Pay differentiation means paying people in the same role differently, through bigger merit increases, bonuses, equity or faster promotions, because they contribute more. Done well, it keeps the people you can least afford to lose. Done badly, it turns colleagues into competitors and creates pay gaps you cannot explain.
This guide shows what the research says about pay dispersion and team performance, how far a typical merit budget really stretches, how to design a merit increase matrix with a worked example, which tools reward high performers outside base pay, and what the EU Pay Transparency Directive changes. It also uses TalentUp data on more than 23,000 job ads to show how much room the market itself leaves for differentiation.
Key takeaways
- Research does not say “differentiate” or “don’t”: pay dispersion helps when work is independent and differences are explained by contribution, and tends to hurt in interdependent teams when it is not.
- With a 3.5% budget, a top performer can get about 2x the average increase only if the top-rated group is small. Beyond that, someone gets zero.
- In 2026 job ads for the same role and level, the 75th percentile salary is typically 40% above the 25th. A 3-point merit premium would take about 11 years to cover that gap.
- Use the merit matrix for gradual differentiation and promotions, market adjustments, bonuses and LTIs for big moves.
- Once the Directive is transposed, differentiation must rest on objective, gender-neutral criteria, be documented, and show up cleanly in variable pay gap reporting.
What Pay Differentiation Is (and Why It Is a Dilemma)
Most organisations say they pay for performance. In practice, many spread increases fairly evenly, because differentiating is uncomfortable: managers must explain why one person got 6% and a teammate 2%.
The dilemma is real. Not differentiating tells your strongest people that extra effort is invisible, and they are usually the ones with the best outside options. Differentiating aggressively in a team where results depend on cooperation can make people guard information, avoid unglamorous work and compete for credit.
The useful question is not “should we differentiate?” but “how much, through which pay element, and on what evidence?” The answer depends on how interdependent the work is, how big the budget is and how clear your performance criteria are.
| Lever |
What it rewards |
Permanence |
Best for |
| Merit increase |
Sustained performance and growth in the role |
Permanent, compounds |
Gradual differentiation within a band |
| Market adjustment |
Being paid below market for the role |
Permanent |
Scarce skills, retention risks |
| Promotion |
Bigger scope and responsibility |
Permanent |
Outgrowing the role |
| Annual or spot bonus |
Results in a period or a one-off contribution |
One-off |
Top performers high in their range |
| Long-term incentives |
Multi-year value creation and retention |
Vests over years |
Critical talent and leaders |
| Development and recognition |
Potential and contribution |
Non-cash |
All employees, especially when cash is tight |
Peer-reviewed research on pay dispersion (how spread out pay is within a team or organisation) is mixed, but the pattern is consistent enough to guide decisions:
| Study |
Setting |
Finding |
| Bloom (1999), Academy of Management Journal |
Major League Baseball teams |
More compressed pay was positively related to several measures of individual and team performance. |
| Shaw, Gupta and Delery (2002), Strategic Management Journal |
Trucking and concrete pipe companies |
Dispersion improved results where individual incentives existed and work was independent; compression worked better where work was interdependent. |
| Trevor, Reilly and Gerhart (2012), Academy of Management Journal |
Interdependent team performance |
Dispersion explained by valued inputs (skills, performance used to attract and retain people) was positively related to team performance; unexplained dispersion had no effect or was harmful. |
| Garbers and Konradt (2014), Journal of Occupational and Organizational Psychology |
Meta-analysis of 146 studies, about 31,900 people |
Individual incentives improved performance (g = 0.32), team incentives too (g = 0.45), and equitably distributed team rewards outperformed equally distributed ones. |
| Bucciol, Foss and Piovesan (2014), PLoS ONE |
Italian Serie A football, 2009–2011 |
The effect of dispersion changed sign depending on how the team was defined: negative among players on the pitch, positive across the whole squad. |
Three practical lessons follow:
- Interdependence is the key variable. The more an outcome depends on people working together, the more you should rely on team rewards and the less on large individual gaps in base pay.
- Explained differences are less damaging than unexplained ones. Trevor and colleagues’ result is the strongest argument for documented, criteria-based differentiation, which is also what pay transparency law requires.
- “Equal” is not the same as “fair”. Even for team rewards, the meta-analytic evidence favours distribution in line with contribution over identical payouts.
