Internal equity vs external competitiveness is the core trade-off in compensation design: internal equity asks whether people doing work of comparable value inside your organization are paid consistently, while external competitiveness asks whether your pay matches what other employers offer for the same jobs. Markets move fast and unevenly; pay structures move slowly. The gap between the two is where pay compression, inversion and equal pay risk appear.
This guide defines both concepts, explains why they conflict, shows how to measure each with compa-ratio, range penetration and market ratio (with formulas and a worked example), and gives a decision framework for when to follow the market and when to protect equity. It closes with the constraint that matters most in Europe: equal pay for work of equal value under the EU Pay Transparency Directive, as national laws apply. Along the way, TalentUp job-postings data shows how much one hot skill, AI, is pulling salaries apart.
Key takeaways
- Internal equity = consistent pay for work of comparable value inside the company; external competitiveness = pay in line with the relevant market.
- The conflict is structural: hot skills and new-hire premiums move faster than bands, creating compression and inversion.
- Measure both with three ratios: compa-ratio and range penetration (internal) and market ratio (external).
- In TalentUp’s UK data, AI and machine learning engineer postings advertise a median salary 33% above software engineer postings in 2026, up from 22% in 2024.
- Pay above the band through documented, reviewable mechanisms (market zones, time-limited premiums), never informal exceptions: where the EU directive applies, the employer must prove differences are objective and gender-neutral.
Internal Equity vs External Competitiveness: Definitions
Internal equity (internal pay equity) is the principle that employees whose jobs are of comparable value to the organization are paid consistently, with differences explained by objective factors such as level, skills, performance and experience. It is built through job evaluation and a job architecture of families and levels.
External competitiveness (also called external equity or market competitiveness) is how your pay compares with what other employers pay for the same jobs in the labor markets where you recruit. It is built through market pricing: matching jobs to salary data and deciding whether to lead, match or lag the market.
|
Internal equity |
External competitiveness |
| Question it answers |
Are comparable jobs paid consistently here? |
Can we attract and keep people at this pay? |
| Reference point |
Other jobs and employees in the organization |
Pay for the same jobs at other employers |
| Main tools |
Job evaluation, job architecture, salary bands |
Salary benchmarking, market pricing, pay policy line |
| Key metrics |
Compa-ratio, range penetration, pay gaps between comparable groups |
Market ratio, positioning against market percentiles |
| Risk if ignored |
Perceived unfairness, disengagement, equal pay claims |
Failed hires, regretted turnover, rising time to fill |
| Legal angle in the EU |
Equal pay for same work or work of equal value |
Pay ranges for candidates where transparency laws apply |
Neither is optional. A structure that is perfectly fair but 15% below market cannot hire; a structure that copies every market quote ends up paying two people in the same job very differently for no reason you can explain.
Why Internal Equity and Market Pay Conflict
The two principles collide for a handful of recurring reasons.
Hot Skills Move Faster Than Bands
Markets reprice individual skills, not whole job families. When demand for a skill such as AI, cybersecurity or a specific regulatory specialism spikes, the market rate for those profiles rises while adjacent roles in the same grade stay flat. Bands that were calibrated on last year’s data suddenly sit below market for one group and on target for everyone else.
New-Hire Premiums
Employers usually have to pay current market rates to hire, while existing employees move only with the annual salary review. If market rates rise faster than merit budgets, every new hire arrives closer to, or above, colleagues who have been in the job for years.
Pay Compression and Pay Inversion
- Pay compression: the difference in pay between experienced and new employees, or between adjacent levels, becomes too small to reflect real differences in skills, experience and responsibility.
- Pay inversion: the extreme case, when a new hire earns more than a tenured peer in the same job, or a team member earns more than their manager.
Both are signals that external pressure has overtaken internal structure. They rarely stay hidden: in markets where job ads publish salary ranges, employees can see what new hires are offered. Our analysis of transparent salary job postings shows how common published ranges have become.
Each above-band offer approved “just this once” becomes a precedent. After a few hiring cycles the structure no longer explains actual pay, and nobody can say why two people in the same grade earn different amounts. That is exactly the situation pay transparency rules are designed to expose.
TalentUp Data: How Hot Skills Pull Salaries Apart
To show the market pressure on internal equity, we compared the median posted salary of AI and machine learning engineer postings with software engineer postings in the same country and year, using the structured salary field of TalentUp’s job-postings database.
