Analysis · October 13, 2026 · data as of October 12, 2026

Which feed a job came from predicts pay text far better than region

In 21,378 Hired Creative listings, the share showing pay runs from 0% to 80.6% by feed, and feed accounts for about five times as much of the variation as region does.

In Hired Creative's listings export (21,378 listings, read on 2026-10-12), 1,983 listings (9.3%) are marked by our pay field as showing pay. That rate runs from 0% to 80.6% depending on the feed a listing was collected from. In our listings, which feed a job came from accounts for about five times as much of the variation in whether it shows pay as the coarse region field does (52.6% against 10.3% of deviance explained; in-sample and descriptive). A regional pay table built from a mixed set of feeds like ours therefore measures the feed mix as much as the regions.

Key points

• In 21,378 Hired Creative listings read on 2026-10-12, the feed a listing came from explains about five times as much of whether it shows pay as its region does.
• Feeds differ from 0% to 80.6%: Reed shows pay on 258 of 320 listings, while the six LinkedIn feeds together show it on 15 of 10,008 (0.1%).
• The very low rates in the Asia-Pacific, Africa and MENA groups are largely a feed effect. The Americas-to-Europe gap (16.6% against 8.0%) is not.
• A regional pay table built from mixed feeds shows the feed mix as much as the regions. It does not show what employers in any region do.

What the data covers

Hired Creative reads job boards and applicant-tracking feeds, and each feed covers some regions far more than others. The pay field and the region field therefore both depend on which feeds are read. A gap between regions could be a source-mix effect, so this piece tests that before treating any gap as a market fact. The figures describe one export on one date. They do not show what employers publish, and they carry no trend claim.

The feed sets the rate

Three feeds supply 79.2% of all listings that show pay (1,570 of 1,983): Remote Rocketship, Ashby and Reed. Eleven of the 25 feeds with at least 100 listings show pay on none of them.

Public documentation shows one reason feeds can differ. Greenhouse's Job Board API returns pay ranges only when the caller sets pay_transparency=true on the single-job endpoint. Ashby's public posting API returns compensation only when the caller adds includeCompensation=true. The data do not show whether Hired Creative's collectors send these parameters, so a feed's rate cannot be assigned to the board, the employers or the collection method.

Region differences

Among the five region groups shown, raw rates run from 16.6% in the Americas to 0.6% in the MENA group.

Source mix explains part of this. One LinkedIn feed makes up 75% of the Asia-Pacific listings, 74% of the MENA group and 93% of the Africa group, and those feeds show almost no pay anywhere. If every listing showed pay at its own feed's overall rate, Asia-Pacific would sit at 3.1%, Africa at 1.3% and the MENA group at 2.0%, against 9.3% overall. That is most of their shortfall.

For the Americas the same calculation gives 12.6%. Source mix therefore accounts for up to about two thirds of the Americas-to-MENA gap (10.6 points expected against 16.0 raw). This is an upper-end estimate, because each feed's overall rate itself contains regional effects.

It does not explain Europe. Europe's expected rate is 12.9% against an actual 8.0%, while the Americas run above expectation (16.6% against 12.6%). The Americas-to-Europe gap is not a feed artefact in these data.

Inside one feed

Region still matters when the feed is held fixed. Remote Rocketship shows pay on 56.8% of its Americas listings and 19.2% of its European ones.

Ashby points the same way (74.9% of 630 Americas listings, 9.1% of 175 European ones), but a small number of employers account for much of that gap. Neither contrast shows why regions differ. Pay rules, employer type and the roles posted can all differ by region, and none is controlled here.

Same country, different boards

Listings whose location text names the UK or a UK city (1,590 listings) show pay on 10.9% overall. By feed the spread is wide: 77.4% of Reed's UK listings (123 of 159) and 0.0% of the main European LinkedIn feed (0 of 520).

