Apify + Google Sheets: Track Greenhouse Jobs via n8n

Track Greenhouse jobs and keep your Google Sheets register up to date with n8n + Apify

This n8n workflow continuously follows public Greenhouse job postings and automatically maintains a current Google Sheets register of new matches, updates, and confirmed closures—no AI service and no applicant data required.

What this workflow does

• Validates your watch setup before paid work: up to three boards, job filters (company/keyword/department/location/remote mentions), a stable watch name, a target Sheet destination, and a charge cap.
• Starts one pinned Apify Actor run, then polls that same run until completion. If the run fails, delivery stops—there’s no automatic paid retry.
• Guarantees coverage: it reconciles dataset counts and estimated Actor charges, with support for a maximum of 100 current jobs.
• Detects NEW vs repeats on first run (including older postings): repeats distinguish new matches, updates, unchanged jobs, and confirmed CLOSED or LEFT_FILTER results.
• Upserts deterministically using a watch-and-job key: existing rows are updated, while confirmed closure rows remain.
• Returns a delivery receipt: the sheet is a current register, not a permanent event history. Verified empty checks create no new rows.

Use cases

• Maintain a live hiring tracker for your SaaS team by company, keyword, department, and location (including “remote” mentions) in Google Sheets.
• Run scheduled monitoring for niche roles on Greenhouse and get automatic updates when jobs change or close.
• Operators needing a reliable, cost-controlled workflow to track up to 100 current Greenhouse listings.

Technical details

• Integrations: Apify (Header Auth with Authorization: Bearer <token>) and Google Sheets (OAuth2).
• Setup: create a Greenhouse Watch tab and import sheet-header.csv to preserve all 15 column names.
• n8n nodes used: if, set, code, wait, and http request (plus a sticky note for guidance).
• Execution guidance: start with the Airbnb remote engineering example, then repeat sequentially without overlapping runs.