The Bid & Proposal Salary Index is built from the salaries employers and agencies publish in job adverts for bid, proposal and tender roles. It isn't based on payroll records or a survey, and it won't tell you what one particular person ought to be paid. What it does give you is a clear, regularly refreshed view of what the market is advertising, role by role and country by country.
This page explains how we get from a job advert to a published figure, what the confidence labels mean, and where the data has its limits.
Where the data comes from
We draw on licensed job advert data covering more than 50 countries. Advert sources don't capture every job in every market, so please treat the index as a well-evidenced indicator of advertised pay rather than a complete picture of worldwide salaries.
Both direct employer and recruitment agency adverts are included.
How adverts become figures
Reading the adverts
Salaries aren't always set out neatly. Some sit in a dedicated pay field and others are buried in the body of the advert. We use AI to help read and classify adverts, for example to pick out a salary written in the text, group the benefits on offer, or fill in a missing country or posting date.
AI can get things wrong. Every salary it extracts goes through a series of automated checks before it's used, and we remove duplicate adverts. These checks cut errors down a great deal, but no process like this is perfect.
We leave out internships, apprenticeships and work-student roles, figures that are placeholders rather than pay (such as '$1 per hour'), and any salary below a tenth of the typical figure for bid roles in its country.
We don't keep the advert wording once a salary has been checked. Both the original text and any extracted quote are deleted.
Working out the currency
Where an advert doesn't state its currency plainly, we work it out from the evidence in the advert and its context, giving more weight to the most reliable signals. If we can't establish the currency with reasonable confidence, the salary is left out.
Converting between currencies
Where every advert in a sample uses the same currency, we calculate the median in that currency first and only then convert it. This stops exchange rates distorting the result. Where a sample mixes currencies, we convert to the currency used most often in that sample. We use fixed reference exchange rates, which we update periodically, and we don't convert currencies we hold no reliable rate for.
So a country's figure can include a few adverts paid in another currency: an advert for a role in India paid in Canadian dollars is converted into rupees and counted with the rest. Where a figure includes any, the page says so beside it, naming the currency and the date of the exchange rates used.
Contract day rates
Day rates are reported separately and are never turned into annual salaries. Converting one into the other means guessing at working days, holidays, tax status and employment costs, and those guesses would make the numbers less trustworthy.
What we publish
Advertised base salary
All figures are medians of advertised base salary. We don't add estimated bonuses, pension contributions or the value of benefits.
Confidence labels
Every usable salary counts, even where only one advert is available. To make clear how much weight a figure can bear, each one carries a label:
- Strong: a large, varied sample and a tight range around the median
- Fair: a reasonable sample with a moderate range
- Indicative: a usable figure, but the range is wide, so read it as a guide
- Low data: too few salaries, too few employers, or one employer accounts for much of the sample
The labels come from a statistical range calculated around each median, alongside our own checks on sample size and how many different employers it includes.
The spread of salaries is always drawn the same way, as the number of adverts in each fixed salary band, however few adverts there are. A small sample is marked by its label, not by a different kind of chart.
How many adverts state pay
We also show the share of adverts that state a usable salary. It is measured over the same permanent adverts as the figures on the page, for the same role, country, experience level and period, whether or not they list pay. So the number stating pay is the figure's own sample size, and the total is the count the benefits section uses. Contract and part-time adverts are not included; contract day rates are reported separately. Below 100 adverts or 10 employers the share is still shown, marked Low data. At the time of writing, roughly a third of the bid and proposal adverts we see worldwide include a salary.
Benefits
AI groups the benefits mentioned in adverts into categories. It can occasionally miss a benefit or file it under the wrong heading. Benefit percentages only cover adverts that mention benefits.
Pay trends and inflation
Trends follow the median advertised salary month by month. Months with only a handful of salaries are shown less prominently and left out of trend calculations, and the current month is excluded until it's complete. Where we only have a short run of data, the result is labelled Short history and we show the total change rather than an annual rate.
Every trend starts in August 2026, the first month our daily collection covered in full. The months before it hold only a handful of adverts each, collected a different way, so a line drawn through them would measure our collection rather than the market. Adverts posted earlier still count in current figures.
To show how advertised pay has moved once inflation is taken into account, we use the standard real-terms adjustment:
(1 + pay change) ÷ (1 + inflation) − 1
Each country has its own price index, so there's no single worldwide inflation comparison.
Inflation data: International Monetary Fund, Consumer Price Index (CPI), data.imf.org. The IMF does not endorse this index or our calculations.
Updates and why figures move
The index refreshes daily. Each figure shows two dates: when it was last calculated, and the posting dates of the adverts behind it.
Adverts sometimes reach us after their original posting date, so recent figures can shift a little as more data arrives. Some markets are refreshed more often than others.
Limits worth knowing about
- No cost-of-living or tax adjustment. A higher advertised salary in one country doesn't necessarily mean more spending power.
- Pegged currencies. Several Gulf currencies are tied to the US dollar, so a Gulf salary shown in pounds will mostly move with the pound-to-dollar rate.
- Earlier data. Adverts posted before August 2026 are used in current figures but not in trends, and early advert counts should be read as minimums. The date range shown with each figure starts at 1 August 2026, even where an earlier advert is included.
- Uneven coverage. Smaller markets have smaller samples, which the confidence labels reflect.
How this compares with our Salary Survey
Bid Solutions also publishes the UK Permanent Salary & Consultant Day Rate Survey. The 2025 edition reports a mean of £68,940 and a median of £65,000 from 1,422 respondents, with Bid Managers at £58,886.
The index will usually come in lower, for a few reasons. Advertised salaries and range midpoints can sit below final offers. Survey respondents tend to be more engaged with the profession, and often more senior. And the survey leads with its mean, whereas the index uses the median. The two are designed to be read together.