competency-modeling

Context

Use in Phase 2 when an engagement requires building (or rebuilding) an organization's competency architecture. Owned by the Learning & Development Specialist, with structured hand-offs to the skills that consume the finished model.

Positioning vs. learning-curriculum-design: that skill includes a competency-model design step as the ENTRY to curriculum work — a just-in-time model for one learning program. THIS skill is the organizational framework: the enterprise competency dictionary, its proficiency ladder, and the assessment design that makes competencies measurable across talent processes (performance rubrics, succession readiness, curriculum gap analysis). When a full framework exists, learning-curriculum-design consumes it instead of re-deriving.

When to Use

• Client has no competency framework, or one that is stale, unlevelled, or unusable as a rating basis
• A competency model must serve MULTIPLE consumers (performance + learning + succession), not one program
• Behavioral anchors or proficiency levels need to be defined or harmonized
• Assessment instruments (interview guides, work samples, 360s) must be designed against the model

Execution Steps

1. Scope the Architecture

• Decide the model's span: core/values competencies (all roles), functional/technical competencies (job families), leadership competencies (management levels). Record which are in scope.
• Align to strategy: the 5-8 competencies that distinguish great performance in THIS organization's next 3 years — not a generic library dump. Cite the strategic input (business plan, org-design output from org-design-framework if present).

2. Collect Evidence (never invent from the armchair)

• Critical-incident interviews with high performers and their managers: "describe a time this role demanded excellence — what did the person DO?"
• SME panels / focus groups per job family to draft functional competencies
• Existing artifacts: job descriptions, performance forms, exit-interview themes
• Every competency must trace to at least two independent evidence sources; label single-sourced candidates as DRAFT.

3. Draft the Competency Dictionary

One entry per competency:

Field	Rule
Name	Short, observable behavior phrase ("Builds Stakeholder Trust" — not "Integrity")
Definition	1-2 sentences, behavior + outcome, no traits or personality language
Behavioral anchors	3-5 observable indicators per proficiency level — what a rater would SEE
Distinctness	No overlap with another competency — merge or split until clean
Evidence	The sources supporting the entry

Ban list: circular definitions ("X is demonstrating X"), unobservable adjectives ("passionate", "driven"), and trait nouns.

4. Level the Model (proficiency scale)

• Define ONE proficiency scale across the model (recommend 4 levels: Foundational → Applied → Advanced → Expert/Mastery), with a one-line anchor defining what progression means generically.
• For every competency × level: a concrete behavioral anchor. Test each anchor with "could two raters watching the same person agree this level is shown?" — if not, the anchor is too vague.
• Map levels to the client's career ladder where one exists (individual-contributor tiers vs. manager tiers); note where the model and ladder disagree rather than forcing alignment.

5. Validate

• SME review of definitions and anchors (clarity + completeness)
• Incumbent + manager review of LEVELING at 2-3 real roles: run a mini calibration — discrepancies are anchor defects, not rater errors
• Record validation status per competency (Validated / Draft)

6. Design the Assessment Layer

Choose methods per competency type:

Competency type	Primary assessment methods
Behavioral (core, leadership)	Structured behavioral interviews (STAR probes against anchors), 360 feedback
Functional/technical	Work samples, simulations, credential verification
Integrated (complex judgment)	Case/simulation exercises scored with the anchor rubric

Design rules:

• Every instrument's scoring rubric quotes the model's anchors verbatim — paraphrased rubrics break comparability
• Structured format: same probes, same order, same scoring guide for all candidates/incumbents
• Multiple assessors for high-stakes decisions; document inter-rater resolution
• Fairness guardrail: check for adverse impact across protected groups during validation runs — route measurement questions to data-analyst; labor-market legality questions (which characteristics may lawfully be assessed under the active country profile) to labor-compliance-analyst
• Assessment ≠ performance rating: assessments measure current proficiency against anchors; performance management measures results against goals — keep the instruments distinct

7. Govern and Hand Off

• Version the dictionary (semver); record review cadence (annual or on org-design change)
• Hand-offs: behavioral anchors + levels → performance-system-design (rating rubrics); gap analysis → learning-curriculum-design (curriculum priorities); readiness levels → career-path-succession-planning (succession criteria); "% roles with defined competency profile" → hr-metrics-analysis per the ISO 30414 mapping (Professional competence and development area)

Output Format

• Competency Dictionary: entries per the Step-3 table, with validation status
• Proficiency Leveling Matrix: competency × level anchor grid
• Assessment Plan: per competency — method(s), instrument sketch, scoring source (which anchors), assessor structure
• Governance Note: version, review cadence, evidence log locations

Related Skills

• learning-curriculum-design — downstream consumer: converts assessed gaps into curriculum design
• performance-system-design — downstream consumer: anchors become rating rubrics
• career-path-succession-planning — downstream consumer: levels become readiness criteria
• org-design-framework — upstream: strategy and job-family structure input