The Ghost Job Index

A “ghost job” is a listing kept up with no live intent to hire. Intent cannot be measured from the outside, so this index does not try. It measures the three things the term is attached to in practice — listings that are old, listings that never come down, and listings that come back under the same title — for every employer with enough history to be measured, from 27,368 live postings at 300 employers’ own career pages.

Published by FindRole · computed · Methodology · Download CSV

Key findings

  • As of October 2026, 26.6% of the 27,368 live US technology listings tracked by FindRole have been up for more than 60 days (7,283 listings). The median open listing is 30 days old.
  • Across the 71 employers with enough closure history to be measured, the median employer still has 67.8% of its listings up 30 days after they appear. At 21 of them (30%), more than half of all listings outlive our 68-day observation window.
  • 26.8% of the 14,255 listings we watched close at those employers came down while an identical title stayed posted. The range runs from 0.0% (Target) to 94.5% (Opendoor).
  • At 7 of the 71 employers, more than half of the listings a job seeker sees today are older than 60 days. Opendoor scores highest on the index at 83; The Coca‑Cola Company lowest at 1.

1. How to read the score

Each employer gets three shares, each between 0 and 100, and the score is their plain average. It is deliberately unweighted: any weighting would be a claim about which signal matters more, and nothing here can establish that.

Stale share

Of the listings open today, the share that has been up for more than 60 days.

What a job seeker meets on the careers page right now.

Still up at 30 days

The share of listings still posted 30 days after they first appear, estimated over both closed and still-open listings.

Whether listings come down at all.

Repost rate

Of the listings we watched close, the share that closed while the same employer had another live listing under the same title.

Whether a closed listing was really a closed opening.

A high score means listings at that employer tend to be old, tend to stay up, and tend to come back under the same title. It does not mean the roles are fake: a hard-to-fill senior opening, an evergreen pipeline requisition and a listing nobody remembered to close all look the same from outside. A low score means the employer takes listings down promptly and rarely reposts a title — which is the behaviour a job seeker can rely on, whatever the reason.

2. The index, employer by employer

71 employers clear the publication gates: at least 20 listings we have watched close, enough of them for a survival estimate, and at least 10 listings open today. Employers below any gate are absent, not shown with a caveat. Scores run from 1 to 83; the median employer scores 34.

# Employer Score Stale share Up at 30d Half down by Reposted Open now Closures seen
1 Opendoor
83
90% 66% 48 days 95% 19 91
2 Brex
66
64% 100% not reached 33% 85 24
3 Walmart
66
51% 65% not reached 83% 333 64
4 Booz Allen Hamilton
64
21% 99% not reached 71% 1,388 761
5 Fortinet
63
48% 79% not reached 63% 83 27
6 Amazon Inc
62
33% 98% not reached 57% 617 30
7 SpaceX
62
58% 92% not reached 37% 788 273
8 Oracle
60
36% 79% 68 days 66% 568 360
9 Broadcom
56
36% 98% not reached 33% 228 45
10 Lam Research
55
33% 75% 53 days 58% 137 218
11 Adyen
53
63% 91% 53 days 7% 16 46
12 Nvidia
53
33% 99% not reached 29% 1,404 28
13 Amd
51
32% 98% not reached 24% 487 21
14 MongoDB
51
36% 74% 45 days 44% 28 48
15 General Dynamics
49
33% 87% not reached 26% 1,106 145
16 Affirm
47
42% 98% 68 days 2% 43 45
17 Balyasny Asset Management
47
45% 88% not reached 10% 47 42
18 Leidos
47
18% 79% not reached 44% 345 103
19 State Street
45
27% 96% not reached 11% 308 117
20 Stripe
45
46% 77% not reached 13% 145 93
21 Twilio
44
52% 68% 45 days 11% 27 44
22 DoorDash, Inc
43
44% 71% 50 days 13% 81 170
23 Arm Holdings
42
36% 88% not reached 4% 107 28
24 Datadog
42
51% 68% 47 days 8% 61 144
25 Chime
41
46% 73% not reached 3% 37 34
26 Reddit
41
39% 77% not reached 8% 77 76
27 Robinhood
40
28% 82% 64 days 9% 50 53
28 Anduril Industries
39
34% 55% 35 days 29% 916 1,700
29 Salesforce
39
14% 94% not reached 9% 167 33
30 Northrop Grumman
38
20% 78% not reached 16% 469 108
31 PNC
38
37% 60% 53 days 18% 131 61
32 GitLab
37
29% 80% 54 days 3% 52 60
33 Samsung Semiconductor
37
27% 74% not reached 10% 37 39
34 Block
36
33% 70% 44 days 6% 46 55
35 Citi
35
16% 69% 62 days 21% 268 139
36 Cloudflare, Inc
34
37% 62% 46 days 4% 100 154
37 JLL (Jones Lang LaSalle)
34
13% 51% 31 days 39% 24 130
38 JPMorgan Chase
33
22% 58% 39 days 20% 1,164 1,615
39 Lyft
33
28% 68% 57 days 5% 25 22
40 Samsung Electronics
33
20% 75% 54 days 5% 82 58
41 The Walt Disney Company
33
20% 54% 33 days 25% 118 142
42 Accenture
32
1% 87% not reached 7% 220 31
43 General Motors (GM)
31
26% 61% 49 days 7% 176 96
44 Electronic Arts
30
15% 65% 35 days 11% 39 101
45 Airbnb
29
40% 47% 28 days 1% 43 85
46 Elevance Health
29
8% 40% 28 days 39% 13 26
47 PayPal
29
2% 20% 14 days 66% 68 149
48 Qualcomm
29
2% 70% 48 days 16% 553 543
49 AT&T
28
24% 36% 14 days 23% 79 116
50 Dell Technologies
28
1% 64% 39 days 18% 69 44
51 Flex (Flextronics)
28
0% 66% 61 days 17% 26 24
52 Bristol Myers Squibb
27
3% 69% 56 days 10% 37 42
53 Target
27
17% 63% 48 days 0% 66 52
54 Cisco
25
8% 42% 26 days 25% 242 441
55 Microsoft
25
16% 24% 16 days 37% 599 1,844
56 SoFi
24
15% 53% 35 days 3% 13 34
57 McDonald’s Corporation
22
0% 39% 13 days 26% 15 23
58 GE Aerospace
18
0% 48% 26 days 7% 37 27
59 USAA
18
23% 27% 7 days 3% 43 78
60 Caterpillar
17
6% 18% 12 days 27% 135 317
61 CVS Health
16
3% 25% 13 days 22% 100 366
62 Humana
16
0% 28% 13 days 20% 37 92
63 IBM
16
0% 36% 25 days 13% 265 482
64 Marathon Petroleum
9
7% 14% 14 days 6% 15 50
65 US Bank
9
8% 18% 13 days 2% 51 207
66 Boeing
8
3% 15% 14 days 6% 68 278
67 FedEx
8
0% 0% 7 days 24% 11 67
68 Johnson & Johnson
6
3% 12% 14 days 2% 36 237
69 Wells Fargo
6
0% 6% 6 days 13% 20 429
70 Medtronic
4
0% 5% 11 days 7% 27 396
71 The Coca‑Cola Company
1
0% 1% 12 days 2% 16 132

