Guide · Market pricing

How to benchmark pay across countries

Price each role in its own market, its own currency and on its own country's idea of pay. Then make the rules once, for everyone.

Short answer: Benchmarking pay across countries works when each role is priced against local market data, in local currency and on that country's own definition of pay. The decisions are then made once for the whole company: One job framework, one stated target against each market and one way to approve exceptions. Converting a home-country range at today's exchange rate prices the currency, not the job.

One spreadsheet, six countries

Picture a 170-person software company in Cleveland that sells planning tools to manufacturers. It started hiring abroad three years ago, one role at a time. It now has engineers in Poland and India, a customer support team in the Philippines, two salespeople in the UK and a small product team in Germany. About a third of its people work outside the United States.

Each of those hires was priced differently. The Polish engineers came through an employer of record, whose recruiter suggested a salary. The Indian team was set up by a local manager who used a free salary site and offers from rival firms. The German product manager negotiated hard and got close to what a US hire would have cost. The support team in the Philippines was priced as a share of the US support rate, because that is what the CFO had done at a previous company.

Then the board asked for one view of pay across the company before next year's budget. The head of people built a spreadsheet with every role converted to US dollars. It raised more questions than it answered. Indian salaries had gone up by several percent in dollar terms since last year, though no one in India had received a raise. Engineers at the same level in Poland and India were paid in a ratio that nobody could explain. The German offer looked generous in dollars and turned out to be about right for Berlin. The Philippine figures left out a month's pay that the law requires.

The CFO asked the obvious question: Are we paying people fairly? Are we paying too much? The spreadsheet could not answer either one, because it compared numbers that did not mean the same thing.

This is a composite of a common situation at companies this size, not a client. The details are illustrative. The pattern is common because international hiring usually starts as a series of one-off deals. Nobody sets the rules until there are enough people abroad for the inconsistencies to show.

Why benchmarking pay across countries is hard

Benchmarking means comparing a company's pay for a job with what other employers pay for the same job. Within one country, the hard parts are matching jobs well and picking the right data. Across countries, three more things change at once: The market, the currency and the meaning of the word "pay".

Two good goals pull against each other. The first is local accuracy. People compete for jobs in their own labor market. An engineer in Kraków compares an offer with other offers in Poland, in złoty, after Polish taxes. A company that ignores local pay levels overpays in some countries and cannot hire in others.

The second goal is global consistency. The company wants a senior engineer to mean the same thing in every country. It wants promotions, raises and exceptions to follow the same rules wherever people sit. It wants the board to see one picture. Those are decisions about the company, not about any one market.

The common mistake is to solve one goal by giving up the other. Some companies chase consistency by taking one country's pay as the base and converting or discounting it for everyone else. Others chase accuracy by letting each country do what it likes. The first gets local pay wrong. The second leaves the company with no rules at all.

Pay also moves at different speeds in different places. In a survey of 1,799 organizations published in July 2026, projected average salary increase budgets for 2027 ranged from 2.7% in Switzerland to 3.6% in the United States and Germany, 5.5% in Poland and 9.2% in India [1]. A pay structure set once and left alone drifts out of line with each market at a different rate.

Our guide on one global rate or pay by location covers the policy question: How much location should matter and what to tell staff. This guide covers the method underneath it: How to get a sound market number for each role in each country and how to keep those numbers comparable.

Five ways cross-country benchmarking goes wrong

Each of these starts from a reasonable wish: To keep things simple, to be fair, to be consistent or to respect local knowledge. The damage comes from treating a different market as a version of the home one.

"Convert the US range at today's exchange rate"

What it looks like: The company has good US pay data. For other countries it takes the US range and converts it into local currency. A softer version applies a fixed country factor: India at a set share of US pay, Poland at another.

Why smart teams fall into it: It uses the data the company already trusts. It needs no new survey. It looks consistent, because every country hangs off one number.

What it costs: It ties pay to the currency market instead of the labor market. On 23 October 2025 one US dollar bought about 87.85 Indian rupees. A year later it bought about 94.94 [2]. A dollar-based range converted into rupees rose by about 8% over that year with no change at all in what Indian employers were paying. The reverse happens when a currency strengthens: Local staff see their pay "fall" against the home figure and the company thinks it is overpaying. A fixed country factor avoids the currency swings but hides a bigger problem. The gap between US and Indian pay is not the same for a support role, a junior engineer and a senior product manager. One factor cannot be right for all of them. Each country's range should start from that country's market data, in that country's currency.

"Compare base salary to base salary"

What it looks like: The benchmark compares annual base salary in each country with a base salary figure from each country's data, as it would in the United States.

Why smart teams fall into it: Base salary is the number on the offer letter and the number most surveys lead with. It seems like the cleanest like-for-like figure.

