The job market, in numbers.
Every figure below is computed from the PivotHop corpus — 4,401 live listings from 1,357 companies across 65 countries, mapped onto standardized occupations and refreshed nightly. Data as of 27 Jul 2026. Nothing here is invented; each number links to the page that proves it.
Scale
The corpus reads more than 100,000 live job postings and maps them onto 177 standardized occupations, refreshed nightly.
Job titles, deprecated →In a single month, 42,254 distinct job-title strings mapped to no standard occupation at all — titles fragment while the skill demand under them clusters.
the thesis →AI exposure
4.9% of all postings now demand LLM or agent-tooling skills by name, across 43 of 177 occupations — including lawyer, recruiter, and motion designer.
AI and jobs, checked →Every occupation in Trades, Healthcare, Construction, and Transport shows zero AI-skill demand; in Technology, only 37% do. The hands-on economy is not being rewritten around AI.
the two AI-proof workers →Adjacency
Of the 3,521 scored occupation-to-occupation pairs, 55% sit under 20% skill readiness. Skills are specific, not a universal solvent — which is exactly why measuring their reach matters.
every measured route →The most transferable careers have a data core or a coordination core: sales engineer and operations manager each reach 13 destinations at 45%+ readiness; data scientist reaches 12.
the widest-exit ranking →The confused pairs
A typical project manager’s skills cover only 24% of what product-manager postings demand; the reverse shares too few skills to score. Same first word, different jobs.
compared →Graphic designers and UX designers share 13% and 12% of a skill set in each direction, against posted bands of $42k–$73k versus $74k–$151k — the most expensively confused pair in design.
compared →A registered nurse reads 94% ready for nurse-practitioner work on skills alone — and no skill overlap shortens the graduate degree and license between the two titles.
compared →Bridge skills
Data analysis is the market’s reserve currency — in the top-20 posted demand of 62 of 177 occupations. LLM and agent tooling already sits in roughly a third.
the bridge-skill table →Method, in one paragraph
Skill readiness is the share of a destination occupation’s posted skill demand that a typical origin profile already covers, measured directionally from live postings; pairs sharing fewer than three skills are left unscored rather than guessed. Salary bands are posted 25th–75th percentiles, stated salaries only. A required license is shown as a gate and floors the transition estimate — a 90% skill overlap does not shorten a three-year degree. The full method is in What is career adjacency, and every number recomputes with the nightly scrape.
Cite as: PivotHop Adjacency Index, 27 Jul 2026 (pivothop.com/adjacency-index).