Job Analysis, Job Architecture and Job Evaluation

Introduction and context

Compensation decisions are only as reliable as the understanding of the work on which they are based. An organisation cannot design equitable pay structures, career paths or meaningful incentive systems if it has not established what jobs involve, how jobs relate to one another and how their relative contribution should be assessed. Job analysis, job architecture and job evaluation therefore form the analytical foundation of total rewards management.

These activities are related but distinct. Job analysis identifies the tasks, responsibilities, outputs, relationships and requirements of work. Job architecture organises jobs into a coherent structure of families, sub-families, levels and career paths. Job evaluation determines the relative value or contribution of jobs for internal equity purposes. Market pricing then supplements, but does not automatically replace, internal analysis.

In contemporary organisations, the object of analysis is increasingly fluid. Agile teams, project-based work, hybrid employment, digital tools and skills-based deployment challenge traditional assumptions that a job is a fixed bundle of tasks performed in one position. The central challenge is to create documentation and pay structures that are sufficiently stable for governance while remaining adaptable to changing work.

Learning objectives

By the end of this chapter, the reader should be able to:

1. Explain the purpose, scope and methods of job analysis.
2. Produce clear and defensible job documentation.
3. Design job families, sub-families, career levels and competency or skills frameworks.
4. Compare ranking, classification, point-factor and market-based evaluation methods.
5. Apply a point-factor method to a practical example.
6. Diagnose problems of job comparability, bias and outdated job design.
7. Recommend governance practices for equitable and future-ready job architecture.

1. Job analysis: meaning, purposes and outputs

Job analysis is a systematic process of collecting, validating and interpreting information about work. It examines what is done, why it is done, how it is done, the context in which it is done and the capabilities required for effective performance. It is concerned primarily with the job or work role, not with the personal characteristics of the current jobholder.

A useful distinction is between:

• Task: a discrete unit of activity, such as preparing a variance report.
• Duty: a cluster of related tasks, such as financial reporting.
• Responsibility: an area for which the role is answerable, such as budget control.
• Accountability: the obligation to achieve outcomes and explain decisions.
• Competency: an observable combination of knowledge, skill, ability and behaviour.
• Job specification: the requirements needed to perform the work.
• Job description: a structured statement of the work, purpose, accountabilities and requirements.

Purposes of job analysis

Job analysis supports several HR decisions:

• recruitment and selection criteria;
• onboarding and training needs;
• performance objectives and standards;
• work redesign and workforce planning;
• reasonable accommodation and occupational safety;
• internal equity and job evaluation;
• career development and succession planning;
• compliance with employment and equal-pay obligations.

A job analysis should not become an inflated list of every activity a person might perform. Excessively detailed descriptions become obsolete quickly and may unintentionally narrow managerial flexibility. Conversely, descriptions that use vague language such as “performs other duties” without identifying the role’s core purpose provide little analytical value.

Sources and methods of data collection

No single method is universally superior. A robust analysis normally triangulates several sources.

Interviews should distinguish between activities, outputs and decision rights. For example, “attends meetings” is an activity; “coordinates cross-functional decisions” is a responsibility; and “reduces service delays” is an outcome. The latter two are normally more valuable for job architecture.

A practical job-analysis protocol

A defensible process includes the following stages:

1. Define the purpose and scope. Clarify whether the analysis supports evaluation, redesign, recruitment or another decision.
2. Identify stakeholders. Include jobholders, supervisors, internal customers, HR, employee representatives where relevant and subject-matter experts.
3. Collect evidence. Use interviews, questionnaires, observation, records and workflow information.
4. Separate job content from jobholder characteristics. Record required capability, not the incumbent’s exceptional personal attributes.
5. Validate the draft. Ask knowledgeable stakeholders whether the document reflects normal, substantial and continuing work.
6. Evaluate and approve. Apply the organisation’s evaluation method and governance rules.
7. Review periodically. Trigger review after major technology, process, regulatory or organisational changes.

2. Job documentation

A modern job description should be concise enough to remain usable but sufficiently specific to support decisions. A recommended structure is:

Accountabilities should use outcome-oriented language. “Maintains the enterprise data governance framework” is stronger than “responsible for data governance.” Requirements should be genuinely necessary. Requiring a particular degree when equivalent expertise is sufficient can create unnecessary exclusion and may weaken the validity of selection decisions.

