Guide

Remote jobs with a salary listed: what 132,942 postings show about pay

Last updated: November 2026. Every figure below was measured from the RemoShift index on 12 November 2026.

Live remote roles
132,942 at 18,992 companies
Publish a figure
50,070 (37.7%)
Ranges, not numbers
47,246 of 49,987 (94.5%)
Best market
US 57.0%, Canada 51.5%
Worst market
India 9.4%, Brazil 11.5%

Pay transparency is usually discussed as a matter of employer virtue: firms that publish a range are called open, firms that do not are called evasive. The data does not support that reading. On 12 November 2026 our index held 132,942 live remote listings from 18,992 companies, and whether a listing carries a number is predicted far better by the country it is open to, the currency it is quoted in and the seniority in its title than by anything about the company's culture.

50,070 listings published a pay figure โ€” 37.7%. That is the headline, and it is the least useful number in this article. The rest of it is about where the other 62% sit, what the published numbers actually are, and how to read one before you apply.

37.7% publish a number, and 94% of those are ranges

Of the 50,070 listings that state pay, 49,989 carry a readable figure and 47,246 of them (94.5%) are ranges rather than single values. A typical posting does not say what the job pays; it says what band the job sits in. Across 35,814 annual ranges wide enough to compare, the median ceiling is 1.33 times the median floor.

That ratio is the single most practical number here. A range of 100,000 to 133,000 is the norm, not an outlier, and the ceiling is not a promise โ€” it is the top of a band that also contains people with five more years of experience than you. Treat the floor as the offer you will probably be made and the ceiling as the argument you can have.

The units are messier than the numbers. Of the 49,989 readable strings, 29,423 state an annual figure explicitly, 9,571 state no period at all, 8,167 quote an hourly rate and 2,584 quote a monthly figure. A listing that says 90,000 without a period is annual; a listing that says 90,000 and means rupees a month is not. The currency mix is heavily dollar-denominated: 44,384 of 49,989 (88.8%) are in USD, against 1,626 EUR, 1,303 CAD, 960 GBP, 320 PHP and 188 PLN.

The transparency map: 57% in the US, 9% in India

Disclosure rates by country eligibility, over listings that name at least one country:

Read against market size, the ranking is uncomfortable. Germany is the fourth-largest market in the index by listings (5,085) and publishes pay in 14.7% of them โ€” below Spain (17.8%), Portugal (17.1%) and France (15.4%), and above only Poland and Romania among the eight European markets with more than a thousand listings. India is the fourth-largest market overall (5,173) and the least transparent of the eighteen countries with more than a thousand listings. A candidate searching from Berlin or Bengaluru is working in a market that publishes pay roughly three to six times less often than the one they are competing against for the same remote roles.

One counter-intuitive result: listings open anywhere, rather than to a named country, disclose slightly more often (39.8% of 2,649), not less. Global hiring tends to mean a company that has already decided how it prices a role across geographies.

The currency on the listing matters more than the country

This is the finding worth carrying into your search. Compare the same country's listings that quote dollars against those that quote the local currency, annual figures only, on 12 November 2026:

These are not the same job priced two ways; they are two different employer populations. The dollar-quoted group is mostly US companies hiring internationally, or remote-first firms that set one global band. The local-currency group is mostly domestic employers paying a local market rate. But the pattern repeats in every market where both exist, and it has a direct consequence: on a remote board, the currency in the salary field tells you more about the offer than the country in the location field.

It also reframes the geography question. A UK-based candidate scanning "remote jobs in the UK" sees two markets stacked on top of each other, one paying US bands in dollars and one paying UK bands in pounds, and the filter that separates them is a line of text most people skip.

Read the unit before you read the number

The monthly and hourly quotes are where honest listings get misread as bad ones. Measured medians by currency and unit:

Hourly work is a sixth of the disclosed market and a different market, not a cheaper version of the salaried one. It covers part-time support, task-based AI training contracts, tutoring and interim project work. Compare it to a salary only after annualising it honestly โ€” hours per week, weeks per year, no paid leave, no benefits โ€” which is what our salary negotiation guide walks through.

Which functions publish, and which do not

Disclosure by category, over listings our index has labelled:

The pattern is not "technical roles hide pay". It is that functions with an internal grading structure publish it, and functions where pay is set per hire do not. Sales, the second-largest category, publishes at 46.4% because commission structures force a number into the posting.

Entry-level postings are the least transparent of all

Grouped by the seniority language in the title: junior, entry-level and internship titles publish a figure in 27.5% of 2,842 listings; senior, lead, principal and staff titles in 40.0% of 32,258; manager, director and head-of titles in 42.3% of 32,482.

The gradient runs the wrong way for the people who need it most. A senior engineer negotiating a 150,000 band has leverage, alternatives and a market they already know. A candidate applying for their first remote role is 12.5 percentage points less likely to be told what it pays, in exactly the situation where not knowing costs them the most. If you are early-career, assume the number is negotiable precisely because it was never published, and ask in the first screening call rather than the offer call.

It is an employer decision, not an industry rule

The spread between individual employers is far wider than between functions or countries. Among companies with at least 100 live remote listings on 12 November 2026:

Two employers in the same industry, hiring the same roles, can sit at 95% and 0%. That matters for how you read an absent number: it is not a signal that the company is hiding something unusual, it is a signal about that company's policy, and the policy is knowable before you apply by looking at its other postings on the same board.

How to use a disclosed range

Methodology

All figures measured on 12 November 2026 against the live RemoShift index, filtered to listings with status = active and is_remote = remote: 132,942 listings from 18,992 companies. "Publishes a figure" means the source posting supplied a salary field containing at least one number; nothing is inferred or estimated, and 53 listings whose salary field held only text such as "competitive" are counted as not disclosing. Salary bands are medians of the parsed floor and ceiling of annual-quoted figures, excluding quotes below 15,000 or above 900,000 to remove unit errors; monthly and hourly figures are reported separately and never mixed into an annual median. Country counts sum to more than the listing total because one listing can name several eligible countries (158,419 country tags across 121,724 listings), and disclosure rates are computed per country tag. Category, seniority and employer groupings are derived from the listing title and our own category labels.

For the broader picture of what remote employers are hiring for and where, the State of Remote Hiring report covers the full distribution. If you are choosing a market rather than reading one, start with working remotely for a US company โ€” the largest and most transparent cohort in this data โ€” or the finance and healthcare role hubs, which publish pay more often than any other function.