The short answer
There is no single reliable number for what a job pays, and any site that gives you one confidently is guessing. The way to build a defensible figure is to triangulate: read advertised ranges in your own market, use published pay scales where they exist, ask people who do the work, and adjust for level and location yourself.
What matters most
- Advertised ranges in your own market and level are the most current evidence available to you.
- Public-sector and government pay scales are genuinely published and are useful reference points even for private-sector roles.
- Job titles are unreliable. Compare the described work, not the label.
- Triangulate at least three sources and build a range, not a single figure.
Why we do not publish salary figures
It would be easy to fill this site with average-salary tables, and every competitor does. We do not, for a straightforward reason: we have no verified data source we could stand behind, and a number presented with false confidence is worse than no number at all. Pay varies by location, sector, company size, seniority and the specific responsibilities of a role, and a national average collapses all of that into a figure that is wrong for almost everyone who reads it.
The failure is asymmetric, which is what makes it serious. If a table understates the market and you anchor on it, you may accept less than you could have had for the entire time you stay in the job, and that difference compounds into every future rise. So instead of giving you a number, this guide gives you the method for building one, using sources you can verify and cite in the conversation itself.
- Aggregator averages. Frequently based on small samples, self-reported, undated, and blending several seniorities under one title. Useful only as a very rough sanity check.
- Old data. Pay in a fast-moving specialism can shift substantially within a year. Any figure without a date attached is not evidence.
- National figures for a local decision. The same role can differ enormously between two cities in the same country. National means nothing to your negotiation.
- Title-based comparison. An "analyst" at one company is a graduate role and at another runs a team of six. Compare the work described, never the label.
The sources worth using
- Advertised ranges
- The most current evidence you can get, and in a growing number of jurisdictions employers must publish them. Collect postings for roles genuinely comparable to yours, in your location, and record the range with the date and the actual responsibilities listed. Twenty of these is a dataset you built and can defend.
- Published public pay scales
- Government, healthcare and education pay structures are published in most countries, and they are real, verifiable numbers. Even if you are heading into the private sector they anchor the general shape of a career ladder in your field.
- People who do the work
- Former colleagues, people you trained with, professional community members. Ask about ranges and bands rather than personal figures, which is a much easier question for people to answer. This is the only source that reflects what is actually being paid rather than advertised.
- Recruiters in your specialism
- A specialist recruiter who places your kind of role sees actual offers weekly. They have an interest in the numbers they quote, so treat it as one input, but they are closer to the market than any website.
- Professional bodies and unions
- Where they exist for your field, they often publish member surveys or negotiated scales. Sector-specific and usually far better grounded than general aggregators.
Building your range
- 1Define the job, not the title
Write five lines describing what you actually do: scope, people, budget, systems, decisions you own. This is what you will compare against, because titles are noise.
- 2Collect fifteen to twenty live postings
In your location or the location of the job, for work that genuinely matches your five lines. Record the stated range, the date, the company size and the responsibilities. Where no range is stated, note the posting anyway — the description is still evidence about level.
- 3Discard the outliers and read the middle
You are looking for where the bulk of comparable postings cluster, not the highest number you can find. One unusually high posting is a different job with the same title.
- 4Adjust for the things postings do not say
Company size, sector, whether the role is remote or requires a commute, and how the total package differs. A lower base at an employer with a strong retirement contribution can be the better-paid job.
- 5Sense-check it with a person
One conversation with someone doing the work will tell you whether your range is plausible faster than another ten hours of reading. Ask about bands, not their personal salary.
- 6Write down the range and the sources
A range with reasoning attached is what you take into a negotiation. "Comparable roles in this region are advertising in this band" is a sentence you can say with confidence, because you built it.
Reading a posted range properly
Pay transparency rules in a number of US states and, increasingly, across the EU mean more postings now carry a range. That is genuinely useful, and it needs to be read with some care, because a posted range is not a promise about you.
| What you see | What it usually means |
|---|---|
| A wide range spanning several levels | The posting probably covers more than one seniority. Work out which end your experience maps to before anchoring on the top. |
| A narrow, specific range | A well-defined band. Where you land within it is usually about experience relative to the requirements, and it is reasonable to ask how that is assessed. |
| A range with "depending on experience" | Standard language. Ask directly what would place a candidate at the upper end — most employers will tell you, and the answer is a negotiation roadmap. |
| No range at all | Either the jurisdiction does not require one or the employer prefers not to say. Asking early saves everyone time and is a completely normal question. |
Using what you found
Your research does two jobs. It sets your own three numbers before an offer conversation, and it gives you a sentence that makes an ask credible instead of arbitrary. That sentence should be short and calm — a statement about the market, not a demand. Nobody needs to see your spreadsheet, but you need to have built it, because the confidence in how you say the number comes from knowing where it came from.
From here, how to negotiate a salary offer covers the conversation itself, and the offer comparison tool handles the part where two packages have to be compared on more than base. If your research suggests you are already underpaid where you are, asking for a raise is the other place this work pays off.
Questions people actually ask
Why does this site not publish salary figures?
Because we have no verified data source we could stand behind, and a confident-looking average is worse than no number. Pay varies too much by location, sector, company size and actual responsibilities for a national figure to be useful, and anchoring on a low one can cost you money for years. We teach the method for building your own instead.
Are salary comparison websites accurate?
Treat them as a very rough sanity check rather than evidence. Many are built on small self-reported samples, blend several seniorities under one job title, and carry no date. Advertised ranges for genuinely comparable roles in your own location are far better evidence and are available to anyone.
How do I research pay for a city I do not live in?
Use postings for that specific location rather than national data, and compare what is left after housing and any local taxes rather than the gross figures. Published public-sector scales that include locality adjustments are useful for seeing how much a role is expected to shift between regions.
How many sources should I use?
At least three of different kinds — advertised ranges you collected yourself, a published scale or professional body figure, and at least one person who does the work. Agreement between different types of source is what makes a range defensible; twenty results from the same website is one source, not twenty.
What if the posted range is much lower than I expected?
Check first that you are comparing the work rather than the title, and that you are looking at your location and level. If the range genuinely is the market, the useful questions are whether the total package changes the picture, and whether a different sector or company size pays differently for the same skills.