Job evaluation methods all answer the same question: what is each role in your business worth, relative to the others? Five methods dominate in the UK. Ranking, paired comparison and classification compare whole jobs. Point-factor and factor comparison break jobs into scored parts. The Hay method is the best known branded version of the scored approach. 

They do not produce equally defensible answers. Pick the wrong one and you either bury a 30-person business in scoring admin, or build a pay structure that folds the first time an employee questions it. 

This article compares the five, shows where the legal line falls between them, and gives you a decision path by business size. If you want the full process for actually running an evaluation, from job descriptions through to pay bands, our step-by-step job evaluation guide for UK employers covers it end to end. This article helps you choose the method before you start. 

Job evaluation methods fall into two families 

Every method sits in one of two families, and the family matters more than the method. 

Non-analytical job evaluation compares whole jobs against each other. You judge that the operations manager role is bigger than the administrator role without breaking either job into parts. Ranking, paired comparison and job classification all work this way. They are quick, cheap and easy to explain. They are also judgements of impression, and they inherit whatever assumptions the judges bring into the room. 

Analytical job evaluation breaks each job into factors: knowledge, decision-making, responsibility for people or budgets, working demands. Each factor is scored against defined levels, the scores are weighted and totalled, and the totals set the rank order. Point-factor schemes and factor comparison work this way, and so do proprietary schemes like Hay. 

The split decides three things. How long the exercise takes. How consistent the results are. And how much weight the results carry if pay is ever challenged. That last one has a statutory edge, covered below. 

One boundary before the comparison starts: every method here grades the role, never the person in it. If your annual review cycle currently decides both grades and performance ratings, read job evaluation vs performance reviews as well, because mixing the two contaminates both records. 

Non-analytical job evaluation methods: fast but fragile 

Three methods make up the non-analytical family. Each trades rigour for speed in a different place. 

Ranking 

The ranking method of job evaluation is the simplest on the list. Line up every role from biggest to smallest by overall judgement. A morning with the senior team and a whiteboard produces a complete rank order. 

Its strength is speed. Nothing is faster or cheaper. For a firm with ten roles and no pay friction, ranking is often enough to sense-check salaries against each other. 

Its weakness is that the gaps are invisible. Ranking says role three is bigger than role four, but not by how much, so grade boundaries end up arbitrary. It also records no reasons. Two years later nobody can explain why the office manager outranks the senior technician, least of all to the senior technician. 

Best for: businesses under ten roles taking a first look at pay structure, as a temporary measure. 

Paired comparison 

Paired comparison is ranking with the arithmetic forced into the open. Every role is compared against every other role, one pair at a time. The bigger role in each pair scores two points, equal roles score one each. Point totals produce the rank order. 

Because it forces a decision on every single pair, it is noticeably more consistent than straight ranking. It also scales badly. Ten roles need 45 paired judgements. Twenty roles need 190. Somewhere past 15 roles the exercise collapses under its own admin. 

Best for: businesses under 15 roles that want more rigour than ranking without building a factor scheme. 

Job classification 

Classification, also called the grading method, works in the opposite direction. You write the grade definitions first: Grade 1 covers routine work under close supervision, Grade 2 adds independent judgement, and so on up the ladder. Then you slot each job into the grade whose definition fits best. 

Its strength is structure. New roles slot into an existing grade without re-running the whole exercise, which is why large public bodies used variants of it for decades. 

Its weakness is the slotting judgement. Real jobs rarely match one definition cleanly. The temptation is to slot each job into the grade that matches its current salary, which quietly recycles the very problem the exercise was meant to fix. 

Best for: businesses of roughly 20 to 50 roles that want a stable framework without factor scoring, and can resist grading to the payroll. 

Analytical job evaluation methods: slower but defensible 

Two methods make up the analytical family, and for most SMEs the choice between them takes one paragraph. 

Point-factor 

The job evaluation point factor method is the workhorse of the analytical family. It runs in four moves. 

1. Choose factors that matter in every role. Knowledge and skills, decision-making, responsibility for people, money or safety, and working demands are the usual set. 
2. Define levels within each factor, from lowest demand to highest, in plain written statements. 
3. Weight the factors, because decision-making usually deserves a larger share of the total than working conditions. 
4. Score every role against every factor, total the points, and let the totals set the rank order and the grade boundaries. 

The output is a scored record of why each role sits where it sits. That record is the difference between "we felt the role was bigger" and "the role scored 340 points against published factors". One is an opinion. The other is evidence. 

The cost is real but smaller than most owners expect. A practitioner shortcut: fully evaluate eight to twelve benchmark roles that everyone recognises, then slot the remaining roles against those benchmarks. And keep the scheme lean. Four to six factors are plenty for an SME. A twelve-factor scheme doubles the admin and almost never changes the rank order. 

Best for: any business past about 20 roles, and any business of any size where pay is already being questioned. 

Factor comparison 

Factor comparison also assesses jobs factor by factor, but ties each factor directly to money rather than to points, building pay rates from benchmark roles as it goes. It was influential in its day and survives mostly inside older schemes. 