There is also a fairness risk inside the process itself. In a large US service organisation, Castilla (2008) found that women and ethnic minorities received smaller salary increases than white men with the same performance ratings, job, unit and supervisor. The bias appeared when ratings were converted into pay, not in the ratings themselves.
How Much Differentiation a 3–4% Budget Really Allows
Company merit budgets for 2027 in Europe mostly cluster around 2.5–3%, with large differences by country and sector (see our salary increase budget 2027 guide). With a fixed pool, every extra point for one person is a point taken from someone else.
The table below shows the arithmetic for a team with equal salaries and a 3.5% budget, already at the generous end (illustrative):
| Scenario |
Top-rated share |
Top performers get |
Everyone else gets |
Top vs others |
| Flat increase |
0% |
3.5% |
3.5% |
1.0x |
| Small top group, 2x premium |
20% |
5.8% |
2.9% |
2.0x |
| Small top group, 3x premium |
20% |
7.5% |
2.5% |
3.0x |
| Large top group, 2x premium |
40% |
5.0% |
2.5% |
2.0x |
| 20% top at 2x, 10% get zero |
20% |
6.4% |
3.2% |
2.0x (0 for 10%) |
Two things stand out. First, rating inflation kills differentiation: when 40% of people are “top”, the top increase falls to 5.0% even with a 2x premium. Second, the money involved is modest: on a €45,000 salary, the gap between 5.8% and 2.9% is about €1,300 a year before tax.
That is why many reward teams use a carve-out: a slice of the budget (for example 0.5 points of a 3.5% pool) reserved for the top-rated group before the rest is distributed. It protects differentiation from being diluted, but it only works if ratings are calibrated and the top group stays small.
TalentUp Data: How Much Room the Market Leaves for Differentiation
40%
typical gap between the 25th and 75th percentile salary for the same role and level
11 years
of a 3-point merit premium to close that gap
16%
median width of a salary range stated in a job ad (min to max)
0.9%
of job ads explicitly link pay or pay rises to performance
Source: TalentUp job postings database, January–September 2026.
Employers do not pay the same for the same job title. Across 34 role and country combinations with enough data (about 23,400 job ads, mid-level employee roles only), the salary at the 75th percentile was typically 40% higher than at the 25th percentile:
| Role (mid-level) |
United Kingdom |
Spain |
France |
Italy |
Netherlands |
| Software engineer |
+62% (1,580) |
+44% (1,230) |
+40% (570) |
+25% (1,290) |
+29% (1,300) |
| Accountant |
+35% (1,280) |
+33% (1,620) |
+40% (870) |
+44% (630) |
+40% (360) |
| Project manager |
+44% (1,070) |
+39% (270) |
+31% (300) |
+40% (240) |
+30% (340) |
| Account manager |
+56% (1,200) |
+75% (470) |
+59% (210) |
+71% (220) |
n/a |
| Customer service |
+19% (1,480) |
+42% (1,200) |
n/a |
+49% (160) |
+42% (290) |
| HR generalist / HR business partner |
+42% (590) |
n/a |
n/a |
+50% (200) |
+35% (180) |
| Warehouse operative |
+22% (620) |
+24% (1,300) |
+11% (220) |
+24% (440) |
+21% (200) |
| Nurse |
+42% (710) |
+31% (550) |
+50% (200) |
n/a |
n/a |
How much the 75th percentile advertised salary exceeds the 25th percentile, number of ads in brackets. n/a = fewer than 150 ads.
What this means for differentiation:
- The market differentiates far more than a merit cycle can. At a 3-point annual premium, closing a 40% gap takes about 11 years; at 2 points, about 17. If a high performer is worth a top-quartile salary, a merit increase alone will not get them there in time to keep them.
- Sales-facing roles show the widest spreads (account managers: +56% to +75%), consistent with more negotiated pay and variable components. Operational roles such as warehouse work, where pay is closer to collective agreements and minimum wages, show the narrowest (+11% to +24%).
- Individual ranges are narrow. Among ads that state a salary range, the median range is 16% wide from minimum to maximum (UK 16.7%, n ≈ 750; Spain 14.3%, n ≈ 4,300; France 20.6%, n ≈ 630; Italy 15.6%, n ≈ 380; Netherlands 29.2%, n ≈ 270). Hiring ranges leave limited room, so most differentiation happens after hiring.
