"Remote" Restricts You to One Country in 95.7% of Postings, and That Barely Changes by Field

Bars scaled to 100%

Baseline is the figure across all remote postings with a known location. Every field sits within a few points of it.

"Restricted to one country" means the posting's resolved location maps to exactly one country code. Postings tagged remote but with no location text we could resolve to any country are excluded from every denominator here, so this measures the postings where we could actually check, not every remote posting that exists.

Source: Seeker job corpus, 445,429 active job postings, snapshot 2026-09-04. seekerscore.com/insights/remote-means-one-country-2026

Restriction is close to universal, and does not depend on what you do

95.7% of remote postings in our corpus name a single country as the eligible location. That figure holds almost regardless of field: the lowest is marketing at 93.4%, the highest is education at 97.9%, a spread of 4.5 points across 180,429 postings and ten fields. Whatever field you work in, if a posting is remote, the odds it actually means "anywhere in the world" are close to five percent, and that number does not meaningfully improve by choosing a different field.

This is the opposite of how remote eligibility itself behaves

We have separately shown that whether a field offers remote work at all varies hugely, some fields are mostly remote and others almost never are. That variation disappears once you filter down to postings that are already remote. The gate on whether you can work remotely is field-dependent. The gate on where you can live once you clear it is not.

What this means for how you search

A remote label tells you almost nothing about geographic reach on its own. If global eligibility matters to you, you need to read the location field or the posting text itself, because field choice will not shift your odds. The 4.5-point gap between marketing and education is real but small enough that it should not drive a career decision by itself.

How we measured it (and what we did not claim)

We took every active posting tagged remote whose location resolved to a known country, 180,429 of them, and counted how many resolved to exactly one country versus more than one. We hand-checked 13 rows against their raw, unprocessed location text, 8 single-country and 5 multi-country, and all 13 matched what the raw text actually said. We are not claiming this covers every remote posting in the corpus: postings whose location text did not resolve to any country are excluded entirely, and we do not know whether that excluded group looks like the postings we could measure or looks different, so treat 180,429 as the population we can see, not the full population that exists. We are also not claiming a causal reason for the pattern, only that it holds at the size and consistency we measured it.

Frequently asked questions

Related Data

Guides

• Remote Job Search →