In November 2022, Amazon stock hit a 12-month low ($84.18), shortly after removing half of its job openings from its career site. The year was one of the retail giant’s worst ever; its stock lost 51 percent of its value from the start of the year.
But the most interesting thing (in our opinion) is what happened in October 2022, when its stock was still trading right around $150: Amazon started pulling job listings from its career site (see Figure 1).
Figure 1: Amazon job openings vs. closing price, Oct. 2022 – Jan. 2024
Job listings are often a leading indicator of company performance. But publicly available data sources, like BLS’s Job Openings and Labor Turnover Survey (JOLTS) don’t offer numbers for individual organizations. Rather, JOLTS and other BLS reports focus on aggregated data, which offers useful indications of the overall direction of the economy at a macro level.
Any organization interested in individual company performance, however, needs to find another data source. That’s exactly what our data can do.
As you probably know, JOLTS data comes from a survey the BLS conducts each month. By its nature, survey data is outdated by the time you see it: BLS surveys organizations, processes their responses, and publishes the resulting data.
But this data may also be less accurate than we generally assume it to be: BLS only surveys about 21,000 companies to get its numbers, and response rates have fallen in recent years. Today, only about 7,000 organizations actually respond to surveys, meaning the official JOLTS data we rely on for so much of our economic analysis is based on just a fraction of employers in the US.
It’s no wonder some economists have questioned the reliability of JOLTS in recent years.
This lack of precision is particularly frustrating in an era where real-time data is the norm in so many other contexts.
Luckily, there’s no reason you can’t have access to real-time data for the jobs market – in fact, that’s exactly what we track at Aspen Tech Labs. Rather than surveying organizations, we scrape career sites, meaning we gather data on actual postings that are actually live.
And we don’t do this for a sampling of US companies; we cover every large employer in the country and a steadily growing portion of small employers. Today, our JobsIndex includes roughly eight million job listings from more than 120,000 organizations (including the S&P 500, the DAX, the FTSE 100, and the Russel 3,000).
We update this data daily. If you’re interested in real-time data, we can also provide more frequent refreshes to our data sets.
What’s more, data scrapes are a primary line of business for us, meaning that many of our customers pay us to complete these scrapes. They use the scraped data to power various parts of their businesses, which means it must be accurate and timely.
Many of our competitors do not sell web scrapes as a standalone product, which means they are not as likely to notice broken scrapes and other issues as quickly as we do. Anyone interested in access to aggregated labor market data is unlikely to find a source that’s more consistently timely and correct than ours.
We couldn’t have told the Amazon story we led with without historical data. That data is a crucial piece of the puzzle when assessing any real-time trends. We offer robust historical data (160 million job listings) dating back to September 2022.
In addition to daily counts (including total live jobs, the number of jobs created, and number of jobs deleted), you can also request data down to the job level. This includes any detail based on the job posting, including the following:
In addition, you can request data samples by keywords – for example, every US job listing from the last year that mentions “Databricks” (see Figure 2) or “AI.” You could then track those organizations’ stock prices in the same period to determine whether hiring for AI roles or technology that enables AI work correlates with stock price.
Figure 2: Top 10 companies by job postings featuring the keyword “Databricks,” last 12 months
You could conceivably do this for any other keyword you suspect might correlate with market cap: “DEI,” “remote,” or “hybrid,” for example – or for specific data (like salary information) included in the listing.
In addition to being more timely than JOLTS data, our data is more granular. You can use it to track, for example, specific job titles, like account executives or developers, to determine which are leading indicators of growth.
You might also track vacancy duration, salary, benefits, and any other attribute you deem interesting.
We’re also able to deliver data in the format that works best for your existing operations: as XML, API, in a configured dashboard, or in some other format. We customize data delivery for many of our customers, and are always happy to accommodate specific needs.
The official sources of labor market data are woefully inadequate for any predictive applications or activities that require real-time accuracy. Our data – from the source, updated as often as needed, meticulously cleaned, and customized to spec – delivers.
When you’re ready to explore our data set and ask about customizations to suit your needs, get in touch. We’d be happy to set up a free, one-month trial so you can see the power of real-time data in action.