Job documentation must also distinguish between essential and desirable requirements. This distinction is important for fair selection, disability accommodation and skills-based hiring. A qualification should be labelled essential only when the work or a legal/regulatory condition requires it.

3. Job architecture

Job architecture is the organising framework that connects work, career movement and reward. It provides a common language for comparing roles without assuming that all work follows a single hierarchy.

A typical architecture contains the following elements:

• Job family: a broad group of roles with related professional purpose, such as Finance, Technology or Human Resources.
• Sub-family: a more specialised grouping, such as Financial Planning within Finance.
• Job profile or discipline: a recognisable type of work, such as Compensation Analyst.
• Career stream: a route of progression, commonly managerial, professional, technical, operational or sales.
• Career level: a stage defined by scope, complexity, accountability, influence and capability rather than title alone.
• Job code: a controlled identifier used in HR information and payroll systems.

Designing career levels

Levels should describe increasing contribution, not merely years of service. Relevant dimensions include:

1. Problem complexity: from established procedures to ambiguous, enterprise-wide problems.
2. Autonomy: from close supervision to independent judgement and strategic discretion.
3. Impact: from local output to functional, business-unit or enterprise consequences.
4. Influence: from individual contribution to persuasion across boundaries and external representation.
5. Accountability: from assigned activities to outcomes, resources, risk and long-term direction.
6. Knowledge and capability: from foundational application to recognised expertise and thought leadership.

A level descriptor should be broad enough to apply across a family but specific enough to distinguish adjacent levels. Titles should not be the primary evidence of level: organisations frequently use the same title for materially different work or different titles for comparable work.

Career streams and dual ladders

A dual-ladder architecture permits progression through management or through specialist expertise. This reduces the pressure to promote excellent specialists into people-management roles for which they may not be suited. The two streams should be comparable in contribution, although their work may be evaluated through different forms of influence and accountability.

Competency and skills frameworks

A competency framework describes behaviours and capabilities associated with effective performance, often including collaboration, judgement, leadership and customer orientation. A skills framework describes more specific and assessable capabilities, such as statistical modelling, labour-law interpretation or cloud-security configuration.

A useful framework distinguishes:

• foundational capabilities expected across the organisation;
• functional capabilities specific to a discipline;
• technical skills associated with tasks or methods;
• leadership capabilities;
• proficiency levels, such as awareness, working, advanced and expert.

Frameworks should be evidence-based. Proficiency descriptors need observable indicators. “Advanced communication” is ambiguous; “translates complex workforce analytics into decisions for non-specialist executives” is more assessable. Skills should not be treated as permanent labels: proficiency may decay, tools may change and new combinations of skills may become valuable.

4. Job evaluation and internal equity

Job evaluation is a systematic process for determining the relative value of jobs within an organisation. It does not evaluate the performance, personality or market worth of an individual employee. Its purpose is to support a rational ordering or grouping of jobs, which can then inform grades, pay ranges and career structures.

Internal equity means that jobs of substantially comparable value are treated consistently, while differences in reward are explainable by relevant factors such as job value, market conditions, performance or scarce skills. Internal equity is not identical to equal pay for identical work; it concerns the defensibility of comparisons among different jobs.

Ranking method

The ranking method orders whole jobs from greatest to least value. It is inexpensive and understandable, making it useful in small organisations with relatively few jobs. However, it provides limited explanation of why one job outranks another and becomes unreliable when jobs are numerous, diverse or changing.

A paired-comparison variant compares jobs two at a time. If there are n jobs, the number of unique pairwise comparisons is:

Here n is the number of jobs. The expression applies when each pair is compared once and order does not matter. For 10 jobs, there are 10 × 9 ÷ 2 = 45 comparisons, illustrating why the approach becomes cumbersome at scale.

Classification or grading method

Classification places jobs into predefined grades using written grade definitions. A job is matched to the grade whose description best fits its overall scope and complexity. The method is practical and easy to communicate, but borderline cases may be difficult, and evaluators may anchor on titles or existing pay.

Point-factor method

The point-factor method evaluates jobs against compensable factors, assigns points to factor degrees and totals the points. Common factors include knowledge, problem solving, accountability, communication, working conditions and people leadership.

A generic structure is:

P is the total job score; wᵢ is the maximum point weight for factor i; dᵢ is the proportion of that factor’s maximum level achieved by the job, usually between 0 and 1; and k is the number of factors. This proportional form assumes factor scores are represented on a common 0-to-1 scale. An equivalent implementation assigns a direct point value for each degree and sums those values.

