What Is Competency? (From Dictionary Definition to Organizational Evidence) A Clear Framework for L&D

The word competency can appear on a company intranet as “expected behavior,” and on that same company’s performance form as a “scored criterion.” Both can seem correct, but they are not the same thing. This confusion quietly makes L&D’s job harder: producing training gets easier, producing evidence gets harder.

I’m going to take competency out of the dictionary and bring it down to the field: role → behavioral indicator → evidence. Because if there are no behavioral indicators and no evidence, a competency model resembles Borges’ infinite library: shelves are full, but no one knows whether the book you’re looking for truly exists (Borges, “The Library of Babel”, 1941).

Also, a small human observation: the same manager can say “We want competency-based development” in a meeting and five minutes later say “Everyone should take this training.” This contradiction doesn’t make me angry; I just haven’t fully modeled it yet. The sentence “everyone should take it” is often a security blanket replacing a lack of evidence.

1) A dictionary definition isn’t enough: What exactly is competency?

In everyday language, “competent” means “can do the job.” In corporate language, competency usually tries to carry three things at once:

• The knowledge required by the job
• The skill that turns that knowledge into practice
• The attitude / values / habits that make that practice consistent

But when I reduce competency to a single sentence, I use this:

Competency = A performance capacity that emerges through observable behaviors in a specific role and context, and can be verified with evidence.

Two words are critical here: observable and evidence. You can’t write behavioral indicators for something you can’t observe; and you can’t manage what you can’t prove. Wittgenstein’s idea that “meaning is use” helps here: the meaning of a word is determined by how you use it in practice (Wittgenstein, Philosophical Investigations, 1953). Which decision does your organization use the word “competency” to support—assignment, promotion, training, pay? When the answer changes, the model changes too.

2) Competency, skill, knowledge, attitude: The four most confused concepts in L&D

Creating a “competency map” without separating these four concepts is like setting out on a journey without marking north on the map. It works for a while, then everyone heads in a different direction.

I use the table below often when working with teams:

Let me add a correction: saying “attitude can’t be measured with a survey” would be too harsh. It can be measured; but it is not evidence on its own. A survey traces intention and perception; it doesn’t prove behavior.

From an L&D perspective, the practical result is this: your training catalog may be full of “knowledge modules,” but if the organization wants “competency development,” this is exactly where the catalog alone falls short.

3) No behavioral indicators, no model: The 5 most common mistakes

The most common mistakes I see when writing competency models (and yes, I fell into these early on too):

1. Settling for abstract words: “Strong communicator,” “proactive,” “strategic.”

   • Question: What do they do that makes them proactive? What do they not do when they’re not?

2. Describing personality instead of behavior: “Born leader,” “positive.”

   • A personality judgment locks development design.

3. Not writing the context: “Customer-focused.”

   • Which customer—internal or external? Which channel? Which risk?

4. Leaving level differences vague: If junior and senior are measured with the same sentence, scores become political.

5. Not defining the evidence mechanism: People say “we’ll measure it,” but don’t write how.

Writing behavioral indicators is like the “show, don’t tell” rule in literature. Le Guin’s writing notes have this discipline: instead of issuing judgments to the reader, you build a scene (Le Guin, Steering the Craft, 1998). In competency work too, instead of saying “communicates well,” you say “when an objection comes, asks 2 clear questions and reframes the proposal.” You build a scene.

4) The evidence layer: What makes competency organizationally “real”

Competency becomes real with evidence. And by evidence, I don’t mean only a test score; I mean an evidence pack appropriate to the nature of the role.

I group evidence into three baskets:

• Mini assessments: short quizzes, checkpoints, scenario questions
• On-the-job evidence: real work output, observation, sample case resolution
• Simulation: decision-based scenarios, branching flows, role-play recordings

At this point, my job gets easier at Nextrain: with AI Gates, I can embed “advance if successful / repeat if not” logic into a learning flow. So evidence isn’t collected at the end of the journey; it’s collected inside the journey, at decision points. I can also produce real-time tests and checkpoints, decision-based simulations, and interactive branching scenarios to make evidence more “behavior-adjacent.”

And there’s Saadet; when customers say “Let’s go competency-based,” the first sentence that usually follows is: “Okay, but how will we show this in an audit?” Saadet’s job is to remind people that this question is not panic—it’s a design question. (Sometimes the sentence “An HSE audit is coming up” is the moment organizations take competency seriously for the first time. An interesting source of motivation.)