- Employers rarely advertise pay for performance. In a random sample of about 24,400 job ads in 11 European countries, only 0.9% explicitly link pay or pay rises to performance. It is under 0.5% in nine countries; the exceptions are Germany (3.7%), where most mentions are the generic leistungsgerechte Vergütung, and Sweden (4.0%), where public-sector ads routinely state individuell lönesättning (individual pay setting under collective agreements).
How to Design a Merit Increase Matrix
A merit increase matrix sets the increase for each combination of performance rating and position in the range, usually measured by compa-ratio (salary ÷ range midpoint). It differentiates by performance while moving people towards the right pay level, so strong performers who are underpaid catch up faster than strong performers already paid above midpoint. For the metrics, see range penetration vs compa-ratio.
An illustrative matrix for a 3.5% budget:
| Performance |
Compa-ratio below 0.90 |
0.90 – 1.10 |
Above 1.10 |
| Exceptional |
7.0% |
5.5% |
3.5% + lump sum |
| Strong |
5.0% |
4.0% |
2.5% |
| Solid |
3.5% |
2.5% |
1.5% |
| Below expectations |
0–1.0% |
0% |
0% |
Worked example (illustrative)
A team of eight in the same role, with a range midpoint of €50,000 and payroll of €401,500:
| Employee |
Rating |
Salary |
Compa-ratio |
Increase |
Cost |
| A |
Exceptional |
€46,000 |
0.92 |
5.5% |
€2,530 |
| B |
Exceptional |
€43,000 |
0.86 |
7.0% |
€3,010 |
| C |
Strong |
€51,000 |
1.02 |
4.0% |
€2,040 |
| D |
Strong |
€56,500 |
1.13 |
2.5% |
€1,413 |
| E |
Solid |
€49,000 |
0.98 |
2.5% |
€1,225 |
| F |
Solid |
€52,000 |
1.04 |
2.5% |
€1,300 |
| G |
Solid |
€57,000 |
1.14 |
1.5% |
€855 |
| H |
Below expectations |
€47,000 |
0.94 |
0% |
€0 |
| Total |
|
€401,500 |
|
3.1% |
€12,373 |
The matrix spends 3.1%, leaving about €1,700 of a 3.5% budget. That remainder can fund a one-off payment for strong performers high in the range, or a reserve for equity corrections. Note what the matrix does: B gets the largest increase because B is exceptional and low in the range, while G, a solid performer already paid well above midpoint, gets the smallest positive increase.
Design rules that keep the matrix fair and affordable:
- Simulate it on your real population before launch, so the weighted average stays within budget.
- Give managers a range per cell (for example 4.5–5.5%) rather than a single number, and require a written reason for going outside it.
- Keep the top rating rare. If more than about a quarter of people are rated exceptional, differentiation collapses (see the budget table above).
- Benchmark the midpoint. A matrix only makes sense if the range reflects the market; refresh it with current data before each cycle. For the design steps, see how to build salary bands.
The merit matrix is a slow, compounding tool. When someone needs a bigger or faster signal, or is already high in the range, other levers do the job better and protect internal equity:
- Promotion or re-levelling. If a high performer is doing a bigger job, change the job. Fund promotions from a separate pool, not from the merit budget.
- Market adjustment. If the person is below market for a scarce skill, fix it as a market adjustment with its own documentation, not as a disguised merit increase.
- Spot and annual bonuses. One-off cash rewards differentiate without permanently widening base pay gaps inside the team. Our guides on variable compensation and bonus psychology cover design and payout curves.
- Team-based rewards. For highly interdependent work, combine a team bonus with an individual modifier, in line with the evidence that equitably distributed team rewards outperform equal splits.
- Long-term incentives. Shares, options or deferred cash for critical talent tie reward to multi-year value; see long-term incentive plans.
- Development and recognition. Stretch assignments, visibility, training budgets and flexibility are real differentiators, especially when cash is tight; see what HR can offer when there is no raise budget.
Before any large individual move, check two side effects: pay compression with the level above (see our pay compression guide) and internal equity with peers doing work of equal value (see internal equity vs external competitiveness).
Calibration and Bias Controls
Differentiation is only as good as the ratings behind it. A defensible process includes:
- Clear performance criteria published before the cycle starts, with examples of what each rating means for each job family.