The method requires clear factor definitions, non-overlapping dimensions, anchored degree descriptions and trained evaluators. Weights are normative judgements about what the organisation values; they are not discovered facts. A factor should be included only if it is relevant to work value and can be evaluated consistently.

Worked point-factor example

Consider three factors for illustrative purposes:

The total is calculated as:

The points have no natural physical unit; they are ordinal or quasi-interval decision units created by the evaluation scheme. Therefore, a score of 1,010 should not automatically be interpreted as exactly twice the job value of a score of 505. The score supports structured comparison, not a scientific measurement of human worth.

Market-based evaluation and market pricing

Market pricing uses external pay data to estimate the rate associated with comparable jobs. It is particularly relevant where labour markets are competitive or skills are scarce. A market match should compare job content, scope, geography, industry, organisation size and data quality—not title alone.

Market data may be expressed as a percentile, such as the median or 75th percentile. A percentile identifies the position of an observation within a reference distribution; it does not guarantee that a particular survey is representative of the employer’s relevant labour market. Survey ageing, geographic differences, incentives, sample composition and job-matching error must be considered.

Market pricing can produce a coherent external structure but may weaken internal equity if used mechanically. A balanced approach commonly uses job evaluation to establish internal relationships and market data to inform pay-range positioning. The two signals may legitimately diverge, for example when a role has scarce digital skills or when an organisation pursues a lead-market strategy.

5. Comparing and integrating evaluation methods

A hybrid process should not average incompatible numbers mechanically. Instead, it should ask: What does the internal evaluation say about relative contribution? What does reliable market evidence say about external opportunity cost? What explains the difference? The final decision should be documented, especially where a market premium or retention adjustment is applied.

6. Job comparability, bias and governance

Comparability is the foundation of fair evaluation. Jobs are comparable when the same evaluation logic can be applied to them, even if their tasks differ. Comparability does not require identical work. It requires meaningful comparison of scope, complexity, accountability, knowledge and working conditions.

Common threats include:

• Title bias, in which prestigious titles receive excessive weight.
• Incumbent bias, in which an exceptional jobholder causes the job to be evaluated too highly.
• Gender or occupational segregation, in which caring, administrative or relational work is undervalued.
• Technology bias, in which visible technical complexity is valued more than less visible judgement and coordination.
• Proximity bias, in which work performed near senior leaders is assumed to have greater impact.
• Factor contamination, in which a factor includes elements already measured elsewhere.
• Factor omission, in which important work such as emotional labour, risk prevention or stakeholder influence is ignored.

Governance should include a documented methodology, trained panels, calibration sessions, conflict-of-interest rules, version control, appeal or review routes and periodic audits. Evaluation panels should include diverse perspectives and should examine whether factors systematically disadvantage particular occupational groups. Documentation should show the evidence supporting the level, factor degrees and any market adjustment.

Legal requirements vary by jurisdiction, but organisations should generally preserve records sufficient to explain job classifications and pay decisions, apply equal-pay principles consistently, protect personal data and consult employee representatives where required. Job evaluation is not a substitute for legal advice or a guarantee of compliance.

7. Changing work design and technology

Traditional job architecture assumes relatively stable jobs. Contemporary work is often organised around products, projects, platforms and skills. Artificial intelligence may automate some tasks while increasing the importance of judgement, problem framing, data stewardship and human accountability. Consequently, job analysis should examine task bundles rather than assume that a job title captures the whole role.

Useful responses include:

• maintaining core accountabilities while allowing task flexibility;
• recording skills separately from jobs in a skills inventory;
• using project or gig assignments without confusing temporary tasks with permanent job value;
• reviewing roles after major automation or process redesign;
• evaluating human accountability even when decisions are supported by algorithms;
• distinguishing a skill premium from a permanent revaluation of the whole job.

AI can assist by clustering job descriptions, identifying duplicated skills, detecting missing fields and suggesting possible comparators. It should not independently determine job value. Automated text analysis can reproduce historical bias, infer sensitive characteristics or mistake wording similarity for substantive comparability. Human review, traceability, data minimisation and validation against actual work are essential.

8. Applied case: Northstar Health Services

This is an illustrative case.

Background

Northstar Health Services expanded from three hospitals to a regional network. Its pay structure contained 42 grades, inconsistent titles and separate job descriptions created by local HR teams. A “senior analyst” in one hospital performed routine reporting, while another led enterprise workforce analytics. Employees alleged that administrative and patient-coordination roles were undervalued compared with technical roles.