What it costs: "Base salary" does not mean the same thing everywhere. In the Philippines, employers must pay rank-and-file employees a 13th-month pay by 24 December each year [3]. In several European and Latin American countries, annual pay is customarily split into 13 or 14 payments, so a monthly figure multiplied by 12 understates the year. In India, a salary is often quoted as a total cost to the company that includes basic pay, a list of allowances and employer contributions. Since India's labor codes took effect on 21 November 2025 [4], excluded allowances above half of total pay are counted back into wages, which raises the base for provident fund and gratuity [5]. The same total package can now cost the employer more than it did a year ago. The comparable figure is guaranteed annual cash, built the same way in every country. Check how each survey defines it and build your own figures to match. Variable pay, benefits and employer costs are then compared separately, each on its own basis. The EU directive defines pay to include complementary and variable components, in cash or in kind, so the narrow figure will not be the one that counts there either [6].

"Move every country by the same budget"

What it looks like: Finance sets one salary increase budget of, say, 3.5% and applies it in every country. The pay ranges move by the same amount everywhere, or not at all.

Why smart teams fall into it: One budget is easy to approve and easy to explain. It feels fair to give every country the same percentage.

What it costs: Markets do not move together. On the 2027 projections above, a 3.5% budget is generous in Switzerland and close to the market in the United States, the UK and Germany. It falls well behind in Poland and is about half of what Indian employers plan to spend [1]. Over two or three years, the Indian and Polish teams fall below their markets while the Swiss team drifts above it. The same applies to the data. A survey with an effective date in April and another with a date in January need to be moved to one common date before they are compared. Each country's data should be moved at that country's own rate, not at a single global one. Our note on how to age salary survey data covers the method. Budgets can be set by country and still be approved through one global process.

"Use whatever number we can find for the small countries"

What it looks like: The company has good data for its main countries. For the two people in Portugal or the one person in Colombia it uses a recruiter's suggestion, an employer of record's guide or a free salary site.

Why smart teams fall into it: Buying a survey for one or two people seems out of proportion. Any number feels better than none.

What it costs: Thin data looks as precise as good data once it is in a spreadsheet. Even large surveys are thin in smaller markets. The salary budget survey above marks results from fewer than 30 organizations as a small sample and does not show results from fewer than five at all [1]. For a specific job at a specific level, the number of matches is smaller still. A recruiter's figure is a starting point for a negotiation. A free site's figure may mix job levels and company sizes. Our note on what free salary sites are good for covers where they mislead. The sounder approach is to say how strong each figure is. Use the best source available for the country, check it against a well-covered nearby market where that makes sense, mark the figure as low confidence and review it sooner than the rest.

"Let each country decide for itself"

What it looks like: Local managers know their markets, so each country picks its own data, its own target and its own way of mapping jobs. Head office signs off the total budget.

Why smart teams fall into it: It respects local knowledge and avoids head office imposing rules it does not understand. It is often how things started, with one trusted manager in each country.

What it costs: The company loses the ability to compare anything. A "senior engineer" in one country may match a mid-level role in another. One country targets the market median and another the 75th percentile, without anyone deciding that on purpose. Exceptions are approved in one place and refused in another. When someone transfers between countries or a manager leads a team across three of them, every pay conversation becomes a negotiation. Local knowledge matters for the market data and for local law. The rules about what the company pays for belong to the company. Our guide on which market percentile to target covers why that target should be chosen once and on purpose.

What sound cross-country benchmarking looks like at 50 to 500 people

A good method is not the most expensive one. It is one the company can explain to its board, its auditors and an employee in any country who asks how their pay was set. Several qualities set it apart.

One job framework, matched in every country. Each role has one level and one job family wherever the person works. Matching to each country's survey starts from that framework, not from local titles.

Local data, local currency, local definitions. Each country's market figure comes from data for that country, in its currency. It is built on the same definition of guaranteed annual cash as every other country.

A stated target against each market. The company says where it wants to sit in each market, such as the median. If it chooses to pay above the market in one country, that is a written decision with a reason.

One date, each country's own rate. All data is moved to the same effective date. Each country's data moves at its own rate.

Confidence is labeled. Every figure carries a note on the source and how many matches stand behind it. Thin figures are flagged and reviewed sooner.

Pay is compared on pay, budgets on cost. Market comparisons use pay. Budgets use the full employer cost, which includes social charges and statutory payments that differ by country.

Global rules, local review. One process approves ranges, raises and exceptions. Local managers and local payroll or legal advisers check that each country's figures and pay practices fit local law and custom.

The calls only you can make

Benchmarking can be done to a method. The choices inside it belong to the company.

Which market is the market? For most roles, the market is the country where the person works. For a small number of senior or scarce roles, the company may compete for people across a region or globally. Deciding which roles those are and paying them differently is a strategy call.