Evidence pack template (copy–apply)

The template below makes a competency “measurable”:

Competency name:
Role(s):
Context:
Behavioral indicators (3-7 items):
Levels (L1-L4):
Evidence types:
- Mini assessment:
- Simulation:
- On-the-job evidence:
Validity period / renewal (if any):
Risk/compliance note (GDPR, HSE, etc.):

Don’t leave the “risk/compliance” line blank. In areas like GDPR and HSE, competency produces not only development but also compliance evidence. Organizations usually realize this late.

5) A rubric for leveling (junior–mid–senior): An example

Leveling can look like the most political part of a competency model; in fact, it’s the most technical part. Because a level difference doesn’t mean “better”; it means performance in a more complex context, with less support, more consistently.

Let me pick an example competency: “Managing customer objections” (sales context). Rubric:

The beauty of the rubric is this: L&D turns the question “which training?” into “which behavior is missing?” And it selects evidence according to the level.

I can’t skip mentioning Ebbinghaus: the forgetting curve tells us that one-time exposure decays quickly (Ebbinghaus, 1885). A leveling rubric determines what repetition should be tied to: instead of making people rewatch the same content, you make them repeat the same behavior in different contexts.

6) Linking the training catalog to competency: Map → journey

A common scene in organizations: there’s a catalog, lots of training, but the answer to “which training develops which competency?” doesn’t exist—even in Excel.

The linkage logic I recommend has three layers:

1. Competency map: Role → competency set → levels
2. Learning journey: Next step based on the competency gap
3. Evidence loop: Checkpoint + simulation + on-the-job evidence within the journey

In Nextrain, I can design this in practice like this:

• Making the “current level / target level” gap visible with the Competency Scanner
• Dynamically building Learning Journeys based on that gap
• Embedding AI Rules and AI Gates into the journey to change the flow based on evidence outcomes
• Tracking every action at event level (View / Click / Answer / Time) and moving reporting from “did they complete?” to “where did they struggle?”

There’s a risk here: if you turn the competency map into a “giant schema that covers everything,” no one uses it. In Calvino’s Invisible Cities, cities are sometimes bigger than the map; the map is no longer the city (Calvino, Invisible Cities, 1972). A competency map can bloat the same way. I start small: 5 critical roles, 6–10 competencies per role, 3–5 behavioral indicators per competency. Then expand.

Practical checklist: To be able to say “we’re competency-based”

• Is the context of each competency written?
• Are there at least 3 behavioral indicators?
• Are at least 2 different evidence types defined?
• Is the level difference separated not as “better,” but as a more complex context?
• Is every critical training in the catalog linked to at least 1 competency?
• In areas like GDPR / HSE, have renewal and audit evidence been considered?

7) A short example scenario: What does competency look like in areas like GDPR / HSE?

GDPR and HSE trainings are often labeled “mandatory” and the matter is considered closed. I find that incomplete. Mandatory training is only the exposure part of competency; the evidence part still requires design.

Example: the competency “Correctly escalating a suspected data breach under GDPR.”

• Behavioral indicator: Reports the suspicious incident to the correct channel within 24 hours; does not share unnecessary personal data about the incident.
• Mini assessment: A 5-question scenario checkpoint (which information is shared / which is not?)
• Simulation: Steps to take in the first 10 minutes in a case of “email sent to the wrong person”
• On-the-job evidence: Filling out the correct form based on a real (anonymized) incident record

This design is stronger in an audit than saying “they took the training”: you say “they demonstrated this behavior with these pieces of evidence.” Organizations usually relax when they hear the second sentence; that relief is almost tangible.

“Completion” is activity evidence; competency requires behavioral evidence.

Closing: Saving the word “competency”

Competency is one of the most used and least clarified words in organizations. My job isn’t to decorate the word; it’s to make it decision-ready. Write the role, show the behavior, collect the evidence. Then the training catalog starts to make sense on its own: the question “what should we produce?” connects to the question “which behavior is missing?”

If you’ll do just one thing today, pick one competency and write these three lines:

1. In which role, in which context is this competency needed?
2. What are the observable behavioral indicators?
3. How will I collect evidence of this behavior?

Everything else—maps, matrices, catalogs—comes later. If evidence doesn’t come, nothing comes.

Notes

1. Borges, Jorge Luis. “The Library of Babel” (La Biblioteca de Babel), 1941.
2. Wittgenstein, Ludwig. Philosophical Investigations, 1953.
3. Le Guin, Ursula K. Steering the Craft, 1998.
4. Ebbinghaus, Hermann. Über das Gedächtnis (Memory studies), 1885.
5. Calvino, Italo. Invisible Cities (Le città invisibili), 1972.