Six Figure Work Insights

How every number here is made

Name: How every number here is made
Creator: Six Figure Work
License: https://sixfigurework.com/terms

The rules behind every statistic on these pages, in plain words, with the data dictionary at the end.

Live updated 28 min ago (Nov 5, 2026 17:20 UTC)
64,604 postings
2,738 employers
5,255 boards monitored, 99.7% checked in 24 h

Among jobs that publish an annual base of $100K+ on the employer boards we monitor. Not all jobs. About the sample

The rules

Every number on these pages is read from a stored statistic. Nothing is typed in by hand, and each chart can show its own table and calculation.
Every claim carries its sample size, written as n.
A number that rests on too few postings is hidden and labelled "not enough data", with the n it would have had. It is never shown as zero.
Postings we could not classify are kept as their own row, "Not determined". They are never dropped from a chart.
Trends use fixed sets of employers and print the date their record starts.

The sample

The pay statistics cover 64,604 job postings, one row per posting. A posting is in the sample when the employer's own job board states an annual base salary in US dollars whose bottom is $100K or more, the posting is open to the public, and it passed the same checks as every listing on the site (site methodology). When a posting lists several pay ranges, we use the one with the highest floor.

So the sample is six-figure postings on the 5,255 employer boards we monitor. It is not every job, not every employer, and not what anyone is paid. Bonus, commission, equity and hourly pay are not counted.

The pay-disclosure statistics use a wider set: all 625,931 open postings on those boards, whatever they pay. A posting "discloses" when any pay figure appears on it, hourly included. A disclosure rate is not a legal compliance rate: laws differ in scope, exemptions and dates, and we do not judge any posting against a law.

What a median means here

The median floor ($147K right now) is the middle floor among six-figure postings. Because every posting starts at $100K or more, the median describes the upper part of the market only.

The cut matters most for junior roles. Right now the median floor for mid-level postings is $131,250 (n=1,320) and for entry-level postings $130K (n=726). Most entry and mid-level jobs pay under $100K and never enter the sample; the few that do are unusual roles, often in high-paying fields. Read the seniority chart as a description of this sample, not as a career ladder.

A floor is the bottom of an advertised range. Offers can land anywhere in the range, or outside it, and postings do not show where.

Removed is not filled

"Removed" means the posting disappeared from the employer's board. It may have been filled, paused, moved or cancelled. We do not know which, so we never call it a hire.

Time on board is the number of days from when we first saw a posting to when it was removed. For postings removed in the last 7 days the median is 10.0 days (n=83,307). We started watching on Oct 6, 2026, and each board on its own first fetch, so older postings look younger than they are. It is not a time to fill.

Trends and panels

We are still adding employer boards, so raw daily totals mostly measure our own growth. Trends therefore follow a fixed panel: the employers we already watched on a given day, frozen from then on. Each trend prints its panel, the number of employers, n for every day, and the date its record starts ("tracking since").

Panel p20261008: 448 employers, frozen on Oct 8, 2026.
Panel p20261015: 1,932 employers, frozen on Oct 15, 2026.
Panel p20261101: 4,398 employers, frozen on Nov 1, 2026.

A day counts only the employers whose boards were all fetched that day; an employer we could not observe is left out, never counted as zero. A series with fewer than 7 days gets no trend claim. Points are never smoothed. Days with intake events (a backlog cleared, rules changed) carry a numbered note, and a change of calculation method is marked on the chart.

Pay statistics are only tracked from Oct 24, 2026. Before that, pay data was re-evaluated in place, so earlier days cannot be rebuilt honestly.

Coverage and refresh

We monitor 5,255 approved employer boards; 99.7% had a successful fetch in the last 24 hours. Of published postings, >99.9% were re-checked within the last 8 hours.

The live statistics are recomputed about every 30 minutes, and these pages show them within 5 minutes of each run. Each night a copy of the day's values is stored as history; trends are drawn from that history.

Small numbers

Medians and percentiles are hidden when they rest on fewer than 5 postings, and fewer than 30 for two-way cells (job family by state, job family by seniority) and range findings by group.
Disclosure rates are hidden when fewer than 200 open postings sit behind them.
Counts are always shown.
Shares are calculated with their numerator and denominator stored next to them; the tables print both where it helps.

What we do not publish, and why

Pay changes on open postings (direction). Events before 24 Oct 2026 compared different grains; the fixed events need weeks of history.
A named ranking of employers by pay disclosure. Held back until employers can see and correct their own numbers.
Tech-skill pay premiums. Not in the live statistics yet; a premium also has to be measured within job family to mean anything.
Pay trends before 24 Oct 2026. Pay data was re-evaluated in place on 8-12 Oct, so earlier days cannot be replayed honestly.
Headcount, hires, applicants, time to fill. Job postings do not show any of these.

There is no download or API for these statistics yet. When there is, it will be linked here.

