From job catalogue to policy.
- 01Import your job catalogueWord, Excel, PDF or CSV — no retyping.
- 02Algorithmic scoringWith a rationale per criterion; a human decides.
- 03Score approvalAudit trail: who, when, why.
- 04Data-driven gradesBoundaries where the scores have gaps.
- 05ComparisonWith your current pay grades, flagging disputed job pairs.
- 06Policy exportMethodology, grades and classification of all jobs.
We evaluate jobs, not people. The app holds no salaries and no personal data of employees.
Five things you wouldn't have without the app
- 01
Boundaries where your data puts them.
Linear grades cut with a ruler: two almost identical jobs land in different grades, two very different ones in the same grade. Paryty puts the boundary in the gap between scores, so every employee can understand it.
- Linear: boundary by the ruler
- Paryty: boundary in the gap
Linear gradesequal slices of 50 points
Parytyboundaries in the gaps between scores (Fisher–Jenks)
In this example, every boundary sits in a gap larger than a single differing judgement (7 points). Where the gap is smaller, Paryty flags the boundary as weak and won't export the policy without review. A linear boundary sits wherever the ruler falls.Illustration: 35 jobs, 8 grades, illustrative range 100–500, same points in both rows. 7 points is the highest criterion weight under default weights: the total shifts by this much if an evaluator rates one criterion one level differently. - 02
You see a weak boundary first.
Every boundary is marked as strong or weak. The app won't let a weak boundary into the policy until you justify it.
- Strong boundary
- Weak boundary: needs a justification
- R1 | R2strong boundaryWarehouse operative | Storekeeper
- R4 | R5strong boundaryClerk | Senior clerk
- R6 | R7weak boundaryHead of department | Head of division
The jobs on either side are too close. Before the policy is exported, HR must justify the boundary or re-evaluate the jobs.
Illustration - 03
It finds pay pairs you couldn't defend.
The app compares your current pay grades with the difficulty grades and shows every case where a more demanding job is paid less.
- Disputed pair
Comparison with pay gradesDisputed pairs: 135 jobsJob Difficulty Pay grade Purchasing clerk R3 · 194 PR 5 Warehouse shift lead R3 · 191 PR 4 Storekeeper R2 · 176 PR 5 Administrator R2 · 163 PR 4 The shift lead's job is more demanding than the storekeeper's, yet sits in a lower pay grade. The difference must be removed or justified by a separate, gender-neutral reason.
Illustration - 04
A written answer to “why” for every boundary.
A difference in points won't convince a judge. The app compares the highest-scored job in the lower grade with the lowest-scored job in the upper grade and records which tasks, responsibilities and conditions set them apart.
- The criterion that separates the grades
Boundary report · R2 | R3What separates two adjacent gradesstrong boundaryHighest in R2Storekeeper176 pointsgap 15Lowest in R3Warehouse shift lead191 points- Responsibility for people1 → 3Leads a shift of six to eight colleagues and assigns their work.
- Autonomy2 → 3Decides alone how to reassign work when someone is absent or a delivery is late.
- Communication2 → 3Coordinates goods receipt with purchasing and carriers.
The same for both: professional knowledge (2), physical effort (3), working conditions (3).
Conclusion: the difference comes from leading people and autonomy in decisions, not from education. The boundary is justified by the content of the work.
Illustration - 05
You have the proof before anyone asks for it.
An internal policy: methodology, grades and the classification of all jobs in one document, ready for adoption with the trade union or works council.
- One document, ready for adoption
Internal policy · .docxJob evaluation methodology- 1.Purpose and basis
- 2.Criteria, weights and anchor scales
- 3.Rule for grade boundaries
- 4.Grades and classification of all jobs
- 5.Boundary justifications
- 6.Comparison with pay grades
- 7.Audit trail
Adopted through the procedure with the trade union or works councilFor: inspectorate · trade union · courtIllustration
Grades that hold up in a dispute
In court, what counts is why the boundary is there.
- Why is the boundary right here?Linear gradesBecause the range was split into equal parts. The reason is arithmetic, not the work.ParytyBecause there is a gap in job demands here. The rule is published before the calculation; the calculation is reproducible.
- Do the jobs on either side of the boundary really differ?Linear gradesPerhaps by a point or two — less than a single differing judgement.ParytyBy more than one judgement (7 points). A narrower boundary is flagged as weak; the policy can't be exported without review.
- How are jobs grouped into grades?Linear gradesBy the ruler: a grade can split two almost identical jobs while merging jobs far apart (45 points in the example).ParytyBoundaries are set so that the spread within grades is as small as possible for the chosen number of grades.
- What about a job right at the boundary?Linear gradesNothing. One point below the boundary means a lower grade.ParytyWithin 4 points it's a borderline case: compared with similar jobs; decision and reason recorded.
| Inspector, union or court asks | Linear grades | Paryty |
|---|---|---|
| Why is the boundary right here? | Because the range was split into equal parts. The reason is arithmetic, not the work. | Because there is a gap in job demands here. The rule is published before the calculation; the calculation is reproducible. |
| Do the jobs on either side of the boundary really differ? | Perhaps by a point or two — less than a single differing judgement. | By more than one judgement (7 points). A narrower boundary is flagged as weak; the policy can't be exported without review. |
| How are jobs grouped into grades? | By the ruler: a grade can split two almost identical jobs while merging jobs far apart (45 points in the example). | Boundaries are set so that the spread within grades is as small as possible for the chosen number of grades. |
| What about a job right at the boundary? | Nothing. One point below the boundary means a lower grade. | Within 4 points it's a borderline case: compared with similar jobs; decision and reason recorded. |
Once a worker establishes the facts, the employer must prove there was no discrimination (Directive (EU) 2023/970, Article 18). An objective reason must stem from the work.
A formal policy.
- Methodology as policy.docxRules, weights, grades and the classification of all jobs in one document — adopted through the procedure with the union or works council.For: internal rules · works council · inspection · courtMethodology · v2
- Comparison with payCurrent pay grades vs. job-value grades, flagging disputed job pairs.For: management and HR — where the structure doesn't follow job value
- Audit trailWho, when, what and why — for every score, boundary and decision. No personal data of jobholders.For: proof of consistency over time
- Score DM-014 changed 3 → 409:12
- Approval, M. Kralj09:40
- Weight “Responsibility” 12 → 1410:05
- New methodology version v210:06
Employers writing or refreshing their job catalogue
- Manufacturing and servicesThe job catalogue exists, the basis for classification doesn't.
- Companies before a pay system overhaulEvaluation is the input you need before negotiating pay.
- Consultants and HR agenciesOne methodology for many clients, each in its own space.
What people ask us most
How long does it take?
Setting up the methodology takes a few hours. For 50 positions with two evaluators, plan a few days of work spread over a week or two.
Do we need pay grades?
No. If you have them, the number of difficulty grades follows from them and you get a comparison with your structure. If not, you set the number of grades yourself, with a recorded reason.
What about our existing job descriptions?
Import them from Excel, CSV or a Word or PDF job catalogue. The app suggests positions from the document and you confirm them.
Can one person do the evaluation?
Yes, with controls: an approved job description, a delayed blind re-score of part of the scores and a recorded reason for the exception to the four-eyes principle.
What happens when the work changes?
A new methodology version is created. Earlier classifications stay valid for the period they applied to, with their own points and grades.
Does the app say how much anyone should earn?
No. It tells you which positions are equivalent in difficulty. Pay stays your decision.
Who sees the results?
You do. There are three roles: HR admin, evaluator and approver; every access and every action is recorded in the trail.