equipment against their procedures. So managers ignore it, assessments drift back to gut feel, and the taxonomy becomes shelfware.

The libraries that work are built from the organization's roles outward: start with the 20 roles that matter most, define the skills those roles actually require in the words the people doing the work would use, attach scales with behavioral anchors, and grow from there. External taxonomies are useful as a reference to borrow from, not as a foundation to adopt. SkillsDB is built around this approach, with a skills library designed to be authored in the customer's language rather than imported from a catalog.

A short checklist for a working skills map

  • Every skill has one canonical name and one definition.
  • Every skill has a proficiency scale with behavioral anchors at each level.
  • Every role references its required skills and levels from the library.
  • Perishable skills carry a currency rule: recertification interval, decay trigger, or evidence requirement.
  • Assessments, learning content, and job architecture all reference the same library, so a gap means the same thing everywhere.
  • Someone owns the library. Uncurated skills libraries decay as fast as perishable skills do.

The Base Unit

Org charts describe reporting lines. Job titles describe what the company decided to call the work. Headcount describes how many people are on payroll. None of them describe what the organization can actually do.

Skills do. They are the base unit that every other workforce structure is composed from. A role is a skill set with a title. A team is a set of skill sets with a manager. A succession plan is a prediction about which skill sets will be ready when. A training budget is a bet on which skills will move.

When the base unit is undefined, all of that runs on approximation. When it is defined, named, scaled, and held in one place, the approximations become measurements. That is the whole argument for treating skills as infrastructure rather than as a field in an HR record. Define the unit first. Everything else follows.

Frequently Asked Questions

Ready to make skills visible?

See how SkillsDB puts your workforce data to work.