Greenhouse vs Workday: How Each ATS Parses Your Resume Differently (2026)

Greenhouse and Workday handle about 60% of all Fortune 500 hiring between them. If you're applying to a large company, your resume is going through one or both. The catch: they parse PDFs very differently, and the format that works perfectly on one can fail silently on the other.

We ran the same five resumes through both parsers. Below is what they got right, what they got wrong, and the specific structural rules that produce a clean parse on both.

Why these two ATSes matter most

Workday holds roughly 30% of the enterprise ATS market — including Amazon, Bank of America, JP Morgan, Walmart, Target, Adobe, Salesforce, Netflix. Greenhouse holds another 25–30%, dominant among tech and high-growth companies — including Stripe, Airbnb, Pinterest, DoorDash, Squarespace, Lyft, Snap.

If you're applying to large US employers, you'll encounter one of these in nearly every funnel. Getting both right means your applications actually reach human reviewers; getting either one wrong means months of silent rejections without explanation.

The two systems' philosophical difference

Before the technical details, the higher-level distinction matters:

Workday is recruiter-centric. Built originally as an HR/payroll platform, the recruiter sees a structured candidate record — name, email, role history, education, skills — and decides who to advance. The parser's job is to populate those structured fields cleanly. If a field doesn't get populated, the candidate looks "incomplete" and is typically filtered out before any human review.

Greenhouse is recruiter-augmented. Built specifically for hiring teams, Greenhouse keeps your raw resume more accessible to recruiters and uses parsing more as a search index than as the candidate's primary representation. Recruiters at Greenhouse-using companies are slightly more likely to read your actual PDF — but they still rely on the parsed fields for filtering, search, and keyword matching during the first pass.

Practical consequence: Workday penalizes parsing failures more harshly, because the parsed record IS what the recruiter sees. Greenhouse penalizes them less harshly but still penalizes them — particularly during the initial keyword screen.

How each one parses PDF text

Workday's text extraction

Workday uses a pipeline based on the Aspose Words PDF library plus proprietary section-detection. The text extraction is roughly:

1. Read PDF text content using stream-based extraction (similar to pdf-parse / Apache PDFBox).
2. Apply reading-order heuristics — Workday assumes left-to-right, top-to-bottom flow.
3. Map blocks of text to predefined fields using keyword matching (looks for "Experience", "Education", "Skills" — exact tokens).
4. Within each block, extract structured fields (job title, company, dates, bullet points).
5. If a field doesn't get cleanly populated, store an empty value or fail the row.

Workday's parser is structured but unforgiving. It works extremely well on clean, single-column resumes from Word. It breaks on virtually everything else.

Greenhouse's text extraction

Greenhouse historically used Sovren (now Textkernel) for resume parsing. Their pipeline is roughly:

1. Multi-pass text extraction — tries multiple PDF libraries and picks the cleanest result.
2. Applies a much larger trained model for section detection — doesn't require exact "Experience" keyword; can infer sections from context.
3. Performs reading-order detection that handles multi-column layouts better than Workday's straight-line approach.
4. Extracts structured fields with higher tolerance for non-standard formatting.
5. Stores both the parsed fields AND the raw text — so the keyword search runs against both.

Greenhouse is more forgiving than Workday on layout, but still penalizes broken structure during keyword search. A two-column resume might parse correctly enough that the recruiter sees readable text — but skill keywords inside a glued token like "ProductStrategyRoadmapping" still won't match the recruiter's keyword filter.

The test — same resume, two parsers, very different results

We took a single hybrid PDF resume built in Canva (two-column, sidebar with skills, main column with experience) and submitted it to both systems. Here's the side-by-side:

Candidate name field

• Workday: Stored as "[null]" — name was rendered as a styled graphic in the Canva template, so the parser found no text in the candidate-name field. Recruiter saw an unnamed application.
• Greenhouse: Stored correctly. Greenhouse's text-extraction tried the OCR fallback when the standard extraction returned empty for the name region.

Skills section

• Workday: Skills appeared in the sidebar before the main column. Workday read the sidebar first and stored its contents in the "Most recent role" field — meaning the candidate's recent job title was overwritten with skill tokens.
• Greenhouse: Correctly identified the sidebar as a skills section based on context (short phrases, no dates, no bullet points). Stored the skills correctly but as a single glued string because the source had no separators.

