Sources, history, geography, validation, and the limits every buyer should know about.
The best workforce intelligence data connects three things: what employers are hiring for, who actually holds those jobs, and how both align with official statistics. It should do all three in one standardized taxonomy that works at every level, from a single company to a national labor market.
This guide lists eight questions to ask any provider. We answer each one, and we close with where our data has limits.
What is workforce intelligence data?
Workforce intelligence data describes the people who work, the jobs they hold, and the jobs employers are trying to fill. It usually combines job postings, public professional profiles, and government statistics. It standardizes titles, occupations, skills, and pay so you can compare them across companies, regions, and time.
Buyers use it to benchmark competitors, plan hiring, size talent pools, price roles, track labor market conditions, and build investment signals.
Company data and labor market data should come from the same source
Many buyer’s guides split the market in two. One group of vendors covers companies: headcount, hiring, and talent flows. The other covers labor markets: jobs, skills, and regional wages.
That split reflects how vendors were built, and it causes a practical problem. When your company view and your market view come from different sources, they use different taxonomies, and the numbers won’t reconcile.
Revelio Labs builds both views from the same data. The profiles that show, for example, junior roles in US banks shrinking to 4.7% below their October 2022 level, while senior roles kept growing, also roll up into a monthly national employment report.
That report is Revelio Public Labor Statistics (RPLS). In our one-year retrospective, RPLS monthly employment change correlated with the Bureau of Labor Statistics (BLS) at 0.77. Smoothed over six months, the correlation rose to 0.97. RPLS is free and comes out the day before the BLS jobs report, and our postings data arrives a full month ahead of the Job Openings and Labor Turnover Survey (JOLTS).
The value of an independent source showed up in October 2025. A federal government shutdown halted the BLS jobs report, and RPLS still published, estimating that the economy added 60,000 jobs in September.
The same data breaks down to the state, county, and city level, with workforce history back to 2007.
Eight questions to ask any workforce intelligence provider
1. How many sources feed the data?
Data from one source carries that source’s blind spots. Ask how many distinct sources sit behind the job postings, and how many types of sources the provider combines.
Our answer: 5.6 billion job postings from 1.3 million unique sources, plus about 2 billion career profiles. We combine seven source families:
- Public professional profiles
- Job postings
- Employee reviews
- Freelance platform records
- WARN Act layoff filings
- Government data from the BLS, the International Labour Organization (ILO), the census, and the Social Security Administration
- Firmographic reference data
2. How far back does the history go, by dataset?
Workforce history and job posting history often start in different years. Ask for the start date of each dataset you plan to use.
Our answer: Workforce data goes back to 2007. Job postings start in 2021 and are deduplicated by COSMOS, our posting engine. COSMOS normalizes title, company, and location, then compares postings that overlap in time and merges the ones that match.
Because COSMOS reads the full text of each posting, it can measure what employers ask for. We used it to show that tech hiring is getting more specialized, tracking both the years of experience required for each skill and the number of distinct skills per posting.
3. What is the smallest geography you can analyze?
National trends rarely answer a hiring or site selection question. Ask how far down the data goes and whether it holds up at that level.
Our answer: Country, state, county, and city. We cover 195 countries, and all 195 have job posting data. Local detail matters most when activity moves, and surfacing trends is valuable. Our county and city data, for example, shows data centers spreading the AI boom beyond tech hubs.
4. How are titles, occupations, and skills standardized?
Raw job titles can’t be compared. Ask how the provider builds its taxonomy and how many levels of detail it offers.
Our answer: We embed every job title without seniority keywords, measure the distance between titles, and group them into about 15,000 occupation clusters. Skills nest from 15 broad categories down to 33,500 skill clusters. A separate model assigns seniority on a seven-level scale. Occupations map to standard government codes, so our data lines up with official statistics.
5. Is salary data observed or self-reported?
Survey data depends on who answers. Ask what share of the salary data comes from observed records.
Our answer: 1.7 billion salary observations, 1.4 billion of them from job postings. Our salary model is trained on visa application data, self-reported data, and postings, with country multipliers and inflation adjustment.
6. How do you correct for who is missing?
Online profiles overrepresent some occupations and countries. Ask how the provider adjusts for that.
Our answer: We stratify by occupation and country, then reweight against the BLS in the US and the ILO internationally. A nowcasting model corrects for the lag before people update their profiles, and we apply a separate seasonal adjustment. In the US, we cover more than 100 million profiles, about two-thirds of employed people.
7. Is the methodology public, and is accuracy tested in public?
Any provider can claim accuracy. Ask where the methodology is published and how anyone outside the company can check the results.
Our answer: Our full methodology is in our public data dictionary. RPLS is a monthly public test: we publish our numbers before the BLS publishes its own, and anyone can compare them.
8. How fresh is the data, and how is it delivered?
Ask how often each dataset updates and whether you can work with the data in your own tools.
Our answer: We deliver data by API, by data feed, or through the Revelio Labs terminal. We offer different refresh cadences and update frequencies through customized packages.
Which data fits which question?
- Track a company’s headcount, hiring, and attrition. Use workforce dynamics data, standardized at the company level, including subsidiaries.
- Analyze labor supply and demand by location. Combine profiles and job postings at the state, county, or city level.
- Plan workforce needs. Use occupation, skill, and talent flow data together, so demand and supply share one taxonomy.
- Benchmark pay. Use salary data built mostly from job postings.
- Monitor national labor conditions. Use RPLS, which is free and monthly. The latest report shows the economy adding 56.9k jobs in September 2026, even as active job postings fell 1.8%.
Where our data has limits
Every dataset has limits, and buyers should know ours.
- Job postings history starts in 2021. For hiring demand before then, use our workforce data, which goes back to 2007, or government series.
- Profile data can’t see temporary recalls. In the largest post-pandemic months, RPLS undershot BLS, because workers returning from furlough rarely update their profiles.
- Some sector totals differ from BLS. The biggest gap is in health care, where BLS shows larger cumulative gains than we do.
We publish these findings because a dataset is only useful when you know where it’s strong and where it’s weak.
Frequently asked questions
What is the best workforce intelligence data?
The best workforce intelligence data combines job postings, career profiles, and government statistics in one standardized taxonomy. It should cover company and labor market analysis from the same source, publish its methodology, and publicly test its accuracy.
What is the difference between workforce data and labor market data?
Workforce data describes the people inside companies: their roles, skills, and moves. Labor market data describes jobs, wages, and employment across regions and industries. When both come from the same source, a company view and a market view reconcile.
Can workforce intelligence data replace government statistics?
It works best alongside them. Government surveys set the benchmark. Workforce data adds company-level detail, finer geography, and earlier signals. For example, RPLS publishes before the BLS jobs report.
How far back does Revelio Labs data go?
Workforce data goes back to 2007. Job postings data starts in 2021.
What geographic levels does Revelio Labs cover?
Country, state, county, and city, across 195 countries.
Is Revelio Labs data available by API?
Yes. We deliver data via API, data feed, and the Revelio Labs terminal.