Analysis · October 26, 2026 · data as of October 26, 2026

Why 77% of listings carry no employment type

Employment type is blank on 76.6% of Hired Creative's listings, it is filled or blank by source rather than by job, and the 9.2% contract share among filled rows is not a market figure.

In the Hired Creative listings export read on 2026-10-26 (20,556 rows, first seen 2026-06-03 to 2026-10-25), the employment type field, job_type, is filled on 4,800 rows (23.4%) and blank on 15,756 (76.6%). Whether it is filled depends on the board a listing came from, not on the job. All 9,737 LinkedIn rows and all 2,824 Greenhouse rows in this export are blank. The 9.2% contract share among filled rows comes mainly from a few boards and runs from 0.0% to 20.9% by board. It does not measure the creative job market, and freelance cannot be measured from this field at all.

Key points

• Employment type is blank on 76.6% of the 20,556 listings in the Hired Creative export read on 2026-10-26.
• Whether a listing has it depends on the board it came from: all 9,737 LinkedIn rows and all 2,824 Greenhouse rows in the export are blank.
• Among the 4,800 filled listings, contract is 9.2%, from 0.0% to 20.9% by board, and three boards supply 61.9% of them. It is not a market figure.
• No listing carries a freelance value, so freelance work cannot be measured from this field.

What is blank, and where

The field holds four values: full_time (4,044 rows), contract (443), part_time (217) and internship (96). The value freelance appears on 0 rows.

The export draws on 159 sources. 23 fill the field on every row (4,306 rows), 129 fill it on none (15,682 rows) and 7 fill it on some rows (568 rows). So 19,988 of 20,556 rows, 97.2%, sit in a source that is always filled or always blank. LinkedIn feeds are 9,737 rows (47.4% of the export) and Greenhouse is 2,824, none filled in either. Together they are 79.7% of all blank rows. These are facts about this export; the data do not show why a source leaves the field blank.

Coverage by region group follows the LinkedIn share of each group. Filled share against LinkedIn share of rows: Americas 32.0% and 34.8%, Europe 31.4% and 35.2%, APAC 9.0% and 73.9%, MENA 8.4% and 83.3%, Africa 2.7% and 93.3%. Ranking the five groups by LinkedIn share gives the reverse of ranking them by coverage. Because the LinkedIn feeds fill nothing anywhere, the regional gap is a source-mix effect, not evidence that jobs are described differently by region.

What the filled rows show

Among the 4,800 filled rows, 84.2% are full_time, 9.2% contract, 4.5% part_time and 2.0% internship. Three sources supply 61.9% of them: remoterocketship (1,837), ashby (771) and smartrecruiters (362).

Across the 17 sources with at least 30 filled rows, the contract share runs from 0.0% to 20.9%: ashby 2.2% (n=771), remoterocketship 14.5% (n=1,837), himalayas 20.7% (n=164), recruitee 20.9% (n=115). Part-time runs from 0.0% to 50.0% across the same 17.

The pooled 9.2% is a mix of the rates of the boards that fill the field, and three boards supply 61.9% of those rows. It is not a contract share for creative jobs. A row bootstrap, which treats rows as independent, gives a 95% interval of 8.4% to 10.1%. A bootstrap that resamples whole sources gives 3.6% to 12.8%, and still covers only boards that report. The data do not split the spread into a source part and a job part.

Within one board, motion and writing listings are labelled contract more often than design listings. In remoterocketship, contract is 30.2% of motion (n=96) and 28.2% of writing (n=156) against 11.9% of design (n=1,216). In ashby, motion is 15.0% (n=40) against 1.7% for design (n=235). Hired Creative does not rank disciplines from this, and the pooled all-source figures by category mix different boards and are not comparable. Both readings sit inside single boards' own labelling.

Titles and labels

Of 256 labelled rows whose title contains contract, contractor, freelance, freelancer, fixed-term or temporary, 150 (58.6%) are labelled contract, 62 full_time, 43 part_time and 1 internship. For titles with intern, apprentice or trainee words, 77 of 141 (54.6%) are labelled internship. Titles and labels agree about 55 to 59% of the time where both exist, in this export. That may reflect each board's own conventions; Hired Creative does not call the labels wrong.

Titles also cannot fill the blanks. The same kind of title rule fires on 590 of the 15,756 blank rows (3.7%). The other 96.3% give no signal either way. A hand read of 25 of the 590 hits found that all 25 titles state a contract, temporary, freelance, fixed-term, internship or trainee term. That shows the title states a term, not the employer's own classification.

Outside documentation

Google's job posting structured data documentation lists the accepted employmentType values as FULL_TIME, PART_TIME, CONTRACTOR, TEMPORARY, INTERN, VOLUNTEER, PER_DIEM and OTHER. There is no freelance value, and employmentType is listed under recommended properties, not required ones. The Greenhouse Job Board API has no standard employment type field on a job; employers can expose custom fields in its metadata attribute. All 2,824 Greenhouse rows in the export are blank, and the data do not show why.

Method

Data: the Hired Creative listings export read on 2026-10-26, 20,556 rows, first-seen dates 2026-06-03 to 2026-10-25. A row is filled when its job_type value is not empty. Source is the collection feed recorded on each row; LinkedIn pools every feed whose name begins with linkedin. Region group and the nine-bucket category are the fields recorded in the export. Shares are rows with a value divided by rows in the group.

Source-level contract shares use sources with at least 30 filled rows (17 sources). Cells within a single source are reported only at 30 rows or more. Both bootstraps use 2,000 repetitions and report the 2.5th and 97.5th percentiles of the pooled contract share: one resamples the 4,800 filled rows, the other resamples the 17 sources.

Title rules are case-insensitive whole-word matches on the title. Contract: contract, contractor, freelance, freelancer, fixed-term, fixed term, temporary. Internship: intern, interns, internship, apprentice, apprenticeship, trainee, werkstudent, working student. Part-time: part-time, part time.

Countries with fewer than 30 listings, and any whose separate reporting could identify a sensitive location, are pooled.

Limits

• One snapshot (first seen 2026-06-03 to 2026-10-25), not a trend. New sources change the mix, so nothing here says employment type is shifting.
• The 9.2% contract share and the 84.2% full-time share describe only the 23.4% of rows from sources that fill the field, and 61.9% of those come from three sources. Neither is a share of creative jobs. The source interval of 3.6% to 12.8% rests on 17 sources.
• Freelance cannot be measured: no row has that value. This piece says nothing about freelance work.
• Region and discipline gaps among filled rows are confounded with source. Discipline cells inside one source are small outside remoterocketship (ashby and smartrecruiters have only a few cells of 30 or more rows). This piece publishes no region cut inside a single source and no ranking of regions or disciplines by contract share.
• The label disagrees with the title on 41% to 45% of title-flagged rows, so the filled field is not a clean ground truth either. A title may legitimately say freelance while a board classifies the role differently; the export cannot settle which.
• Why a source leaves the field blank is not tested. The data do not show whether the cause lies with the board or with how its listings are read, and this piece makes no claim about it for LinkedIn, Greenhouse or any other source.
• The title rules are crude keyword matches. A critic could say temporary and fixed-term are a different class from contract.

Sources

• Google Search Central: Job posting structured data, opened 2026-10-26.
• Greenhouse Job Board API documentation, opened 2026-10-26.
• Hired Creative listings export, 2026-10-26 (20,556 rows).

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