LinkedIn Jobs Monitor

Track newly observed LinkedIn jobs by keywords, location and target company filters. Keep a reusable hiring watchlist with original job links, source dates and saved history, so repeat scans avoid delivering the same job again.

This Actor collects public job-search cards directly. It does not call a paid source Actor and does not require LinkedIn cookies or a LinkedIn account. It also accepts an existing compatible Apify dataset. Public access can be blocked or incomplete; failed access is reported rather than presented as zero jobs.

Quick start

{
  "sourceMode": "public",
  "keywords": "Revenue Operations",
  "location": "United States",
  "postedLimit": "week",
  "companyIds": [],
  "mode": "new",
  "monitorName": "revops-hiring",
  "maxScanJobs": 25,
  "maxPages": 1,
  "maxResults": 25,
  "proxyConfiguration": { "useApifyProxy": false }
}

Run again with the same monitor name and search/filter settings to receive newly observed job IDs. Set emitInitial: false to save the first matching baseline without delivering job-result events. If you want recurring scans, add an Apify schedule yourself; the Actor does not create one automatically.

For a company watchlist, use numeric LinkedIn companyIds. You can obtain these from the f_C parameter of a LinkedIn Jobs search URL after selecting the company filter. Up to ten IDs are supported. These filters are passed to LinkedIn; the source can broaden search results, and unknown company IDs remain null in output.

Use includePhrases or excludePhrases to filter the collected title, company name and location using literal case-insensitive substrings. Matching evidence includes the actual words and offsets. This does not inspect a full job description or prove a company's purchasing plans.

What is included

• Job ID and available direct LinkedIn job URL.
• Job title, company name/link when available, and location.
• The source's listed date at its original precision and any relative date label.
• Observation and collection times, kept separate from the listed date.
• Literal filter evidence, source page provenance and a stable event ID.

This version reads search cards. Full descriptions, recruiter contact details, applicants, salary enrichment and applications are not collected. It does not publish messages or apply for jobs. “New” means newly observed by this monitor, not proof of a newly created vacancy. Changes to existing jobs do not create new-job events; missing jobs never imply that a role is closed or filled.

SUMMARY reports source mode, page/request counts, coverage and saved output. ERROR records failures. SOURCE_DATA retains the bounded raw observations from a successful collection for inspection. Apify datasets provide JSON, CSV and Excel exports.

Source and limits

The collector reads the public LinkedIn Jobs search page and, if requested, its public guest pagination used by LinkedIn's own website. It does not use an authenticated session. Defaults allow one page and 25 examined jobs; limits are at most four pages and 100 examined jobs per run. LinkedIn can return more cards on a page than the requested examination cap; unexamined cards are not saved as seen.

Results can be ranked, broadened, delayed, regional or incomplete. A successful bounded scan is not a complete job database. A repeat run with zero output does not prove no new jobs exist. A failed required page fails the collection before new history is saved.

Direct requests are the default. If they are blocked from your runtime, explicitly choose an Apify proxy in the input, for example the UNBLOCKER group. Proxy availability depends on your account, and its usage is billed separately. The Actor does not automatically switch to a paid proxy, purchase a third-party scraper, or change your plan. Each request has a timeout, and pages are bounded.

For existing data, use sourceMode: "dataset" and choose sourceDatasetId. Rows should include jobId or id, jobUrl or url, title, optional company: { name, url, id }, location, listedAt and postedAgo. Dataset mode reads the first maxScanJobs rows; it does not reapply live keyword/location/company/date filters. Literal include/exclude phrases still apply.

Pricing

$0.98 per 1,000 delivered job observations, plus a $0.00005 startup event per allocated GB of memory (minimum one event). Apify compute, storage and any explicitly enabled proxy usage are separate.

For example, delivering 25 job observations has a result fee of $0.0245, plus startup and actual platform/proxy usage. In new mode, previously delivered or rejected jobs have no new job-result fee, but a repeat scan can still incur platform/proxy costs.

maxResults limits delivered records. It does not cap request or proxy charges. The Apify maximum Actor charge limits this Actor's events, not every underlying platform cost. maxPages, maxScanJobs and requestTimeoutSecs bound collection work.

Saved progress

Qualified overflow is saved before output and drained on the next run before another search. Keep the same monitor name and settings. Keyword, location, company, window, source or phrase-filter changes create separate history; changing only the output cap does not.

A lock prevents concurrent writers. Delivery is at least once: a crash after output but before its saved checkpoint can repeat and charge the same eventId; deduplicate that field downstream. History never silently expires. At the configured history limit or 16 MB state limit, the Actor stops instead of forgetting jobs and replaying old results.