Freshest Markets
Most new postings from the past 7 days
- 1AtlantaGA · 36 of 8841%
- 2PhoenixAZ · 10 of 3231%
- 3HoustonTX · 24 of 8429%
- 4San JoseCA · 11 of 4425%
- 5AlbuquerqueNM · 6 of 2524%
National 16% posted this week
New corporate job postings
Built Oct 23, 2026. Rebuilt every 6 hours from the live posting pool. Last 30 days by default.
Where employers posted new corporate and office roles over the last 7 to 90 days, how many show ghost signals, and what they pay.
The default 30-day view is based on 12,583 new postings collected by ShouldApply from our job sources. It is a large sample, not every job posted in the US. See what it covers and misses
Last
Postings before October 5, 2026 came mostly from a few metros, so place-to-place comparisons lean on them. Full coverage for this window from November 3, 2026.
Click or tap states on the map, or use the list above. You will see their pooled totals and the plain average side by side, for new corporate postings.
Most new postings from the past 7 days
National 16% posted this week
Highest share of postings showing ghost signals
National ghost rate 3%
Salary after cost of living
National median $118K
Lowest share of postings showing ghost signals
National ghost rate 3%
All 50 states and DC have new postings. Remote and no-location roles are counted nationally and shown in the sidebar, not on a state.
New corporate postings from the last 30 days. A small number often means we search that state less, not that few companies hire there.
Line two cities up side by side: ghost rate, pay, freshness and cost of living.
New postings in a rolling window. Every number counts postings first listed in the last 7, 14, 30, 60 or 90 days, 30 by default. A posting is new when its posting date, or the day we first saw it when the board gave none, is inside the window. The pool keeps expired postings, so windows over 14 days include roles that may have closed since.
Reposts count once. Inside a window, the same employer, title and location is one posting, the newest. Job boards, USAJOBS and employer career boards feed one pool, and every posting goes through the same rules. Each is judged corporate or not from its title, then placed in a US state from its location text. When the city is one of the tracked cities, it counts there too.
Ghost signals, hidden by default. Each posting earns points for signals such as open for weeks, no pay listed, talent-pool wording, a repost, a staffing poster or the same role in many places. At 35 points it is flagged and hidden from counts, pay and freshness. The ghost share counts every posting. A signal is not a verdict. The Jobs list uses a stricter score, so its ghost counts differ from this map.
Everything is accounted for. The salary layers use only postings that post pay, converted to a yearly figure. Remote and no-state postings are counted nationally in their own card. In the last 30 days 591 postings had a location outside the US and 4,116 were not corporate or office work. Both are left out of every figure.
Pair it with the Ghost Job Index for employers, the salary explorer for roles, and the city reports for a deeper read on one place.
A posting earns points for each signal below. At 35 points it is flagged as showing ghost signals, and the switch above hides it. In the last 30 days 314 of 12,583 new corporate postings are flagged. A posting is flagged at 35 points. A few traits get there alone: open 60 days or more (35 points), talent-pool title (35 points) or listed again on 3 or more dates (35 points). No pay listed is worth 15, so no pay plus any trait worth 20 or more also flags: open 30 days or more, talent-pool wording, listed again on a second date, staffing or recruiter posting or apply page gone. Everything else needs company. These are signs, not proof, and not a finding about any employer. The Jobs list uses a stricter score, so its ghost counts differ from this map.
What we cannot see. We do not see applicant counts, whether the employer filled the seat, internal hiring freezes, or whether a recruiter is working the posting. We do not check careers pages or apply links at scale, and we do not use pay-transparency laws because a reliable state list was not available. Postings for hourly and service roles are left out of this map, so their relisting habits do not count here.
Longer windows show more ghost signals. A posting can only earn the open-a-long-time signal after it has been up for weeks, so a fresh posting cannot fire it. A 7 day window therefore shows a lower ghost share than a 90 day window for the same pool. Compare places inside one window, not across windows. The Jobs list uses a stricter score, so its ghost counts differ from this map.
A sample, not the market. The pool is what ShouldApply collected from job boards and job search APIs, federal postings on USAJOBS, and employers' own career boards on Greenhouse, Lever and Ashby. It is large, 12,583 new corporate postings in the last 30 days, and it is far smaller than every job posted in the US. Read each number as a share of what we collected.
Where the postings come from. Job boards and search APIs are searched by job title and place. The titles and places come from what ShouldApply users look for, from the metros our ads reach, and from a scheduled sweep of common titles in the largest US metros. Federal postings come from USAJOBS, one row per posting, and a posting that lists several states counts in its main one. Employer career boards add the US openings of the companies we follow. Only corporate and office roles count, judged from the title, so the clinical, trades, law enforcement and hourly federal postings are left out like any other. A state or city that the sources reach less shows fewer postings. That is a coverage effect, not low demand.
Why some states are thin. In the last 30 days 14 of 50 states with postings have fewer than 30. The smallest: Rhode Island (4), Wyoming (4), Delaware (8), South Dakota (8), Idaho (10), New Hampshire (10). The federal sweep rotates through every state. The scheduled search now also rotates through the largest city of each state it used to skip, inside the same request limit. Employer career boards add postings wherever those employers hire. A small count is still a coverage gap, not low demand.
Older windows lean on a few metros. Before October 5, 2026 most postings came from a handful of metros, Seattle and California above all. In the last 30 days the five biggest states hold 56% of placed postings and Seattle alone 8%. Full coverage from November 3, 2026 for 30 days, December 3, 2026 for 60 days, January 2, 2027 for 90 days. State-to-state comparisons are most reliable on the windows that are already clear.