The same target everywhere? Paying at the median in every country is simple and consistent. Some companies choose to pay above the local market where they hire their hardest roles, or where they are less known as an employer. Either can be right. It should be chosen, not inherited.

Which currency to pay in? Employees usually expect to be paid in local currency. Some contractors and senior hires ask to be paid in dollars or euros. The company has to decide who carries the currency risk and say so in writing.

What happens when someone moves country? A move from Germany to Poland, or from India to the UK, is a move between markets. The company needs a rule before the first request arrives. Our guide on location-based pay covers the options.

How much to share and where? Pay transparency laws differ by country and are changing. In the EU, applicants have the right to the initial pay or its range for the position, based on objective, gender-neutral criteria [6]. Member states were due to bring the directive into national law by 7 June 2026 [6]. Our note on EU pay transparency, country by country tracks where each one stands. A company with staff in several countries has to decide whether to share the same information everywhere or only where the law requires it.

The numbers and the law

  • Salary budgets differ by country. A compensation professionals' association's survey of 1,799 organizations, closed in May 2026 and published in July 2026, shows projected 2027 average total salary increase budgets of 2.7% in Switzerland, 3.6% in the United States and Germany, 3.7% in the UK, 4.9% in Brazil, 5.0% in Mexico, 5.5% in Poland and 9.2% in India. Results from fewer than 30 organizations are marked as small samples and results from fewer than five are not shown [1].
  • Currencies move on their own schedule. The Federal Reserve's daily rate was about 87.85 Indian rupees per US dollar on 23 October 2025 and about 94.94 on 23 October 2026, a change of about 8% [2].
  • Philippines, 13th-month pay. Presidential Decree 851, as amended in 1986, requires employers to pay all rank-and-file employees a 13th-month pay not later than 24 December each year, with some exemptions [3].
  • India, labor codes. The Code on Wages, 2019, the Industrial Relations Code, 2020, the Code on Social Security, 2020 and the Occupational Safety, Health and Working Conditions Code, 2020 took effect on 21 November 2025 [4]. Under the common definition of wages, excluded components such as house rent and conveyance allowances cannot exceed 50% of total pay. Any excess is counted back into wages, which affects provident fund contributions and gratuity [5]. Rules under the codes were still being finalized during 2026, so check the current position with local advisers.
  • European Union, what counts as pay. The pay transparency directive defines pay as the ordinary basic or minimum wage or salary and any other consideration, in cash or in kind, that a worker receives directly or indirectly from the employer. Applicants have the right to the initial pay or its range for the position. Member states were due to transpose the directive by 7 June 2026 [6].

This is general information, not legal or tax advice. Employment, payroll and tax rules differ by country and change often. Have local advisers check pay structures, statutory payments and pay disclosures in each country where you employ people.

Our view

We think most companies with staff in several countries should price locally and decide globally. Every role should be benchmarked against data for the country where the person works, in local currency and on a definition of pay that is built the same way everywhere. The decisions about levels, targets and exceptions should be made once, for the whole company.

We would not convert a home-country range or apply a country factor, even as a stopgap. It is quick, but it sets pay by the exchange rate and builds in differences that nobody chose. The first currency swing or the first transfer between countries exposes it.

And we would be honest about the data. Most companies this size have good data for two or three countries and thin data for the rest. A figure labeled "low confidence, review in six months" is more useful than one that looks precise and is not.

Sources

  1. WorldatWork, 2026-2027 Salary Budget Survey: Top-level results (published 15 July 2026; 1,799 organizations, survey closed May 2026), checked 2026-10-28
  2. Board of Governors of the Federal Reserve System, Indian rupees to one US dollar spot exchange rate (FRED series DEXINUS, updated 26 October 2026), checked 2026-10-28
  3. Philippines, Presidential Decree No. 851 (1975) as amended by Memorandum Order No. 28 (1986), 13th-month pay (LawPhil), checked 2026-10-28
  4. Press Information Bureau, Government of India, Four Labour Codes herald transformational change (21 November 2025), checked 2026-10-28
  5. Kathuria and Ramteke, India's new labour codes: A concise overview of changes, implementation and impact, Bar & Bench (25 March 2026), checked 2026-10-28
  6. Directive (EU) 2023/970 on pay transparency, Articles 3(1)(a), 5(1) and 34(1) (EUR-Lex), checked 2026-10-28

Which salary survey should you buy?

Coverage of your own jobs beats a brand name. Five ways companies choose the wrong salary survey and what a good choice looks like.

Which salary data should you trust?

Salary surveys, live pay data and job ads answer different questions. Five ways teams misread their market data and what a sound source policy looks like.

Pricing a role with no market match

Blend matches openly and never hide a weak one. Five traps when no survey job fits a role and what a sound price for a new or hybrid role looks like.

See Resources for every guide and the self-check.