Gulf Jobs Scraper — compare countries by real openings

Pick one role and two or more Gulf countries, and get back one row per country ranked by the number most job boards never show you: what share of the postings you can actually apply to.

A Gulf search that returns 300 results looks like a strong market. Strip out the postings that restrict applicants by nationality, then strip out the ones that have been sitting there for over a month, and the number that is genuinely open to you can be a fraction of that. This Actor does that arithmetic per country and shows its working.

What you get per country

Field	Example
market	saudi-arabia
openFreshRate	0.42 — share that is neither nationality-restricted nor stale
openFreshRateRange	[0.31, 0.54] — 95% range for that share
openFreshJobs, jobs, jobsDated	13, 30, 31
nationalityRestrictedRate	0.27 — share whose text restricts by nationality
staleRate, medianPostedDaysAgo	0.31, 18
uniqueCompanies, topPosterShare, topPoster	14, 0.30, Gulf Recruitment LLC
rank, notRankedReason	1, or why the row was left out

The run summary reports totalJobs next to totalOpenFreshJobs, so the gap between the headline count and the real one is a single line.

What this Actor refuses to do

Rank by job count. Every country is sampled the same number of postings, so the count reflects your sample size, not the market. Ranking is by rate, with a Wilson confidence range — which stays honest at 0% and 100%, where the textbook formula collapses to zero width and would let a 0-of-8 result read as certainty.

Count a posting as fresh because it could not be dated. Listings whose posted date cannot be read are excluded from both sides of the freshness ratio. Folding them in as "recent" would make the stalest markets look best.

Let one recruiter stand in for a market. topPosterShare is on every row, and the summary names any country where a single company holds 30% or more.

Countries with fewer than 10 datable postings get rank: null rather than a rate that swings on one listing.

Summary (key-value store, BENCHMARK_SUMMARY)

query, markets, ranked, mostOpen, mostOpenRate, topTwoTooClose, medianOpenFreshRate, totalJobs vs totalOpenFreshJobs, restrictedMarkets, concentrated, notRanked.

Input

{
  "query": "data analyst",
  "countries": ["uae", "saudi-arabia", "qatar"],
  "samplePerCountry": 30
}

Available markets: uae, saudi-arabia, qatar, kuwait, bahrain, oman, egypt, jordan, lebanon, morocco. A single country is rejected — one market is not a comparison. You are charged per country, not per job.

Limits, stated plainly

• Nationality restrictions are detected from the wording of the posting. A restriction expressed only during screening will not appear here, so openFreshRate is an upper bound on how open a market is.
• Only job listings are read. Nothing about applicants is collected.
• Bayt refuses direct traffic, so a residential proxy is required.

Unofficial; not affiliated with Bayt.