How the ratings are built
Every figure comes from a public, citable source. Where the data is thin we say so next to the number. Postings are not hires: they show where employers are trying to add people, which is the earliest signal of demand.
Independence rules. Ratings are never paid for, sponsored or adjusted on request, including for clients and partners. A method changes only through the public changelog below, and every issue stays archived as published. Verdicts on claims are published whichever way they fall.
The demand rating
Each occupation gets a rating in each country: rising, edging up, holding, edging down or thinning. When the two signals point in opposite directions, the rating says the signals disagree instead of forcing a call. Every rating names its basis, for example "weekly rising, outlook holding".
Agreeing signals give a firm rating. Where one signal is neutral, the rating leans toward the other. Where signals conflict, we publish the conflict. Where a signal is missing, the basis says so.
Signals
Weekly postings (sector level)
Indeed Hiring Lab job_postings_tracker (CC BY 4.0), US and Canada sector files, total postings index (Feb 1, 2020 = 100, 7-day average). Each issue compares the latest week with the same week a year earlier. The figure is compared with all job postings in the same country: within 5 points of the all-jobs change reads as holding, 5 points or more above it as rising, 5 points or more below it as thinning (this issue's cut-offs: US +8.1% and −1.9%; Canada +8.8% and −1.2%). For the US 2025–35 outlook, growth of 7% or more reads as rising and 0% or less as thinning. Each occupation takes the figure for the Hiring Lab sector it maps to, so all occupations in one sector share that figure. Where our job-board snapshot holds enough live roles for an occupation, its own posting count is shown alongside.
Official outlooks
US: BLS Employment Projections 2025–35, with 2025 median pay from OEWS. Canada: Job Bank 2025–27 outlooks, combined across provinces into a national score, with Job Bank wages.
Live roles and hiring themes
Public job boards of 644 companies, pulled Nov 3, 2026: 28,843 distinct US and Canada roles. An occupation is tagged with a hiring theme (AI, robotics and autonomy, energy and grid, data centres, semiconductors, quantum) when a meaningful share of its live roles involve that technology.
Scope and matching
US: occupations in BLS Employment Projections (Table 1.2) whose typical entry needs postsecondary training, a degree or an apprenticeship. Canada: NOC 2021 TEER 0–3. All management, engineering, science and computing occupations are included. US and Canadian occupations are matched using Statistics Canada concordances and a manual audit. Occupations with no US match appear as Canada only.
Limits
- Sector postings describe the market an occupation sits in, not the occupation alone.
- The job boards miss employers on Workday and similar systems, and lean toward startups and tech companies.
- Hacker News hiring threads skew toward startups; we read them as direction only.
- Nothing is scraped from LinkedIn, Indeed or Glassdoor, and no Reddit data is used.
Publishing schedule
The index publishes every Thursday, using Indeed Hiring Lab data released the day before. Issue 2 publishes Thursday, Nov 12, 2026. Each issue is archived as published.
Changelog and archive
Sources and licences
Indeed Hiring Lab job_postings_tracker and ai-tracker (CC BY 4.0). US Bureau of Labor Statistics Employment Projections and OEWS (public domain). Statistics Canada concordances (Statistics Canada Open Licence). Job Bank Canada outlooks and wages (Government of Canada). Public company job boards, used for aggregate counts only.