Observed Signal · Aug 2, 2026 · Technical Report · Source: DEV Community · Impact: 2/5 · Sentiment: Neutral
Workday job API pagination returns total as zero
A technical analysis of Workday's job API found multiple undocumented behaviors that break naïve pagination and scraping. The API returns a truthful total count on the first page (e.g., NVIDIA: 2000) but reports total: 0 on subsequent pages, which causes common loops to stop early and return only a small fraction of postings. The API enforces a strict page size of 20 (requests with limit > 20 return HTTP 400). The API URL cannot be reliably constructed from the public careers URL because it requires an additional host-label segment and variable data-center subdomain. Job postings returned by the API are sparse (title, externalPath, locationsText, postedOn, bulletFields). Measurements were done on 2026-08-02 across tenants including NVIDIA, Salesforce, Adobe, and Sony.
Practical impact on data ingestion and scraping pipelines for ATS/HR feeds: causes undercounting and forces many round trips; relevant to teams normalizing job data from enterprise Workday tenants.
Track Workday Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Workday job API returns a nonzero 'total' on the first page, then 'total: 0' on subsequent pages for multiple tenants.
- Measured tenants showing this pattern: NVIDIA (2000), Salesforce (1477), Adobe (832), Sony (94).
- The API enforces limit=20; requests with limit > 20 (e.g., 21, 25, 50) return HTTP 400.
- The API path requires a host-label segment not trivially derivable from the public careers URL; guessing parts often produces 404, 422, or other errors.
- Individual job postings payloads are sparse and typically include only: title, externalPath, locationsText, postedOn, and bulletFields.
Connected Companies & Entities
6 Entities mapped“Workday is where large enterprises actually post....”
“NVIDIA has 2,000 open roles there....”
“Measured on four enterprise tenants: ... Salesforce | 1477 | 0 | 0...”
“Measured on four enterprise tenants: ... Adobe | 832 | 0 | 0...”
“Measured on four enterprise tenants: ... Sony | 94 | 0 | 0...”
“Dell reported 0 at offset 0 as well, which is the third trap below rather than a fifth data point....”
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
Aggregators block scrapers; ATS job APIs are public
An analysis published on 2026-08-01 found that major job aggregators (Indeed, Glassdoor, ZipRecruiter, Upwork) often return 403 responses with empty bodies (blocked by CDNs/WAFs), while the original sources—company career boards served by Applicant Tracking Systems (ATS)—expose public JSON job-board APIs that answer anonymous requests with no API key. The author measured responses from Greenhouse, Ashby, Lever, Rippling, Workable, Recruitee and SmartRecruiters and found these endpoints return clear semantics (200 with jobs, 200 empty list for no openings, 404 for unknown board) in most cases, though SmartRecruiters behaves differently. Practical notes include the need to discover board tokens, differences in payload size when requesting full content (Greenhouse descriptions increase payload ~18x), and that some ATS endpoints publish compensation bands behind flags.
One API for Company Job Postings
The article demonstrates that most tech companies use a small set of Applicant Tracking Systems (Greenhouse, Lever, Ashby, Workable, SmartRecruiters, Recruitee) which expose public JSON endpoints. Because of that, job-posting collection is primarily a normalization problem rather than classic HTML scraping. The author shows an Apify actor (fetchbase job-postings-scraper) that auto-detects a company's ATS and returns a single normalized record per job, with honest nulls where sources omit fields like remote or salary. The piece highlights that some ATS boards (notably Ashby) publish fully structured compensation ranges (e.g., OpenAI’s Ashby board: 727 roles all include pay ranges, 460 flagged remote).
Developer Reaches 100 Users for Korean Data Scrapers
A developer publishing Korean data scrapers on Apify reports reaching 100 distinct accounts that ran at least one job against Korean sources (Naver, Melon, Musinsa, Daangn, Bunjang, Yes24 and others). The portfolio (described as 13 Korean scrapers) accumulated 14,541 total runs. Usage is concentrated: Naver Place-related scrapers represent 40 users, while the naver-news-scraper produced 10,942 runs from seven users. The author highlights different usage patterns (few high-volume automation pipelines vs. many low-volume researchers), emphasizes 7-day active users as the real engagement signal, and notes a long tail of niche scrapers. The post also mentions an upcoming Show HN submission and that the author will monitor 7-day active users for ongoing signal.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.