+33%
median posted salary for AI/ML engineers vs software engineers in the UK, 2026
+22%
the same UK premium in 2024
£100,000
UK median for AI/ML engineer postings in 2026, vs £75,000 for software engineers
+8%
AI/ML premium in the US in 2026, down from 16% in 2024
Source: TalentUp job postings database, January–September of each year. UK figures in GBP.
| Market and year |
AI/ML engineer median |
Software engineer median |
AI/ML premium |
Sample (AI/ML · software) |
| United Kingdom, 2024 |
£80,000 |
£65,350 |
+22.4% |
~340 · ~2,580 |
| United Kingdom, 2025 |
£95,000 |
£70,400 |
+34.9% |
~1,270 · ~2,490 |
| United Kingdom, 2026 |
£100,000 |
£75,000 |
+33.3% |
~440 · ~860 |
| United States, 2024 |
– |
– |
+15.7% |
~1,880 · ~6,650 |
| United States, 2025 |
– |
– |
+13.8% |
~3,700 · ~7,250 |
| United States, 2026 |
– |
– |
+7.9% |
~2,860 · ~8,800 |
Two lessons for compensation teams:
- Premiums are large and they move. In the UK, the median posted salary for AI/ML engineers rose 25% between 2024 and 2026 (£80,000 to £100,000), against 15% for software engineers (£65,350 to £75,000). A band set in 2024 for both would now be behind for one population and roughly on target for the other.
- Premiums can also shrink. In the US the AI/ML premium in posted salaries roughly halved over the same period. That is the argument for treating skill premiums as reviewable adjustments rather than permanent changes to base pay.
How to Measure Internal Equity and External Competitiveness
Three ratios cover most of the work. Use them together: each one alone can hide a problem.
| Metric |
Formula |
What it tells you |
How to read it |
| Compa-ratio |
Salary ÷ range midpoint |
Where someone sits against the target rate for their grade |
1.00 = on midpoint; below = room to grow; above = paid beyond target |
| Range penetration |
(Salary − range minimum) ÷ (range maximum − range minimum) |
How far someone has progressed through the band |
0% = at minimum; 100% = at maximum; above 100% = out of band |
| Market ratio |
Salary (or midpoint) ÷ market rate for the job, usually the median |
How competitive pay or the structure is externally |
1.00 = at market; below = lagging; above = leading |
Two more checks complete the picture:
- Group compa-ratios: average compa-ratio by gender, tenure band or hire year within the same grade. A group of recent hires averaging 1.08 while tenured peers average 0.94 is compression in numbers.
- Level differentials: median pay of each level compared with the level below. When the gap between adjacent levels narrows year after year, promotions stop paying and managers can end up earning less than their teams.
Worked Example
A software engineering grade with a range of £60,000–£90,000 and a £75,000 midpoint, set to match the 2026 UK median for software engineer postings above. One AI engineer has been hired into the same grade at the AI/ML market median.
| Employee |
Tenure |
Salary |
Compa-ratio |
Range penetration |
Market ratio (own market) |
| A, software engineer |
6 years |
£72,000 |
0.96 |
40% |
0.96 vs £75,000 |
| B, software engineer |
3 years |
£68,000 |
0.91 |
27% |
0.91 vs £75,000 |
| C, software engineer |
New hire |
£80,000 |
1.07 |
67% |
1.07 vs £75,000 |
| D, AI engineer |
New hire |
£100,000 |
1.33 |
133% |
1.00 vs £100,000 |
Illustrative example. Market rates from the TalentUp UK medians above; salaries and tenures are hypothetical.
What the numbers show:
- Inversion: C, hired this year, earns £8,000 more than A, who has six years in the job. A’s compa-ratio of 0.96 looks healthy on its own; only the comparison by tenure reveals the problem.
- A market outlier, not an overpayment: D is 33% above the band midpoint but exactly at the market for AI engineers. The issue is not D’s pay, it is that one band is trying to price two different markets.
- The fix differs by case: A and B need an equity adjustment over one or two review cycles; D needs a documented market reference zone or skills premium for AI roles, so the difference is explained by the structure rather than by an exception.
To build the midpoints themselves, see our guide to salary benchmarking for pay transparency compliance and how TalentUp builds its market data on the methodology page.
When to Follow the Market and When to Protect Equity
There is no universal rule, but the same questions decide most cases. The more answers fall in the right-hand column, the stronger the case for a market-driven exception.
| Question |
Protect internal equity |
Follow the market |
| Is the role critical to strategy or revenue? |
Supporting or easily covered |
Critical, with a measurable cost of vacancy |
| Is the market premium evidenced? |
Anecdotal or based on one candidate |
Consistent across several data sources and hiring attempts |
| Is the premium likely to last? |
Spike or hype cycle |
Structural shortage of the skill |
| How many employees are comparable? |
Many incumbents in comparable roles |
Small, clearly distinct population |
| Can the difference be explained objectively? |
Hard to tie to skills, responsibility or market evidence |
Clearly tied to a distinct skill set and documented market data |
| Gender composition of the comparison groups |
Higher-paid group mostly one gender, lower-paid mostly the other |
Balanced, or no equal-value comparator affected |
| Can you afford to fix the knock-on effects? |
No budget to lift comparable employees |
Budget available for equity adjustments too |
Three rules keep decisions consistent:
- Decide at the job or skill level, not the person level. If one AI engineer warrants a premium, every comparable AI engineer does.
- Price the knock-on cost before approving. An above-band hire usually triggers equity adjustments for peers; include them in the business case.
