Salary benchmarking from job ads: how to read the market rate
A practical salary benchmarking method using public job ads: match the role, record pay on one basis, and report the median with its sample size.
Free app. Runs on your Mac with the Claude or Codex plan you already have.
What salary benchmarking is
Salary benchmarking is the work of comparing the pay for a role with what other employers pay or offer for the same work, in the same place, at the same time. Done well, it gives a range you can defend, with the number of data points behind it. Done badly, it is one figure from a website, for a job title that only looks like yours.
Pay transparency laws in a growing number of places now require employers to publish a salary range in their job postings, which makes advertised pay a current, public source. The method below is how a compensation analyst builds a benchmark from those ads by hand. Each step says what to do and why, then how 10folds carries it out.
What it does for the business
- Make offers candidates acceptKnowing what competing employers advertise for the same role and location helps you set an offer that is competitive before the first refusal, instead of after it.
- Set ranges you can explainA benchmark with sources and a sample size gives managers and finance a reason for each range. That makes pay decisions easier to defend internally and, where required, publicly. It also keeps people in the same role on a consistent basis.
- See the market as it is nowJob ads reflect what employers offer this month. They complement salary surveys, which are often published months after their data was collected.
- Spot roles drifting out of lineRepeating the benchmark on the same sources shows when advertised pay for a role moves, so you can review salaries before people start leaving.
Salary benchmarking from job ads, step by step
Define the role and the decision
A senior and a junior ad with the same title are two different jobs.
What to do
- Write a short job match description: main responsibilities, level, team size managed and required experience.
- Fix the location or remote scope, and the date of the benchmark.
- State the decision it supports: a new hire, a promotion, a range review.
- Decide which employers count as your talent market: those you hire from and lose people to.
Why
Job matching is the heart of salary benchmarking. Titles vary between companies, so matching on duties and level prevents comparing your mid-level role with someone else's lead role. Fixing location and date matters because pay differs sharply between cities and moves over time. Naming the decision up front also tells you how precise the benchmark needs to be: a single hire tolerates a rougher figure than a company-wide range review.
Choose sources and collect the job postings
A search snippet is a lead; the ad itself is the evidence.
What to do
- Collect job postings from the employers in your talent market, from their own careers pages where possible.
- Add job board listings for the same role and location.
- Open each ad and read the pay statement in the full text.
- Save the address and the date you read each ad.
- Note ads that state no pay; they are part of the picture.
Why
Careers pages are usually the most accurate version of an ad, while job boards widen the sample. The share of ads stating no pay tells you how much of the market is visible. Ads also disappear quickly, so a saved address and date are needed to check a figure later.
Record advertised pay on a comparable basis
Convert nothing without stating the rule.
What to do
- Record the salary range minimum and maximum, the currency and the pay period: hourly, monthly or annual.
- Note whether bonus, commission or equity is included or stated separately.
- Record the level and location each ad states.
- Leave pay empty when the ad gives none, or gives only one bound.
Why
A range quoted per hour and one quoted per year look wildly different until you notice the period. Converting hourly to annual needs an assumption about hours, and including a bonus inflates the figure. Recording the basis first lets you decide later, openly, how to treat each case. Bonus and equity are the parts most often described loosely, so note the wording the ad uses.
Group comparable ads and compute the figures
Six ads are a signal, not a benchmark.
What to do
- Group ads by matched role, location, currency and pay period.
- For each ad, take the range midpoint: the average of minimum and maximum.
- Report the median of those midpoints, the lowest minimum and the highest maximum.
- Show the number of ads in every group beside its figures.
- Remove duplicates of the same ad posted on several sites.
Why
The median salary is less distorted by one unusually high or low ad than the average. The range midpoint is a fair single figure for a posted range, though many employers hire below it. The count of ads tells the reader how far to trust the result.
Interpret the result and set the range
A benchmark is an input to a pay decision, never the decision itself.
What to do
- Check the median against at least one other source, such as a compensation survey.
- Decide where you want to sit: at, above or below the market rate, and why.
- Adjust for what ads do not show: benefits, equity, flexibility.
- Repeat the benchmark on the same sources later to see movement.
Why
Posted ranges can be wide, and employers may advertise one figure and pay another. Combining ads with survey data and your own pay structure gives a range that holds up. A repeat run on the same sources is the only fair way to talk about change. Write down the position you chose against the market, so the next review can test whether it still holds.
Common mistakes
- Matching on title aloneRead the requirements to confirm the level matches.
- Mixing pay periodsHourly and annual figures cannot share a median without a stated conversion.
- Hiding the sample sizeA median of four ads needs to say so.
- Treating the posted range as actual payOffers often land below the top of the range.
- Counting the same ad twiceOne job posted on three boards is one data point.
What a good result looks like
- A clear job match description and location.
- One row per ad with minimum, maximum, currency, period and source.
- The median midpoint and range per comparable group.
- The number of ads behind each figure, and how many ads stated no pay.
- A note of sources that could not be read.
And then what? Using the result
- Compare the median with your current salary range for the role.
- Check one or two figures against a compensation survey.
- Agree with the hiring manager where the offer should sit.
- Run the same benchmark again in a quarter to see movement.
When to run it
- Before opening a new roleA benchmark tells you whether the budget will attract the people you want.
- When candidates keep turning down offersComparing your offer with advertised pay shows quickly whether pay is the reason.
- During an annual pay reviewCurrent ads add a fresh view of the market to survey data that may be months old.
- When you must publish a salary rangeUnder pay transparency rules, a sourced benchmark helps you set a range you can stand behind.
Questions people ask
What is salary benchmarking?
Salary benchmarking is comparing the pay for a role with what other employers pay or offer for comparable work in the same market. It is used to set salary ranges, make offers and review pay.
How do you benchmark salaries?
Write a job match description, collect pay data for matching roles in the same location from ads or surveys, put it on one basis, and report the median and range with the number of data points.
Are salary ranges in job postings accurate?
They show what the employer is willing to offer, which is useful, but they are not proof of what people are paid. Some ranges are very wide, and many hires are made below the top. Read them together with the job's level and location.
How many data points do you need to benchmark a salary?
There is no fixed minimum, but a handful of ads is a signal and a larger set from several employers is more reliable. Always show the count so readers can judge.
Does 10folds convert hourly pay to annual, or one currency to another?
No. It groups ads by currency and pay period and keeps them separate. Any conversion is left to you, with your own rule.
Is this a full market survey?
No. It reads the public sources you name, up to 25. Many ads state no pay, and those stay unknown.
More jobs in Research
Benchmark advertised salaries, on your own Mac.Free app. Runs on your Mac with the Claude or Codex plan you already have. Your files stay on your computer.
Download for Mac