- Calibration sessions where managers compare ratings across teams, so “exceptional” means the same thing everywhere.
- Separation of steps with accountability. Castilla’s study found bias where ratings were converted into pay; require managers to justify recommendations outside the matrix and have HR review them.
- Pre-approval equity checks. Before sign-off, compare proposed increases by gender, part-time status, leave and location at the same rating and compa-ratio. Fix unexplained differences before letters go out.
- Pro-rating rules for part-time workers and people on parental or sick leave, so absence does not quietly turn into a lower rating or a smaller bonus.
- A documented rationale for every increase, bonus and promotion, kept with the pay data.
Pay Differentiation and the EU Pay Transparency Directive
The EU Pay Transparency Directive (2023/970) does not ban pay for performance. Its transposition deadline was 7 June 2026, and its obligations apply through national laws once transposed in each member state. Where they apply:
- Objective, gender-neutral criteria. Pay structures must allow pay to be compared using objective, gender-neutral criteria such as skills, effort, responsibility and working conditions (Article 4). Performance can justify differences, but it must be assessed consistently.
- Transparent pay progression. Employers must make available the criteria used to determine pay, pay levels and pay progression (Article 6). Member states may exempt employers with fewer than 50 workers from the pay progression part.
- Right to information. Workers can request their pay level and average pay levels by sex for workers doing the same work or work of equal value (Article 7). Unexplained differences become visible.
- Variable pay gap reporting. Article 9 indicators include the gender gap in complementary or variable pay and the share of women and men who receive it. Employers with 250+ workers report every year from 7 June 2027; 150–249 every three years from 7 June 2027; 100–149 every three years from 7 June 2031.
- Joint pay assessment. If reporting shows an unexplained gap of 5% or more in any category of workers that is not remedied within six months, the employer must carry out a joint pay assessment with workers’ representatives.
- Burden of proof. In pay discrimination claims, the burden of proof shifts to the employer (Article 18), so undocumented, discretionary differentiation becomes a liability.
The practical response is to write differentiation rules into your compensation philosophy, run proposed increases and bonuses through a pay equity audit before approval, and use our pay transparency resources to check your ranges against the market.
Checklist for a Directive-Ready Pay Cycle
- Decide how interdependent each team is, and set the balance between individual and team rewards accordingly.
- Split the budget into merit, market adjustment, promotion and equity pools.
- Benchmark ranges with current market data, by role and country, before building the matrix.
- Build and simulate the merit matrix, with a carve-out if you want real differentiation.
- Calibrate ratings and keep the top group small.
- Use one-off and long-term rewards for people high in their range or needing a bigger signal.
- Run equity checks by gender and working pattern before approval.
- Document and communicate the criteria, so every employee can understand how their increase was set.
TalentUp’s European Salary Benchmark Report and plans help you set ranges on live market data; our methodology explains how the data is built.
Sources
- TalentUp. Job postings database, January–September 2026: about 23,400 mid-level employee ads with salary in 5 countries, and a random sample of about 24,400 ads in 11 countries (see methodology above).
- European Union. Directive (EU) 2023/970 on pay transparency. EUR-Lex.
- Bloom, M. (1999). The performance effects of pay dispersion on individuals and organizations. Academy of Management Journal, 42, 25–40.
- Shaw, J. D., Gupta, N., and Delery, J. E. (2002). Pay dispersion and workforce performance: Moderating effects of incentives and interdependence. Strategic Management Journal, 23(6), 491–512. doi:10.1002/smj.235
- Trevor, C. O., Reilly, G., and Gerhart, B. (2012). Reconsidering pay dispersion’s effect on the performance of interdependent work: Reconciling sorting and pay inequality. Academy of Management Journal, 55(3), 585–610. doi:10.5465/amj.2006.0127
- Garbers, Y., and Konradt, U. (2014). The effect of financial incentives on performance: A quantitative review of individual and team-based financial incentives. Journal of Occupational and Organizational Psychology, 87(1), 102–137.
- Bucciol, A., Foss, N. J., and Piovesan, M. (2014). Pay dispersion and performance in teams. PLoS ONE, 9(11), e112631. doi:10.1371/journal.pone.0112631
- Castilla, E. J. (2008). Gender, race, and meritocracy in organizational careers. American Journal of Sociology, 113(6), 1479–1526.