Remote Job History: hiring velocity, days-open and ghost jobs by company

Remote Job History is an Apify Actor that returns the remote-hiring history of tech companies: how many remote roles each company opened and closed every week, how long its roles stay open, and which of its open postings look like ghost or evergreen jobs, each with the reason. The data comes from 401 company career pages (Greenhouse, Lever, Ashby) and five remote job boards, checked every hour since 17 July 2026. Most job scrapers show what is open today; this shows what happened over time, which you can't scrape after the fact.

What does Remote Job History do?

Pick a mode:

• company (default): one result per company you name, with its weekly history, summary and suspected ghost jobs.
• market: one result with the weekly trend for the whole remote tech market.
• search: find a company's exact name (up to 50 matches), nearly free.

For each company it returns:

• summary: roles_total, open_now, closed_total, avg_days_open, median_days_open, ghost_open, first_seen, last_seen and tracked_since.
• weekly: one row per week with new_roles and closed_roles.
• suspected_ghosts: open postings that look like ghost or evergreen jobs, each with a reasons string such as "open 208 days (longer than 75% of roles that closed)" or "relisted 3× over 63 days", plus the posting URL.
• coverage: full or new-roles-only, so you always know how much the numbers can tell you (see below).
• definitions: what every field means, in every result.

Coverage, stated plainly

• Remote and remote-eligible roles only. "OpenAI has 503 open roles" means 503 open roles that can be done remotely.
• Full history for 296 companies tracked through their own applicant-tracking system, including OpenAI, GitLab, Canonical, ElevenLabs, Samsara, Grafana Labs, Twilio, Coinbase, Stripe, Ramp, Pinterest and Instacart. A careers page lists every open role, so when a role disappears it has closed; that gives closures, days-open and ghost signals.
• New roles only for about 2,100 more companies seen on remote job boards. Job boards show a rolling window, so a listing disappearing there doesn't prove the role closed, and we don't claim it did.
• History starts 17 July 2026 and grows every day. 319 of the 401 career pages were added on 14 October 2026; each company's tracked_since says when its history begins. Weekly numbers before that date only count roles that were still open then.
• The market trend uses a fixed panel of the 74 sources tracked since the first full week (20 July 2026), so adding sources never shows up as market growth. The result's panel field gives the panel's size and start.
• The data is rebuilt every night; data_as_of says when.

When should I use it?

• Recruiters and sales teams: see which companies are scaling remote hiring and which have quietly stopped.
• Job seekers and career coaches: check whether a company is really hiring and skip postings that have been relisted for months.
• Investors and analysts: hiring velocity is an early signal of growth or cutbacks.
• Journalists and researchers: a weekly, defensible measure of the remote tech job market.

Input

{
  "mode": "company",
  "companies": ["Stripe", "OpenAI", "Grafana Labs"],
  "weeks": 12,
  "includeGhosts": true
}

Field	What it does	Default
mode	company, market or search	company
companies	Company names, case-insensitive, up to 100 per run. Names we don't have are skipped and not charged	—
query	Part of a company name, for search mode	—
weeks	How many weeks of history to include (1–104)	12
includeGhosts	Include suspected ghost jobs with their reasons	true

Not sure of the exact name? Run search with part of it first (for example graf finds "Grafana Labs").

Output

One dataset item per company (shortened):

{
  "type": "company-history",
  "coverage": "full",
  "company": "Stripe",
  "summary": {
    "roles_total": 254, "open_now": 121, "closed_total": 138,
    "avg_days_open": 87.7, "median_days_open": 64.2, "ghost_open": 41,
    "first_seen": "2024-06-14T22:04:50+00:00", "tracked_since": "2026-07-17T13:52:28+00:00"
  },
  "weekly": [{ "week": "2026-10-12", "new_roles": 7, "closed_roles": 7 }],
  "suspected_ghosts": [{ "title": "…", "url": "…", "reasons": "open 208 days (longer than 75% of roles that closed)", "first_seen": "2026-03-31T22:06:45+00:00" }],
  "data_as_of": "2026-10-26T16:34:28+00:00",
  "definitions": { "closed_roles": "distinct roles that disappeared from the company's own ATS that week …" }
}

Export the dataset as JSON, CSV or Excel, or pass it to the next Actor.

How much does it cost?

You pay per result, and only for companies we have:

Event	Price
Company with full history (coverage: full)	$0.03
Company with new roles only (coverage: new-roles-only)	$0.003
Market trend	$0.02
Company search	$0.001

Names we don't recognise are skipped and cost nothing. Apify's free plan includes $5 of monthly credit, enough for about 160 full company histories.

FAQ

How do you know a role closed? Only from the company's own careers page (Greenhouse, Lever or Ashby), which lists every open role. When a role is no longer listed, it closed that week. Disappearances from job boards are never counted as closures.

What is a ghost job here? A posting that is open much longer than the company's roles usually stay open, or one that has been taken down and relisted several times. It's a signal with a stated reason, not a certainty; some evergreen postings are real.

Why are the numbers lower than the company's total job count? Only remote and remote-eligible roles are tracked. Office-only roles are not included.

Can I get history from before July 2026? No. The history is what was observed from 17 July 2026 on. Older weeks only contain roles that were still open when tracking began, and tracked_since marks where real tracking starts for each company.

Is any personal data included? No. Results contain company names, job titles, posting URLs and counts. There are no candidate or recruiter details.

Can my AI assistant use it? Yes. Apify exposes every Store Actor as a tool through its MCP server (mcp.apify.com), so Claude, Cursor and other assistants can call it.

Related

• Job Posting Parser: turn job ad text into seniority, job function, remote regions, salary and skills.
• Resume Parser + Job Match: score a resume against a job posting.
• keith-apis-mcp: an MCP server with job, resume and web page tools for AI assistants.