- Write down the reason. The documented rationale is what lets you explain the difference to employees, managers and, if needed, a court or labor inspectorate.
Total rewards practitioners frame the dilemma the same way. Pavithra Venkatesh, a total rewards leader at Dar in Dubai, argues that the answer is not rigid bands but explicit rules for when to flex them:
Pavithra VenkateshStrategic Total Rewards Leader, Dar (Dubai)
Take a common scenario hiring a high impact performer from a competitor. […] Forcing that hire into a predefined band creates a false choice. Stay within range and risk losing talent, or stretch the framework and challenge internal equity. […] The answer isn’t tighter or looser governance within the bands, but clarity on when to flex and how to protect trust when we do.
Job Architecture and Job Evaluation
A clear structure of job families, levels and level descriptors is the foundation of internal equity: it defines which jobs are comparable. Use a consistent, gender-neutral job evaluation method so that “comparable” rests on the work, not on job titles or historical pay.
Market Pricing and a Pay Policy Line
Match each benchmark job to market data, choose a target percentile (for example the median, or a higher percentile for critical families) and set midpoints from that pay policy line. Refresh the pricing at least once a year, and more often for families where the market moves fast. A cross-country reference such as the European Salary Benchmark Report helps when you hire in several countries.
Market Reference Zones and Skills Premiums
- Market reference zones (or market pay zones) give different midpoints to different job families within the same level, so an AI engineer and a software engineer can share a career level but not a market rate.
- Skills premiums or market adjustments are separate, named pay elements for a scarce skill, with eligibility criteria, an amount or percentage, and an annual review date. Because they are visible as a separate component, they can be reduced or removed if the premium disappears, as the US data above suggests can happen.
Salary Review Budgets
Split the annual budget into explicit pots instead of a single merit percentage:
| Budget pot |
Purpose |
Typical use |
| Merit |
Reward performance and progression |
Standard individual increases |
| Market adjustment |
Move structures and people that fall behind the market |
Families where market ratios dropped below target |
| Equity fund |
Close unexplained gaps between comparable employees |
Compression, inversion and gender gaps found in audits |
Ring-fencing an equity fund stops compression fixes from competing with merit increases every year.
Non-Base Levers
Sign-on bonuses, retention bonuses and long-term incentives can close a competitive gap without permanently moving base salary above the band. Use them with care: under equal pay rules, variable and complementary components are still pay, and they need the same objective criteria as base salary.
Communication
Employees accept differences they understand. Publish the structure and the criteria for positioning and progression, train managers to explain why someone sits where they do, and decide how much to disclose using our guide on how much pay information to share.
Equal Pay for Work of Equal Value: The EU Directive Constraint
The EU Pay Transparency Directive (EU) 2023/970 had to be transposed by 7 June 2026, and its obligations apply to employers through national laws, once transposed. Where it applies, it turns internal equity from good practice into something employers must be able to demonstrate:
- Pay structures must allow comparison of the value of work on objective, gender-neutral criteria that include at least skills, effort, responsibility and working conditions (Article 4).
- Criteria for pay setting and progression must be accessible to employees (Article 6), and workers can request their individual pay level and average pay levels by gender for workers doing the same work or work of equal value (Article 7).
- Burden of proof shifts to the employer in equal pay disputes (Article 18).
- Gender pay gap reporting: employers with 250+ workers report annually from 7 June 2027; 150–249 every three years from 7 June 2027; 100–149 every three years from 7 June 2031. A joint pay assessment is required when an unexplained gap of 5% or more in a category of workers is not remedied within six months.
Can the Market Justify a Difference?
The directive’s criteria do not include labor market shortage, and market pressure is not a blanket justification. Under the Court of Justice’s Enderby judgment (C-127/92, 1993), a shortage of candidates can objectively justify a pay difference between jobs of equal value, but only to the extent that market forces demonstrably cause it. In the UK, which is not covered by the directive, an employment tribunal ruled in 2024 that Next could not rely on market rates to justify paying its mostly female store staff less than its mostly male warehouse staff.
In practice, that means a market premium for a hot skill is defensible when it is evidenced with market data, applied consistently to everyone with that skill, documented and periodically reviewed. An informal exception for one person is not. For the full compliance picture see our EU Pay Transparency Directive guide and the 2026 transposition timeline.
Conclusion
Internal equity and external competitiveness are not opposites to choose between; they are two constraints a pay structure has to satisfy at once. The practical recipe:
- Measure both with compa-ratio, range penetration and market ratio, by grade, tenure and gender.
- Refresh market pricing often enough to catch skills that move, like AI, where TalentUp data shows a UK premium of about a third.
- Handle hot skills through structure (market zones, time-limited premiums) instead of exceptions.
- Fund equity adjustments explicitly so compression gets fixed, not postponed.
- Document every reason, because where the directive applies, the employer must prove its pay differences are objective and gender-neutral.
Sources