MEASURED: AI-generated editorial illustration, not real profiles, job results or a product screenshot.
I am Agnes Maina, and I built the Indeed actor and optional LinkedIn People Search actor linked here. MEASURED means read in documentation or cited sources, not a new live test. INFERRED marks recommendations and example arithmetic.
What a job post does and does not say
INFERRED: A job post is evidence of an advertised role. It does not establish current hiring intent, budget for your product, a vendor shortlist or a purchasing deadline. The listing could be stale. The hire could also be the company's way of solving the problem without buying anything. Treat the posting as a reason to research an account.
INFERRED: Start with a narrow question. If you help teams maintain a CRM, which roles mention CRM administration or cleanup? A broad search for every sales role gives you more rows, but it does not answer that question. Write down why the role matters before collecting it.
What the actor returns
MEASURED: The Indeed actor accepts keywords, location, country, an output mode and onlyNew. In companies mode, maxCompanies limits delivery and the actor groups collected matching jobs by employer. Three documented details matter:
-
openRolescounts collected matching results, not the employer's total vacancies. - Job descriptions are disabled in companies mode. Use jobs mode separately for available descriptions.
- Monitor history controls previously delivered results. A result new to your monitor is not necessarily newly published. Keep one schedule per configuration with no overlapping runs.
MEASURED: Console pricing read October 11, 2026 lists $10 per 1,000 hiring companies, $2 per 1,000 jobs with descriptions, $1 per 1,000 without, plus $0.00005 per start. Check the actor's current pricing before running.
INFERRED: Twenty chargeable company rows at $0.01 each imply $0.20, plus the start fee, or $0.20005. A separate run delivering 60 chargeable jobs with descriptions implies $0.12 plus its start fee, or $0.12005. These are price calculations, not actual bills, profit or measured results. They assume every row qualifies for its stated event and prices remain unchanged.
Steps
INFERRED: Build a research queue you can explain:
- Define three to six roles specific to your offer. Record why each one could be relevant.
- Enter one focused keyword search and location. Set
countrytous,outputtocompaniesandmaxCompaniesto20. - Inspect
openRolesand the job source links. A count is only as useful as the posts behind it. - For employers worth closer inspection, use jobs mode separately and read available descriptions. Note named tools, team responsibilities and process problems.
- Verify each role on the employer's own careers page. Mark anything you cannot confirm as unresolved.
- Remove duplicates. Exclude staffing listings and reposts where the actual employer is unclear or outside your target market.
- Record “observed posting” with its date. Do not automatically label the company “ready to buy.”
- Optionally use LinkedIn People Search to find public profiles by title, company and location. Public profile data is self reported, not verified employment. This is a manual follow-up option, not an integration built for this article.
- Review the evidence before deciding whether the account deserves further attention. This workflow stops at research; it sends no messages.
Worksheet
INFERRED: These rows are illustrative, not collected results.
| Company | Role observed | Source link | Observation date | Careers page checked | Description notes | Status |
|---|---|---|---|---|---|---|
| Example Co | Revenue operations analyst | Paste job URL | Oct 12 2026 | Role present | CRM cleanup named as a duty | Observed posting |
| Example Staffing | Same title, unnamed client | Paste job URL | Oct 12 2026 | Unresolved | None available | Excluded, employer unclear |
INFERRED: Description notes are where a signal becomes useful or falls apart. A title can match your keywords while the duties have nothing to do with your offer. Reading the description and checking the careers page can catch false matches before you spend more time researching the account.
INFERRED: To collect your first capped company sample, use the Indeed Jobs Scraper. The source checks and relevance decisions remain yours.
Top comments (1)
Agree with the core point: a posting is a reason to research an account, not buying intent on its own. Two things I would add from the people side. The hire that results from the posting is the stronger signal: the person starts, inherits a stack and has a mandate, and the start is dated on the profile, so a watchlist can be built on starts rather than on listings. And a posting can be stale, but a start date cannot.
I wrote up the job-change version of this watchlist today, with the counts by function and the 12-month window you need because start dates arrive late: dev.to/kevin_hall/a-job-change-sig...
Hope it is useful next to the job-post queue.