Labor Market Intelligence
From Real-Time Job Data
Go beyond periodic labor statistics with granular employer demand signals that reveal how occupations, skills, industries, and regional labor markets evolve over time.
A Dataset Built for
Labour Research
Total records
5B+ deduplicated, enriched job records
Source
Employer career pages
Geographic coverage
100+ countries and territories
Languages
40+ languages in source text
Update frequency
Real-Time: as new postings detected within hours of going live
Historical data
Point-in-time job data available for historical analysis and panel datasets
Job title normalisation
Standardized job titles across records
Skills extraction
NLP-extracted from description text. 40,000+ term taxonomy
Salary data
Stated where posted; AI-estimated via four-signal model where not
Location
Geocoded to city, region, country, and coordinates
Deduplication
Duplicate job postings identified and removed
Delivery formats
MCP, Parquet, CSV, JSON, S3, API, data warehouse
Inside a Job Record
job_idUnique canonical identifier for the job opening
employer_nameCanonical employer name — normalised and deduplicated
employer_idPropellum employer entity identifier — maps to parent/subsidiary structure
title_rawJob title as posted by the employer
title_normalisedNormalised title — mapped to Propellum taxonomy
functionJob function — engineering, sales, finance, operations, etc.
senioritySeniority level — entry, mid, senior, director, VP, C-suite
skills_requiredArray of required skills — NLP-extracted, mapped to 40,000+ term taxonomy
skills_preferredArray of preferred skills — NLP-extracted
salary_minMinimum salary — stated or AI-estimated
salary_maxMaximum salary — stated or AI-estimated
salary_currencyCurrency of stated or estimated salary
salary_estimatedBoolean — true if AI-estimated, false if employer-stated
location_cityCity — geocoded
location_countryCountry — ISO 3166-1 alpha-2
location_latLatitude coordinate
location_lngLongitude coordinate
remote_typeRemote, hybrid, or on-site — classified from description
employment_typeFull-time, part-time, contract, temporary
date_first_seenDate posting first detected on employer career page
date_last_seenDate posting last confirmed active
snapshot_datePoint-in-time snapshot date — for historical analysis
Research Questions the
Dataset Can Answer
Skills Demand & Occupational Change
Track how skill requirements change within and across occupations to study emerging skills, occupational change, and polarisation.
Labour market concentration
Measure how employer hiring demand is distributed across firms, industries, and labour markets.
Wage and compensation dynamics
Analyze advertised salary ranges across roles, sectors, locations, and time to study compensation trends.
Geographic labour market analysis
Use city-level job data to measure where demand for specific occupations is concentrating, dispersing, or emerging over time.
Technology adoption and diffusion
Identify when new technologies and skills enter employer demand and measure how adoption spreads across firms, industries, and regions.
Policy and programme evaluation
Use point-in-time job data to compare employer demand before and after policy or programme interventions.
Data collection and
enrichment methodology
Understanding how the dataset was constructed is as important as knowing what it contains.
Data collection
Agentic crawler monitors employer career pages directly. Covers static HTML, JavaScript-rendered portals, paginated listings, and ATS-hosted pages. Adapts to site redesigns automatically.
Source
Employer's original job posting, not downstream job boards.
Deduplication
Cross-source deduplicate and standardize records for consistent analysis.
Title normalisation
Raw employer titles mapped to a consistent taxonomy using a proprietary classification model.
Skills extraction
NLP applied to job description text. Skills identified regardless of phrasing, synonyms and abbreviations resolved to consistent entities in a 40,000+ term taxonomy.
Salary estimation
Where employers do not post salary: four-signal AI estimation is applied. salary_estimated boolean flag on every record.
Historical Data
Preserve point-in-time job data to study how demand changes over time.
For Teams Turning
Labour Data Into Decisions
Labour economists and academic researchers
Strengthen research with granular evidence of employer behaviour, skills demand, and occupational change. Support deeper analysis beyond traditional aggregate labour statistics.
Policy analysts and government researchers
Evaluate workforce policies, regional economic resilience, and labour interventions with high-frequency employer demand signals rather than lagged administrative reports.
Central banks and monetary policy teams
Track wage pressure dynamics, labour market tightness, and occupational hiring shifts to complement macroeconomic forecasting with direct point-in-time posting data.
Workforce development boards and education planners
Identify emerging skills, declining roles, and geographic talent shortages to align vocational training and educational curricula directly with market demand.
Think tanks and public policy institutions
Produce empirical, peer-level research on technological adoption, future-of-work trends, and economic inequality backed by historical data across 100+ countries.
Economic consulting and advisory practices
Deliver rigorous market concentration studies, talent feasibility analyses, and regulatory evidence to enterprise and institutional clients with audit-ready job data.
Labor Market & Economic Research,
frequently asked questions
Evaluate the dataset
before you commit.
Tell us your research design, target geographies, time period, and required fields. We deliver a structured sample with full field definitions, methodology documentation, and coverage statistics. No commitment required.