For workforce boards and colleges
Where to put training money, job by job, in your region
The question a workforce board, a community college or an economic development team brings is not whether AI matters, but which local jobs it reaches, when, what stays with the people in them, and what they would need to learn. The Workforce Scan answers it for every state and metro area from the jobs up: the tasks of each occupation, how long each takes a person, and when AI models can finish work that long, weighted by how many people do each job there.
1. Every area, free
A page for each of the 50 states and DC, 393 metro areas and 137 nonmetropolitan areas, from BLS counts of the people in every occupation there (May 2025): the share of paid hours now and at the end of each year to 2030 on three paces, the jobs and tasks that carry the most hours, what handing the work over would take, and what changes for the people: the share of their working time that stays with them, that becomes checking a model's work, and what most of them would learn first.
Find your state or metro area or see every state side by side
2. The regional report: the whole workforce
An area page lists its largest jobs. Its regional report reads all of them: every occupation BLS publishes there that the site has task data for (most of the people; each area page says how many), at its local headcount and pay, run through the same scan companies use. The preview is free. The full report is $249, or included in Pro: every occupation at every date and pace, the tasks behind the hours, what changes for the people in each job and what they would learn, with an Excel workbook to cite and a slide deck to present. It is dated, and the method is published.
Open an area, then "Scan this area's workforce" Sample: Worcester, MA, ExcelSlides
3. The data, to use in your own work
Every area's numbers are an open CSV file under CC BY 4.0, free to use with credit, and the API (included in Pro) returns any area or industry in full. The numbers are rebuilt when METR publishes a new measurement of what models can do, and every earlier projection is kept, dated, on the record.
How far the review has come
What stays with people and what to learn come from a review of every task, one at a time: how much of it is documents, business systems, conversation, in-person or hands-on work, and what a mistake would cost. 326 of 923 occupations are done, holding 91% of the jobs BLS counts, as of October 1, 2026; each area page says how many of its people that covers. Data checked October 1, 2026.
What the numbers are, and what they are not
Within reach means a model could finish work of that length at the chosen reliability, if the work were set up for it. It is not a forecast of jobs lost, of hiring or of wages, and it says nothing about whether any employer will hand work over or when. The estimates rest on METR's measured task horizons, O*NET's task statements and BLS employment and pay; the method page gives the sources and the ranges.
A report for your region
To see the full regional report for your area before buying, write to info@stratussc.com.