Extract hiring signals from a careers page with Claygent
Converts any careers page into a structured hiring-signal record: open role count, departments, up to three notable titles, a GTM-hiring boolean, and a one-sentence signal summary. Clean JSON per row means you can filter, score, and merge hiring triggers into copy without parsing free text.
You are a hiring-signal analyst running inside an enrichment pipeline. Your job is to read a careers or jobs page and extract structured hiring intelligence that sales teams use as buying triggers. You know that a company hiring 3 SDRs signals outbound investment, a first Head of RevOps signals tooling budget, and 40 engineering roles signals a funded scale-up — and you report signals, never interpretations you cannot support.
Work from the careers page you are given (a URL to browse or scraped page content). If the page links to a hosted job board (Greenhouse, Lever, Ashby, Workable), follow it if you can reach it.
Extract only what is actually listed. Do not count expired postings if they are marked closed. Do not infer roles from boilerplate like 'we're always hiring'.
OUTPUT CONTRACT — follow exactly:
Return a single JSON object with exactly these keys:
{"open_roles_count": <integer>, "departments": [<strings, e.g. "Sales", "Engineering">], "notable_roles": [<up to 3 exact job titles that signal growth or new initiatives>], "gtm_hiring": <true|false — any open sales, marketing, or revops role>, "signal_summary": "<one sentence, under 20 words, stating the strongest hiring signal>"}
- Return ONLY the JSON object. No markdown fences, no commentary.
- If the page is unreachable, has no job listings, or shows zero open roles, return exactly: {"open_roles_count": 0, "departments": [], "notable_roles": [], "gtm_hiring": false, "signal_summary": "NOT_FOUND"}
- Never estimate counts you did not see. Never invent job titles.How to use it
- 1
Copy the prompt into Claude, ChatGPT, or any LLM — or into a Clay Claygent column.
- 2
In Clay, map a careers-page URL column as the input (build it with a formula like domain + '/careers' or find it with a preceding Claygent step).
- 3
Test on 10 known accounts and spot-check counts against the live pages before a full run.
- 4
Parse the JSON with Clay's built-in extraction to spin each key into its own column, then filter or score on gtm_hiring and open_roles_count.
Best practices
Route gtm_hiring = true accounts into a dedicated campaign that references the specific notable_roles title in the opener.
Re-run monthly — hiring signals decay fast, and a stale 'saw you're hiring SDRs' line is worse than no signal.
Hosted boards like Greenhouse are more reliable than marketing careers pages; when the prompt finds one, trust those counts over homepage claims.
Example: what this looks like in practice
An SDR team selling sales-onboarding software runs this across 2,500 Series A–B SaaS accounts. 640 come back with gtm_hiring = true; 210 of those list 3+ open sales roles. The team builds a campaign referencing the exact title from notable_roles — 'Six AE openings in one quarter usually means ramp time just became a board topic' — and books 22 meetings in three weeks. The zero-role JSON fallback also proved useful: 400 accounts with empty careers pages were deprioritized automatically, saving both credits and sender reputation.
Best fit
This prompt is one gear in a bigger machine. We orchestrate 20+ tools into outbound systems our clients own — and guarantee the results.
Apply for a Pilot Spot → →Frequently asked questions
They're one of the best. Hiring is public, current, and directly tied to budget: a company posting sales roles is investing in growth, one posting a first RevOps hire is buying tooling. Unlike funding news, which every vendor references the same week, careers pages are under-mined — a specific role reference still feels like real research.
More personalization at scale (clay & claygents) prompts
Generate a cold email first line from scraped research data
The workhorse of AI-personalized outbound: converts whatever research your enrichment waterfall produced into one 12–22-word observation-led first line per row. The swap-test rule and fact-invention ban are what separate lines that read researched from the AI slop prospects now delete on sight.
Turn a Claygent website scrape into a personalized one-liner
Produces a sub-25-word, website-specific opening line for every row in a Clay table — no human in the loop. The output contract forces plain text only and a NOT_FOUND fallback, so unreachable sites and thin homepages return a filterable sentinel instead of a hallucinated compliment that ends up in a prospect's inbox.
Analyze a pricing page for sales-relevant intelligence
Extracts a structured pricing profile from any pricing page: model, tier count, exact displayed prices, free-offer type, and a one-line read on their sales motion. At scale it segments prospect lists by how companies charge — which predicts budget, buying process, and which of your angles will land.