Versions

Every stored value carries its calculation version. When a formula changes the version changes, and trend charts mark the day. Changes so far:

insights pages (2026-10-24, vw-insights-build): /insights and six topic pages read live_stat / daily_stat only; no formula changed, calculation_version stays snap-v2.
snap-v2 (2026-10-24, vw-insights-s0): insights.daily_stat / live_stat / panel / intake_note (migration 049); statistic formulas above; salary_changed payload at posting grain; past-date disclosure rates re-aggregated from market cells as snap-v2-mkt. snap-v1 snapshot tables unchanged.
snap-v1 grain note (2026-10-08, vw-company-entity): daily_company_snapshot grouped by company_id going forward; formulas unchanged, calculation_version not bumped because employer:company is 1:1 today.
snap-v1 (2026-10-08): initial.

Questions or corrections: [email protected].

Data dictionary

Every statistic these pages can show, generated from the same registry the pages use. "Rows" is the number of stored live values (one per group).

Data dictionary
Id	Statistic	Formula	Denominator	Hidden when	Version	Rows
A1	Median advertised base floorA floor is the bottom of an advertised range, not pay. Only floors of $100K+ are in the sample.	50th / 25th / 75th percentile of each posting's lowest published annual base	published postings (one row per posting, $100K+ annual base, open to the public)	quantiles hidden when n < 5	snap-v2	4
A2	Floor distributionThe sample starts at $100K by design, so the left edge is cut, not empty.	postings by floor band ($25K steps to $200K, then $50K, then $300K+)	published postings (one row per posting, $100K+ annual base, open to the public)	counts are never hidden	snap-v2	14
A3	Median floor by job familyJob family comes from our taxonomy; 'not determined' is shown as its own row.	median / p25 / p75 floor and share of postings, by job family	published postings (one row per posting, $100K+ annual base, open to the public)	quantiles hidden when n < 5	snap-v2	150
A4	Median floor by state	median / p25 / p75 floor and share, by the posting's state	published postings (one row per posting, $100K+ annual base, open to the public)	quantiles hidden when n < 5	snap-v2	270
A6	Median floor by seniorityThe $100K floor removes most entry and mid-level postings, so the juniors left are the best-paid ones.	median / p25 / p75 floor and share, by seniority level	published postings (one row per posting, $100K+ annual base, open to the public)	quantiles hidden when n < 5	snap-v2	45
A7	Median floor by workplaceMost postings do not state a workplace; those are 'not determined'.	median / p25 / p75 floor and share, by stated workplace (remote, hybrid, on-site)	published postings (one row per posting, $100K+ annual base, open to the public)	quantiles hidden when n < 5	snap-v2	20
A8	Median floor by company size bandSize = open postings we see on the company's boards, not headcount.	median / p25 / p75 floor and share, by the company's open postings we see	published postings (one row per posting, $100K+ annual base, open to the public)	quantiles hidden when n < 5	snap-v2	30
A9	Median floor by job-board platform	median / p25 / p75 floor and share, by the vendor that hosts the board	published postings (one row per posting, $100K+ annual base, open to the public)	quantiles hidden when n < 5	snap-v2	50
A10	Federal vs private floors	median / p25 / p75 floor and share; federal = USAJOBS postings open to the public	published postings (one row per posting, $100K+ annual base, open to the public)	quantiles hidden when n < 5	snap-v2	10
A11	Share of floors at or above a threshold	postings with floor >= $150K / $200K / $300K, divided by all published postings	published postings (one row per posting, $100K+ annual base, open to the public)	counts are never hidden	snap-v2	3
A12	Floor by job family and state	median / p25 / p75 floor in each job family x state cell	published postings (one row per posting, $100K+ annual base, open to the public)	quantiles hidden when n < 30 (two-way cells and finding rows)	snap-v2	4,124
A13	Floor by job family and seniority	median / p25 / p75 floor in each job family x seniority cell	published postings (one row per posting, $100K+ annual base, open to the public)	quantiles hidden when n < 30 (two-way cells and finding rows)	snap-v2	940
B1	Range widthA range is an advertisement. Where offers land inside it is not visible in postings.	for postings whose top is above the floor: quantiles of top minus floor ($) and of top divided by floor	published postings (one row per posting, $100K+ annual base, open to the public)	quantiles hidden when n < 5	snap-v2	8
B2	Narrow and wide ranges	share of ranges with top/floor < 1.2, >= 1.5, >= 2, and width >= $100K	postings with a real range	counts are never hidden	snap-v2	4
B3	Range width by floor bandBands can be led by one large employer; its share is printed.	median width and top/floor ratio by floor band; largest single employer's share of the band	postings with a real range	quantiles hidden when n < 30 (two-way cells and finding rows)	snap-v2	20
B4C	Range width by job family	median width and ratio by job family; largest single employer's share	postings with a real range	quantiles hidden when n < 30 (two-way cells and finding rows)	snap-v2	115