For a UK SME today there is little reason to choose it. It is harder to explain than point-factor, harder to keep current as market pay moves, and the direct link between judgement and cash makes bias harder to spot. Know that it exists, then use point-factor instead. 

The Hay method in brief 

The Hay method is the most widely used proprietary analytical scheme. It scores every role on three factors, know-how, problem solving and accountability, using structured guide charts refined over decades. Large corporates and much of the public sector either use it or use schemes descended from it, which makes Hay results unusually easy to benchmark against market pay data. 

For an SME the trade is straightforward. You buy consistency, external credibility and comparability. You pay in licensing, training and time, and you accept a scheme you cannot freely rewrite. 

We have covered the method properly in its own article, including how the three factors work and what implementation involves. Read the Hay method job evaluation guide for the detail rather than a summary of a summary here. 

Job evaluation methods side by side 

Effort assumes the benchmark-role shortcut for the analytical rows. Score everything from scratch and the effort roughly doubles. 

Where the legal line falls between the two families 

The family split is not academic. It decides what your evaluation is worth as evidence. 

Under section 65 of the Equality Act 2010, equal work includes work rated as equivalent under a valid job evaluation study, and the section describes a study that values jobs in terms of the demands made on the worker. That is a description of analytical evaluation. A scheme that scores demands factor by factor fits it. A ranking exercise built on whole-job impressions does not. Section 65 applies in England, Scotland and Wales. Northern Ireland has its own equal pay law, the Equal Pay Act (Northern Ireland) 1970, which carries the same work rated as equivalent concept. 

The practical consequence runs in both directions. A sound analytical study lets you know which roles count as equal work before anyone else forces the question. And when an employee asks why a colleague earns more, your answer starts from scored evidence rather than recollection. The right to equal pay covers workers, apprentices and contractors hired to do the work personally, not only employees, so include those roles when you evaluate. ACAS guidance on equal pay points the same way: be clear, open and consistent about how pay is set. 

Two cautions. A study only carries weight if it was applied consistently and kept reasonably current. A point-factor scheme scored once years ago and ignored since carries little more weight than a whiteboard ranking. And no method makes a pay gap lawful by itself. Evaluation establishes which jobs are equal, which is where every equal pay question begins, not where it ends. 

The standing note: this is education, not legal advice. If a live equal pay question is already on your desk, take advice on your specific facts before restructuring anything. 

How to choose: a decision path for UK SMEs 

Match the method to your headcount, your risk and your appetite for admin. 

1. Under 10 roles and no pay friction: rank, sense-check the salaries, and diary a proper look for when you pass 15 roles. 
2. 10 to 20 roles: paired comparison, while the arithmetic is still manageable. 
3. Past 20 roles, or hiring fast: point-factor, built on benchmark roles. 
4. Pay already questioned, or visible gaps between men and women doing similar work: analytical only, whatever your size. This is the one situation where the cheap option is the expensive option. 
5. Competing with corporates for senior talent and buying market data: consider Hay or another proprietary scheme, because the benchmarking is the point. 
6. NHS or local government: sector-wide schemes already exist for your roles. Use them rather than inventing your own. 

To see the shape of the decision, picture a 28-role engineering firm with three pay complaints in a year, two of them from women comparing themselves with male colleagues on similar work. Ranking is out, because whole-job impressions are exactly what is being disputed. The firm builds a point-factor scheme with five factors, scores ten benchmark roles across two half-day panels, and slots the remaining eighteen against them. Total cost: around four working days spread over a month. The two disputed pairs come out one equal, one genuinely two grades apart. The equal pair gets aligned. The different pair gets an explanation backed by score sheets rather than seniority folklore. Whatever happens next, the firm now answers pay questions from evidence. 

Treat the families as an upgrade path, not a lifetime commitment. Plenty of firms rank at ten roles, classify at twenty-five and move to point-factor at fifty. The benchmark roles you define early carry forward at each step, so the early work is not wasted. 

Whichever method you pick, run a felt-fair check at the end. Show the results to two or three trusted senior people and ask what looks wrong. Then investigate the surprises through the scoring, never by pencil-editing the outcome. An evaluation record with unexplained manual adjustments is worse in a dispute than no record at all, because it proves the scheme bent under pressure. 

And plan for year two, because that is when most SME evaluations quietly die. The exercise happens once, then the next three hires get slotted by salary negotiation instead of scoring. Keep one page per role, score new and changed roles on a trigger list, and the structure stays alive for the cost of an hour here and there. 

Before you pick a method, know what it has to hold up 

Method choice is the second decision. The first is an honest view of where your pay structure is exposed, and most owners have not looked since the business was half its current size. 

The free HR Health Check takes under one minute and lets you know where your HR is solid and where it is exposed, across the areas that turn into real disputes, pay included. Start there, then choose your method with the gaps in full view. 

Take the free HR Health Check before your next hire reopens the pay conversation.