Problem

The organisation needed a defensible architecture for 1,200 jobs while preserving career opportunities for clinical, administrative, technical and managerial employees. Market data were available for some roles but weak for several public-service and specialist occupations.

Analysis

A project team first reviewed documents and conducted interviews and workshops. It identified five broad streams: clinical, corporate professional, technology, operations and people leadership. Within each stream, jobs were grouped into families and assessed against common dimensions: knowledge, problem solving, accountability, influence and working conditions. The team found that existing titles and salaries had influenced earlier evaluations, creating circular reasoning.

The evaluation panel therefore anonymised job titles during initial scoring, used benchmark jobs for calibration and separated job value from incumbent performance. It also examined whether coordination, emotional labour, regulatory risk and patient consequences were adequately represented.

Approach and solution

Northstar introduced eight broad career levels with stream-specific descriptors. A professional and a managerial path were made parallel from level four upward. Point-factor evaluation established internal relationships, while market data informed range positioning for roles with reliable external matches. Where market premiums were needed for scarce digital skills, the premium was documented separately from the underlying job score.

Job descriptions were rewritten around purpose, accountabilities, scope, decision rights and essential capabilities. A review trigger was established for major technology, regulatory or service changes, with a formal review at least every two years.

Outcome

The new architecture reduced duplicate titles, clarified progression and identified several compression problems between adjacent grades. Some pay relationships changed; others did not, because the review showed that perceived unfairness arose from inconsistent titles rather than job value. Northstar created an appeals process and published the evaluation principles, although individual pay data remained confidential.

Learning points

1. Architecture must be built from work evidence rather than inherited titles.
2. Internal evaluation and market pricing answer different questions.
3. Parallel career streams support expert careers without forcing management promotion.
4. Governance and communication are as important as the scoring model.
5. Review triggers are essential because work and skills change continuously.

Best-practice checklist

Before approving a job architecture or evaluation exercise, ask:

• Is the purpose of the exercise explicit?
• Are jobs analysed rather than individual personalities?
• Are accountabilities and outcomes distinguished from activities?
• Are career levels defined by scope and complexity?
• Are factors relevant, non-overlapping and evidence-based?
• Are diverse forms of work represented fairly?
• Are market matches based on content and labour market, not title alone?
• Are conflicts, appeals, records and review dates governed?
• Has the design been tested against future work and technology?

Chapter Summary

Job analysis establishes what work entails, why it exists and what capabilities it requires. High-quality job documentation focuses on purpose, accountabilities, scope, decision rights and essential requirements. Job architecture organises work into families, sub-families, streams and levels, creating a common language for career movement and reward. Competency frameworks emphasise observable behaviours and capabilities, while skills frameworks provide more specific and assessable descriptions of expertise.

Job evaluation determines relative internal value, not employee performance or personal worth. Ranking and classification are simple and useful in limited settings; point-factor methods offer greater structure and auditability but depend on sound factor design; market pricing supplies external evidence but is vulnerable to matching and survey-quality problems. A hybrid approach is often most defensible.

Fair evaluation requires attention to comparability, bias, documentation, employee voice and legal governance. Because work is changing through automation, agile organisation and skills-based deployment, architectures should preserve stable principles while allowing roles and skills to evolve. Human judgement remains necessary when technology supports, but cannot replace, evaluation decisions.

Key Terms

Job analysis: Systematic collection and interpretation of information about work.

Job description: Document stating a role’s purpose, accountabilities, scope and requirements.

Job specification: Statement of knowledge, skills, abilities and other requirements needed for work.

Job architecture: Organising framework of families, levels, streams and career relationships.

Job family: Group of jobs with related professional purpose.

Career stream: Route of progression, such as professional, technical or managerial.

Career level: Stage defined by scope, complexity, influence and accountability.

Competency: Observable combination of knowledge, skill, ability and behaviour.

Skills framework: Structured description of specific capabilities and proficiency levels.

Job evaluation: Systematic determination of the relative value of jobs.

Internal equity: Consistency of reward relationships among comparable jobs within an organisation.

Ranking: Ordering whole jobs from higher to lower relative value.

Classification: Matching jobs to predefined grade descriptions.

Point-factor method: Evaluation method that assigns weighted points to compensable factors.

Compensable factor: Job characteristic used to assess relative value.

Market pricing: Use of external pay evidence to inform job or pay decisions.

