United States tech hiring dossier
Figures cover the three months to November 2026, measured since August 2025.
We tracked 20,724 United States tech job postings over the three months to November 2026. The tables below show which skills and roles that demand named, and how each moved against its own recent baseline.
Skill demand in United States
Postings mentioning each skill over the trailing quarter, as a count and as a share of the United States postings we tracked.
| Skill | Postings | Share | vs prior quarter | vs year ago |
|---|---|---|---|---|
| python | 10,161 | 49.0% | — | — |
| databases-sql | 8,130 | 39.2% | — | — |
| aws | 6,955 | 33.6% | — | — |
| system-design | 4,591 | 22.2% | — | — |
| ci-cd-pipelines | 3,904 | 18.8% | — | — |
| docker | 3,715 | 17.9% | — | — |
| kubernetes | 3,616 | 17.4% | — | — |
| ai-engineering | 3,525 | 17.0% | — | — |
| http-apis | 3,517 | 17.0% | — | — |
| java | 2,942 | 14.2% | — | — |
| agile-delivery | 2,871 | 13.9% | — | — |
| javascript | 2,832 | 13.7% | — | — |
| data-engineering | 2,815 | 13.6% | — | — |
| devops | 2,771 | 13.4% | — | — |
| react | 2,759 | 13.3% | — | — |
| machine-learning | 2,530 | 12.2% | — | — |
| linux | 2,521 | 12.2% | — | — |
| typescript | 2,258 | 10.9% | — | — |
| cloud-security | 2,217 | 10.7% | — | — |
| git | 2,113 | 10.2% | — | — |
| terraform | 2,080 | 10.0% | — | — |
| azure | 1,968 | 9.5% | — | — |
| apache-spark | 1,942 | 9.4% | — | — |
| data-science | 1,938 | 9.4% | — | — |
| test-automation | 1,819 | 8.8% | — | — |
| cpp | 1,389 | 6.7% | — | — |
| nodejs | 1,348 | 6.5% | — | — |
| gcp | 1,255 | 6.1% | — | — |
| go | 1,237 | 6.0% | — | — |
| ai-coding-tools | 1,188 | 5.7% | — | — |
A dash marks a growth cell held back by the sample gate: fewer than 12 postings in one of the compared windows, or not enough stable history yet to compare.
Role demand in United States
Postings per role type over the trailing quarter.
| Role | Postings | Share | vs prior quarter | vs year ago |
|---|---|---|---|---|
| Software Engineer | 6,606 | 31.9% | — | — |
| Data Engineer | 1,925 | 9.3% | — | — |
| Full Stack Engineer | 1,799 | 8.7% | — | — |
| Security Engineer | 1,506 | 7.3% | — | — |
| DevOps Engineer | 1,471 | 7.1% | — | — |
| Machine Learning Engineer | 1,391 | 6.7% | — | — |
| Data Scientist | 1,345 | 6.5% | — | — |
| Backend Engineer | 905 | 4.4% | — | — |
| Engineering Manager | 864 | 4.2% | — | — |
| Data Analyst | 770 | 3.7% | — | — |
| QA Engineer | 708 | 3.4% | — | — |
| Embedded Engineer | 543 | 2.6% | — | — |
| Frontend Engineer | 502 | 2.4% | — | — |
| Mobile Engineer | 389 | 1.9% | — | — |
A dash marks a growth cell held back by the sample gate: fewer than 12 postings in one of the compared windows, or not enough stable history yet to compare.
Posting turnover, market by market
How long postings stay open and how often they are reposted, per market we track. Levels over each market's observation window, not a week-over-week figure.
| Market | Postings tracked | Median days open | Reposted |
|---|---|---|---|
| United States (this dossier) | 59,322 | 6.721 | 54.1% |
| Australia | 24,245 | 0 | 21.1% |
| United Kingdom | 23,457 | 0 | 45.6% |
| UNKNOWN | 2,733 | 6.721 | 71.3% |
| Canada | 349 | 6.721 | 57.6% |
| India | 214 | 6.721 | 77.1% |
| Singapore | 167 | 6.718 | 53.3% |
| Ireland | 126 | 6.721 | 55.6% |
| France | 75 | 6.721 | 60.0% |
Where roles sit open longest
The United States role and location cells whose postings stay open longest, and how often the same listing comes back. These are the roles where a candidate has the most time and leverage. Observations over the postings we tracked, not a verdict on any employer. Only cells with at least 30 tracked postings appear, as of the week of 9 November 2026.
| Role | Where | Typical time open | Relisted | Open 30+ days | Postings |
|---|---|---|---|---|---|
| Mobile Engineer | San Francisco | 5.8 weeks | 28.2% | 51.6% | 117 |
| Frontend Engineer | Boston | 3.9 weeks | 27.5% | 50.0% | 69 |
| Frontend Engineer | San Francisco | 3.8 weeks | 27.8% | 34.8% | 241 |
| Embedded Engineer | San Francisco | 3.8 weeks | 18.0% | 41.7% | 78 |
| Embedded Engineer | Austin | 3 weeks | 25.9% | 31.0% | 135 |
| DevOps Engineer | Boston | 3 weeks | 23.8% | 44.9% | 269 |
| Frontend Engineer | Austin | 3 weeks | 21.3% | 29.6% | 75 |
| Frontend Engineer | Atlanta | 3 weeks | 16.1% | 33.3% | 56 |
| Backend Engineer | San Francisco | 2.8 weeks | 24.1% | 29.5% | 295 |
| Machine Learning Engineer | Austin | 2.8 weeks | 20.5% | 31.5% | 239 |
| Data Scientist | San Francisco | 2.8 weeks | 18.1% | 25.3% | 443 |
| Frontend Engineer | Denver | 2.5 weeks | 31.7% | 35.7% | 41 |
Visa sponsorship
Of the postings read in the trailing 90 days, the share that state visa sponsorship is available: 0.8%, and the share that explicitly rule it out: 7.9% (31,371 postings read; most postings never address sponsorship either way).
Methodology
- United States postings are collected weekly from Adzuna and Dice. Published figures are computed on calendar-month and trailing-quarter windows; we do not publish week-over-week movement, because weekly counts move with collection noise rather than hiring.
- Growth cells are suppressed when either compared window holds fewer than 12 postings. Year-over-year cells additionally wait for 15 months of stable history.
- “Gaining momentum” flags use the rule validated on nine years of Swedish national job-ad history: trailing-quarter share of at least 0.5%, at least 1.5× its 12-month baseline, on at least 12 postings. The validation study and its dataset are published in the Global Tech Hiring Bulletin and on the open data page.
- When our own collection coverage shifts (a new board, a capture change), the affected market's trend history restarts at the shift and the page says so. Figures are never compared across a coverage change.
Vacancy and salary data includes Adzuna.