Collection dipped for several weeks. Between August 5, 2026 and September 8, 2026 we collected far fewer postings than usual, though some sources kept going. Postings first listed in those weeks are undercounted in any window that reaches back over them, so a 60 or 90 day total reads low next to a window that does not. The note clears on November 7, 2026 for 60 days, December 7, 2026 for 90 days.
What is left out. Postings outside the US (591), postings that are not corporate or office work (4,116), reposts merged into one (214), and postings with no usable date. A posting hidden as a likely ghost is left out of counts but is included in the ghost share.
Context, not a comparison. The BLS Job Openings and Labor Turnover Survey (JOLTS) put US job openings at 7.1 million and hires at 5.2 million in August 2026 (seasonally adjusted, released September 29, 2026; source: bls.gov). Openings are a stock of open jobs and hires are a monthly flow, across every industry. This map counts new corporate postings first listed in a window, so the two are not like for like.
A posting is new when its posting date, or the day we first saw it when the board gave no date, falls inside the window you pick. Reposts of the same role at the same employer in the same place count once, and the newest is kept. The map counts new postings. It does not count open jobs or total employment. A window longer than 14 days includes postings that may have closed since. Pick 7, 14, 30, 60 or 90 days above the map. The default is 30.
No. It is a sample. In the last 30 days the pool holds 12,583 new corporate postings. They come from job boards and job search APIs, federal postings from USAJOBS, and openings that employers list on their own career boards (Greenhouse, Lever and Ashby). The real number of postings across the US is far larger. Some sources are searched by title and place, and the company boards cover only the employers we follow, so some states and cities are thin. Read the coverage section below before you compare places.
Inside a window, postings with the same employer, title and location count once, and the newest one is kept. Board decoration such as Remote or Full-time in brackets is ignored when titles are compared. Seniority words are not: a Senior Analyst and an Analyst are two openings. A job carried by a second board counts once, and two openings from one board on the same day stay separate. A posting with no employer name, or a placeholder such as Confidential, is never merged. In the last 30 days that merged 214 of 17,504 postings.
Each posting earns points for ghost signals: open for weeks, no pay listed, talent-pool wording in the title or description, being reposted, a staffing firm or recruiter as the poster, the same role listed in many places, a vague location, a junior title that asks for five or more years, a very wide pay range, and an apply page that is gone. A posting is flagged at 35 points. A few traits get there alone: open 60 days or more (35 points), talent-pool title (35 points) or listed again on 3 or more dates (35 points). No pay listed is worth 15, so no pay plus any trait worth 20 or more also flags: open 30 days or more, talent-pool wording, listed again on a second date, staffing or recruiter posting or apply page gone. Everything else needs company. A signal is not proof, and plenty of real roles show one. With the switch on, which is the default, flagged postings are left out of counts, pay, freshness and medians. In the last 30 days that hides 314. Turn the switch off to see everything. The Jobs list uses a stricter score, so its ghost counts differ from this map.
It is the share of new postings in the window that show enough ghost signals to be flagged, including the ones hidden when hiding is on, so it stays meaningful either way. It describes postings, not any employer, and it is not a verdict. A posting can show signals and still be a real opening. A posting can be a ghost and show none, because we cannot see inside an employer.
Most remote postings name no state, so they cannot sit on a state map. They are counted in the national total and shown in their own Remote and no state card, with their own pay, ghost share and freshness. In the last 30 days that is 5,131 postings, 41% of the total.
Postings with a location outside the US, such as Bengaluru, are left out of every figure. In the last 30 days that is 591. Postings that are not corporate or office work are left out too: 4,116 in the last 30 days. Postings with no usable date cannot be placed in a window and are skipped.
Corporate and office roles: marketing, finance, engineering, product, sales, HR, legal, operations, data and design. The label is about the work, not the employer. A federal Program Analyst or IT Specialist counts. A park ranger, nurse or firefighter at an agency does not. The pool has no industry field, so a title is judged by its words. Terms such as analyst, engineer, accountant, marketing, counsel and recruiter mark a role as corporate even when the title also names an industry, so a Restaurant Marketing Manager counts and a Restaurant Manager does not. Left out: trades and field work (1,147 in the last 30 days), hourly and service roles (1,005), food and hospitality (193), and frontline clinical and school roles plus any title no rule matched (1,771). They are listed here so the totals reconcile.
A state needs at least 10 postings before the ghost and freshness layers shade it, and at least 5 postings with a posted salary before the pay layers do. On the New Postings layer only a state with no postings is hatched. Hatched means too few postings to say anything fair, not a low value. A shorter window means fewer postings, so more states are hatched at 7 days. A thin state is mostly a coverage effect: we search fewer places there.
Median posted salary divided by the Regional Price Parity published by the U.S. Bureau of Economic Analysis, where 100 is the US average and housing rents are included. States use the 2024 figures (BEA 26-10, February 19, 2026). City pages use the 2023 metropolitan area figures (BEA 24-59, December 12, 2024, Table 4), because BEA publishes the metro tables later. The index describes the whole metro area, not the city alone. BEA updates the figures each December.
A posting counts toward a city when its location names that city and state. Suburbs such as Bellevue are counted in their own state total but not folded into Seattle.
New postings are collected every 2 hours, with a wider sweep of common titles in the largest metros four times a day. Federal postings from USAJOBS are swept three times a day and employer career boards every 3 hours. The map is rebuilt from the pool every 6 hours, and the date under the headline shows the last build.
5,131 remote and no-state postings, 7,138 placed on a state and 314 hidden as likely ghosts make up the 12,583 new corporate postings above, for the last 30 days.