Verification ledgerghost-index-v2 · 10 Sept 2026
Ghost Job Index
When an employer's own hiring system stops serving a role, does the employer's own careers feed stop advertising it? For most of the listings we can watch on both sides, the answer is no.
What this measures
Myro reads job listings from employer hiring systems directly, then re-checks each one at its source. Two signals come back: whether the role is still being served by the hiring system, and whether it is still present in the careers feed the employer publishes.
A listing counts here only when both signals are conclusive. The index reports what was seen, not why. An ad left up after a role closes may be an oversight, a slow feed, or a role that reopened without our seeing it. Read the method.
| Employer | Still advertised | Closed roles watched | Days up since close | Ads pulled | Days to pull |
|---|---|---|---|---|---|
| Palo Alto Networks | 100% | 23 of 23 | 7.8d | 0 | — |
| CRISIL | 78% | 405 of 521 | 50.1d | 116 | 50.1d |
| Godrej Consumer Products | 75% | 18 of 24 | 13.6d | 0 | — |
| DXC Technology | 70% | 19 of 27 | 3.7d | 0 | — |
| AB InBev | 68% | 17 of 25 | 11.6d | 8 | — |
| Genpact | 40% | 16 of 40 | 14.6d | 21 | 4.8d |
| Cisco | 29% | 39 of 135 | 5.8d | 95 | 0.7d |
| Accenture | 26% | 127 of 493 | 3.4d | 236 | 4.3d |
| AMD | 26% | 8 of 31 | 30.7d | 23 | 26.6d |
| Capgemini | 24% | 13 of 54 | 37.9d | 41 | 0.7d |
| Barclays | 19% | 53 of 276 | 11.2d | 223 | 4.8d |
| Cargill | 16% | 7 of 45 | 23.7d | 37 | 17.9d |
| Amazon | 9% | 5 of 55 | 0d | 50 | 18d |
| Aon | 8% | 2 of 24 | 12.5d | 22 | 25d |
| Carrier Global | 5% | 2 of 42 | 11.2d | 40 | 4.7d |
| 3M | 4% | 1 of 27 | 11.5d | 26 | 4.8d |
| ABB | 3% | 7 of 213 | 8.8d | 205 | 8d |
| Citibank | 2% | 3 of 126 | 2.6d | 123 | 10.9d |
| Deutsche Bank | 2% | 2 of 83 | 11.4d | 76 | 4.8d |
| DBS Bank | 1% | 1 of 160 | 11.4d | 148 | 4.7d |
| Aditya Birla Group | 0% | 0 of 50 | — | 50 | 32.1d |
| Adobe | 0% | 0 of 44 | — | 44 | 0.8d |
| Airbus | 0% | 0 of 28 | — | 28 | 4.8d |
| Sector | Still advertised | Closed roles watched | Days up since close | Ads pulled | Days to pull |
|---|---|---|---|---|---|
| Consumer & Retail | 69% | 60 of 87 | 18.2d | 20 | 4.3d |
| BFSI | 40% | 475 of 1,190 | 44.3d | 698 | 4.8d |
| Consulting & Professional Services | 28% | 181 of 648 | 7.4d | 326 | 4.3d |
| Technology | 27% | 106 of 393 | 20.1d | 263 | 4.8d |
| Diversified | 16% | 88 of 544 | 21.8d | 452 | 8d |
| Manufacturing & Industrial | 0% | 0 of 32 | — | 32 | 4.8d |
| Month closed | Still advertised | Closed roles watched | Days up since close | Ads pulled | Days to pull |
|---|---|---|---|---|---|
| 2026-07 | 77% | 462 of 600 | 49.9d | 138 | 50.1d |
| 2026-08 | 23% | 388 of 1,687 | 11.6d | 1,299 | 4.8d |
| 2026-09 | 13% | 91 of 704 | 4.6d | 412 | 0.9d |
What this does not cover
- 23 employers are published, of 171 with any closed role and 297 tracked in total. An employer appears only after 20 of its closed roles have been watched on both sides. Below that a share is noise, so it is withheld rather than printed.
- 57,850 listings reached no verdict at all. A hiring system that rate-limits our checks, or a listing address that never resolved, is not evidence a role ended.
- 10,715 roles are confirmed closed and 27,158 confirmed live. Only the closed roles that also appear in a careers feed can be judged here, which is why the table totals are smaller.
Employers: if a figure here is wrong, we will check it against the evidence and publish the correction. Write to hello@himyro.com.