B4S	Range width by state	median width and ratio by state; largest single employer's share	postings with a real range	quantiles hidden when n < 30 (two-way cells and finding rows)	snap-v2	205
B5	Range, single figure or floor only	postings with top > floor / top = floor / no top, divided by all published postings	published postings (one row per posting, $100K+ annual base, open to the public)	counts are never hidden	snap-v2	8
B6	Federal vs other range width	median width and ratio, USAJOBS vs every other board	postings with a real range	quantiles hidden when n < 5	snap-v2	8
B7	Who posts the widest rangesNamed because one employer can be a large share of a small group; this is a description of postings, not a ranking of employers.	employers' shares of all ranges $100K wide or more (top five)	ranges $100K wide or more	counts are never hidden	snap-v2	7
B8	Massachusetts range ratio by employer	median top/floor ratio for employers with 30+ ranges in MA	MA ranges of that employer	quantiles hidden when n < 30 (two-way cells and finding rows)	snap-v2	20
B9	VA $300K+ ranges with physician-like titles	share of Department of Veterans Affairs ranges with floor >= $300K whose title names a physician-like role	those VA ranges	counts are never hidden	snap-v2	2
C0	Pay disclosure rateDisclosure here is any stated pay, not a $100K+ base, and it is not a legal compliance rate.	open postings with any stated pay on any variant (any currency or period, hourly included) / all open postings	all open postings on approved boards (open + closure pending), any pay or none	share hidden when the denominator is under 200 postings	snap-v2	1
C1	Pay disclosure by state	the C0 rate by the posting's state	all open postings on approved boards (open + closure pending), any pay or none	share hidden when the denominator is under 200 postings	snap-v2	55
C2	Pay disclosure by pay-transparency law groupA correlation. Law states also differ in employers and job mix.	the C0 rate for postings in states whose law requires pay in postings, states that require it on request, other states, and no state	all open postings on approved boards (open + closure pending), any pay or none	share hidden when the denominator is under 200 postings	snap-v2	6
C3	Pay disclosure by job family	the C0 rate by job family	all open postings on approved boards (open + closure pending), any pay or none	share hidden when the denominator is under 200 postings	snap-v2	30
C4	Pay disclosure by job-board platformSome platforms carry pay in a structured field, which raises their rate.	the C0 rate by board vendor	all open postings on approved boards (open + closure pending), any pay or none	share hidden when the denominator is under 200 postings	snap-v2	11
C5	Companies by disclosure bandCompanies are not named here.	share of companies (by size band) whose disclosure rate is >= 90%, >= 50%, or 0%	companies with open postings, per size band	counts are never hidden	snap-v2	15
E1	Largest publishers of $100K+ rolesPostings we publish, not headcount or hiring.	companies ranked by published postings (top 25)	published postings (one row per posting, $100K+ annual base, open to the public)	counts are never hidden	snap-v2	25
E2	Concentration	share of published postings held by the top 10 / 50 / 100 companies	published postings (one row per posting, $100K+ annual base, open to the public)	counts are never hidden	snap-v2	5
E3	Companies by size bandSize = open postings we see, not headcount.	companies with open postings, by open postings we see	companies with open postings	counts are never hidden	snap-v2	5
D1	New postings per dayNew means newly seen by us, not new hiring. Fixed panels only.	first sightings after a board's first complete fetch, summed over companies with every board fetched OK	companies whose boards were ALL fetched OK in the 24 h window (unobserved is not zero)	counts are never hidden	snap-v2	2
D2	Removed postings and net change per dayRemoved from the board, never 'filled' or 'hired'.	removed = taken off the employer's board; net = new + reopened - removed	companies whose boards were ALL fetched OK in the 24 h window (unobserved is not zero)	counts are never hidden	snap-v2	3
D3	Time on board at removalTime on board, not time to fill. Cut off at 6 Oct 2026 and at each board's first fetch.	median and 90th percentile days from first sighting to removal, postings removed in the last 7 days	postings removed in the last 7 days	quantiles hidden when n < 5	snap-v2	3
D4	Baseline first sightings	postings first seen on a board's first complete fetch (we started watching; says nothing about hiring)	companies whose boards were ALL fetched OK in the 24 h window (unobserved is not zero)	counts are never hidden	snap-v2	1
D5	Reopen rate	reopened today / removed by the same companies in the 7 prior days	companies whose boards were ALL fetched OK in the 24 h window (unobserved is not zero)	hidden when fewer than 30 removals	snap-v2	1
H1	Boards monitored and checked	approved employer boards, and the share with a successful fetch in the last 24 h	approved boards	counts are never hidden	snap-v2	2
H2	Freshness	share of published postings re-checked within 8 hours	published postings (one row per posting, $100K+ annual base, open to the public)	counts are never hidden	snap-v2	1
H3	Boards by platform	boards monitored and share checked in 24 h, by vendor	approved boards of that vendor	counts are never hidden	snap-v2	22