Stale share is of listings open today; up at 30d and half down by are Kaplan-Meier estimates over both closed and still-open listings, with “not reached” meaning more than half outlive the observation window; reposted is of the listings we watched close. Each employer links to its own page, which carries the same figures with their sample sizes. Download this table (CSV).

3. How old the market’s listings are

The age of every live listing in the corpus today, measured from the date the employer published it. This is the whole corpus, not only the 71 employers in the table.

Older than 30 days 49.6% · 13,586 listings
Older than 60 days 26.6% · 7,283 listings
Older than 90 days 13.7% · 3,749 listings

The median live listing is 30 days old. A listing older than 60 days is what the index calls stale — twice the corpus median, so that the stale share says something about the employer rather than about the market.

4. The ends of each scale

Stale inventory. 90% of Opendoor’s 19 open listings are older than 60 days, with a median age of 150 days. At FedEx the stale share is 0% across 11 listings. The median employer in the table has 23% stale.

Survival. Brex still has 100% of its listings up at 30 days (109 listings observed). FedEx has 0% and takes half of its listings down within 7 days.

Reposting. 95% of the 91 listings we watched close at Opendoor came down while an identical title stayed posted; at Target it was 0% of 52. Pooled across all 71 employers the rate is 26.8%; the median employer reposts 13%.

Methodology and limits

The corpus is 27,368 live US technology postings collected from 300 employers’ own career pages — not from an aggregator, and not from candidate self-reports. Figures were computed on from closure history last updated , and refresh daily.

Signals, not verdicts. Nothing here observes whether an employer intends to hire. A long-lived listing may be a hard role, an evergreen requisition or a forgotten one; a repost may be two genuine openings under one title, and large employers hiring the same role in several places register here without doing anything questionable. The index ranks behaviours a job seeker can see, and names an employer only where the sample supports the figure.

This is a boutique sample, and it matters. 300 employers is a narrow slice of US tech hiring, weighted toward larger companies that publish structured career pages, and only 71 of them have enough closure history to appear in the table. The closure ledger has run since September 4, 2026 and the survival window opens on August 1, 2026, so these figures describe recent behaviour over a 68-day window and will get steadier over time.

A listing’s age is measured from the date the employer published it, or from our first sighting where no date is published. A few employers refresh the published date when they repost a listing, which makes such a listing read younger than it is — the error runs against the index, never for it. Survival is a Kaplan-Meier estimate over both closed and still-open listings, so a listing that is still up counts as having lasted at least as long as we have watched it; a plain average of closed listings reads far too short, and we withdrew one on 2026-09-04. Repost rate is over listings we watched close, not over the employer’s total hiring. Full definitions are on the methodology page.

Every number on this page is computed from the corpus at render time. Nothing is hand-entered, so the page cannot quietly go stale.

Download the data (CSV) How we measure this

Free to reuse, including commercially, with attribution to findrole.ai (CC BY 4.0). Questions, corrections or a data request: sales@findrole.ai

FindRole re-checks every one of these 27,368 listings against the employer’s site daily. Search them →