Benchmark job: Job used as a reference point in evaluation or market matching.

Factor contamination: Inclusion of unrelated or duplicated content in an evaluation factor.

Factor omission: Failure to represent a relevant dimension of work in evaluation.

Skills premium: Additional reward associated with scarce or strategically important capability.

Review Questions

Short Answer Questions

1. Distinguish job analysis from job evaluation.
2. What is the purpose of a job family?
3. Identify four dimensions that can distinguish career levels.
4. State one strength and one limitation of market pricing.
5. Why should job descriptions distinguish essential from desirable requirements?

Long Answer Questions

1. Explain how an organisation can design a job architecture that supports both managerial and specialist career paths.
2. Compare ranking, classification, point-factor and market-based evaluation methods.
3. Discuss the principal sources of bias in job evaluation and propose governance controls.
4. Explain how competency frameworks and skills frameworks complement job architecture.
5. Analyse why internal equity and external competitiveness may produce different pay signals.

Analytical / Application Questions

1. An organisation uses the same title for two roles, but one controls a large budget and the other performs standardised reporting. Design an evidence-based process to determine whether they belong at the same career level.
2. Using the point-factor formula in this chapter, calculate the score for a job with maximum factors of 400, 300 and 300 and degrees of 0.75, 0.50 and 0.40. Explain what the score can and cannot establish.
3. A new AI system removes routine tasks from a compensation analyst’s role but increases responsibility for data governance and executive interpretation. Explain how the job description, skills framework and evaluation should be reviewed.

Multiple Choice Questions with Answers

Question 1

Which statement best distinguishes job analysis from job evaluation?

A. Job analysis determines market salary, while job evaluation measures employee performance.

B. Job analysis examines work content and requirements, while job evaluation assesses relative job value.

C. Job analysis ranks employees, while job evaluation identifies training needs.

D. Job analysis applies only to managerial jobs, while job evaluation applies only to operational jobs.

Correct Answer: B. Job analysis examines work content and requirements, while job evaluation assesses relative job value.

Question 2

Which item is most appropriately expressed as a job accountability?

A. Attends weekly meetings

B. Uses spreadsheet software

C. Maintains the enterprise data governance framework

D. Has five years of experience

Correct Answer: C. Maintains the enterprise data governance framework

Question 3

The number of unique pairwise comparisons for 10 jobs is:

A. 20

B. 45

C. 90

D. 100

Correct Answer: B. 45

Question 4

In a point-factor method, the factor weights primarily represent:

A. The employee’s annual performance rating

B. The organisation’s judgement about the relative importance of job dimensions

C. The exact market salary for each job

D. The number of employees in each job family

Correct Answer: B. The organisation’s judgement about the relative importance of job dimensions

Question 5

What is a principal limitation of ranking jobs?

A. It cannot order jobs at all

B. It requires external salary surveys for every job

C. It provides limited explanation and becomes difficult at scale

D. It evaluates only technical skills

Correct Answer: C. It provides limited explanation and becomes difficult at scale

Question 6

A dual career ladder is designed primarily to:

A. Eliminate specialist roles

B. Allow progression through specialist and managerial paths

C. Set identical pay for every job

D. Replace job descriptions with employee biographies

Correct Answer: B. Allow progression through specialist and managerial paths

Question 7

If maximum factor points are 400, 300 and 300, and factor degrees are 0.75, 0.50 and 0.40, the total point score is:

A. 450

B. 500

C. 600

D. 750

Correct Answer: B. 500

Question 8

Which practice improves the validity of market pricing?

A. Matching jobs by title only

B. Ignoring geography and organisation size

C. Comparing job content, scope and relevant labour market

D. Using the highest available survey value for every job

Correct Answer: C. Comparing job content, scope and relevant labour market

Question 9

Which example illustrates incumbent bias?

A. A job is reviewed after a major technology change

B. A role is rated highly because its current jobholder has exceptional personal expertise

C. A panel uses several data-collection methods

D. A job is assigned a code in the HR system

Correct Answer: B. A role is rated highly because its current jobholder has exceptional personal expertise

Question 10

Why should AI not independently determine job value?

A. AI cannot process text

B. Automated analysis may reproduce bias and confuse wording similarity with substantive comparability

C. AI is useful only for payroll calculations

D. Job evaluation never requires evidence

Correct Answer: B. Automated analysis may reproduce bias and confuse wording similarity with substantive comparability