Indeed's Hiring Lab put the share of UK postings with a salary at 56% in March 2026. Hired Creative's 10.9% is not comparable: the period, the population (creative roles in our feeds against all postings on Indeed) and the definition of pay all differ. The comparison is directional only. It does not show that the 45-point gap is purely a measurement artefact.

Same employer, different feed

Among 88 employers that appear on both kinds of feed, most of their listings on pay-carrying feeds show pay and almost none on the others. Of 65 employers with a pay-showing listing on a pay-carrying feed, 57 have none on the other kind. Separately, when the same company and job title appear on two feeds, 215 of 270 such pairs show pay on neither side, 13 on both and 42 on one side only. A title can be open in several cities, so a pair is not proof of one posting. Neither check says whether the board, the employer or the collector is responsible.

What this means for a regional pay table

Feeds differ from 0% to 80.6%, and no region-fair "pay transparency by region" number can be produced from this export. The usable comparison holds the feed fixed and shows region inside it, as in the Remote Rocketship chart. Few cells have enough listings to report.

Method

Data: the Hired Creative listings export of 21,378 listings, read on 2026-10-12. Each listing has a feed (the board or applicant-tracking feed it was collected from), a region group from the export's seven-group region field, a pay text flag, optional salary minimum and maximum fields, a company name and free-text location.

• A listing "shows pay" if its pay text flag is set or a salary minimum or maximum is present. The flag alone covers 1,447 listings (6.8%), the structured fields 1,333 (6.2%), and 797 have both. This is Hired Creative's own field, not checked against collector code.
• Feeds with at least 100 listings are reported separately (25 of 159). The rest are pooled into one group. All listings stay in every total.
• Countries with fewer than 30 listings, and any whose separate reporting could identify a sensitive location, are pooled.
• Within-feed region contrasts use only feed and region cells with at least 30 listings.
• Share of deviance explained: the reduction in deviance of the shows-pay outcome from grouping listings by feed (26 groups), by region (7 groups) or both (110 cells), against the overall rate. This McFadden-style measure rises mechanically with more groups, which favours feed over region somewhat. "About five times" is 52.6% divided by 10.3%.
• Expected rate by region: each listing is given its own feed's overall rate (all regions together), averaged within the region.
• Permutation tests (2,000 shuffles of region within feed, and of feed within region) were run. They treat listings as independent although pay clusters by employer, so this piece does not rely on them.
• UK listings: location text containing a UK term (United Kingdom, UK, England, Scotland, Wales) or one of seven UK city names. No other country was placed.
• Same employer: company names reduced to lower-case letters and digits, with at least one listing in a pay-carrying group (Ashby, Remote Rocketship, Pinpoint) and one in a group that rarely shows pay (Greenhouse, the main LinkedIn feed, Lever, SmartRecruiters, Teamtailor, Personio, Workable). Both groups were fixed from the observed feed rates.
• Cross-feed pairs: the same company and job title on two different feeds.

Limits

• Feed and region cannot be fully separated. Most feeds cover one region, and the LinkedIn feeds are one per region, so feed partly encodes geography. The region field has 7 groups and the feed field 26, so a finer country field might close some of the gap.
• The data do not show whether a low rate reflects the board hiding pay, the employer not publishing it, or Hired Creative's collectors not capturing it.
• The five-times figure is in-sample and descriptive, not causal.
• A critic would say that Greenhouse and LinkedIn carry large employers in places without pay laws, while Ashby and Remote Rocketship skew to start-ups and remote roles that post pay more often. That composition effect is not ruled out. The same-employer check helps but covers 88 employers matched on name text, with groups chosen from the outcome.
• The UK listings are placed by keywords and can include places outside the UK. The Indeed comparison crosses period, population and definition. The export mixes listings first seen between June and October 2026, so nothing here is a trend.

Sources

• Indeed Hiring Lab: Full salary transparency in Europe is still a distant prospect
• Greenhouse Job Board API documentation
• Ashby public job posting API documentation
• Hired Creative listings export, read 2026-10-12 (21,378 listings).

Published by Hired Creative from its own listings and the sources above. Spotted an error? Tell us.