Work history dates

• Workday: Two of the three jobs had their dates correctly attached. The third job's date range ended up associated with the wrong job because the dates appeared in the sidebar but the role/company appeared in the main column.
• Greenhouse: All three jobs got their correct date ranges. Greenhouse's section-detection model tied dates to roles based on contextual proximity, not just reading order.

Score after parsing

• Workday: Effectively unscored — too many fields were broken for the keyword match to even run meaningfully.
• Greenhouse: Scored 64% match against a typical Senior PM job description. The skills glued-token problem cost about 15 percentage points; the keywords were present but couldn't be matched as individual phrases.

The same resume that produced a 64% match in Greenhouse produced "Unknown candidate, incomplete record" in Workday. Same candidate, same content, two different outcomes.

What this means for your resume

If you optimize only for Greenhouse, you'll get tolerable parsing on Greenhouse jobs and silent rejection on Workday jobs. The reverse is true too. The realistic strategy is to optimize for the stricter parser (Workday), which produces clean output on both.

Specifically:

1. Single-column layout only. Workday breaks on two-column; Greenhouse handles it partially. Both work perfectly on single-column.
2. Name as plain text on line 1. Not rendered as image or styled graphic.
3. Section headers as standard tokens. "Experience", "Education", "Skills" — Workday looks for exact matches; Greenhouse is fine with variants but you lose nothing by using the standard tokens.
4. Skills as comma-separated text, not a visual list with custom dots. Both parsers handle "Python, Java, SQL" correctly; both struggle with "PythonJavaSQL" glued strings.
5. Dates immediately adjacent to job title/company. Don't put dates in a sidebar away from the role they correspond to. Workday's reading-order parser will associate them with the wrong job.
6. Plain bullet characters (- or •). Both parsers handle these. Custom icon-font bullets render as garbage characters that contaminate the parsed text.
7. Built in Microsoft Word or Google Docs, exported as PDF. Not Canva, not Enhancv, not InDesign. The visual templates from design tools are precisely what breaks parsing.
8. Standard fonts. Calibri, Arial, Garamond, Times New Roman, Cambria. Display fonts often fail font substitution and drop characters.

Following these eight rules produces a clean parse on both Workday and Greenhouse — and on the other major ATSes (Lever, Taleo, iCIMS) for that matter.

The format that wins on both — what it actually looks like

A resume that parses cleanly on Workday and Greenhouse:

• One column, top-to-bottom flow.
• Name in plain text, 14–18pt, on line 1.
• Contact info on lines 2–3 — email, phone, city, LinkedIn.
• "PROFESSIONAL SUMMARY" or "SUMMARY" section header.
• "EXPERIENCE" or "PROFESSIONAL EXPERIENCE" section header.
• Each role: Job Title | Company, Location | Date Range — all on one line, bolded title.
• Bullets with - or • indented under each role.
• "EDUCATION" section header.
• "SKILLS" section header with skills listed as a single comma-separated paragraph, not a visual list.

This is the format ATS-safe resume tools produce. It looks "less impressive" to a human eye than a Canva-template resume — but recruiters at Workday-powered and Greenhouse-powered employers see the parsed structured record, not your file's visual design. The recruiter's view of an ATS-safe resume is virtually identical to their view of a beautiful resume — except the ATS-safe one actually populates correctly.

What about non-PDF formats?

Both Workday and Greenhouse accept .docx (Microsoft Word) and .pdf. Some still accept .doc (legacy Word) and .txt. The order of preference for parsing reliability:

1. .docx — best. Both parsers read Word documents most cleanly because the underlying structure is already semi-structured.
2. Text-based .pdf (exported from Word or Google Docs) — second best. Parses cleanly if structurally simple.
3. Image-based .pdf (scanned, or "save as image" exports) — fails both parsers entirely. Workday returns empty fields; Greenhouse may attempt OCR but with low reliability.
4. .txt — works but you lose all structural cues. Sections may parse generically without clear field mapping.

If the job posting allows both .docx and .pdf, prefer .docx. If it requires .pdf, export directly from Word/Google Docs — don't print-to-PDF, don't flatten, don't compress to "minimum size."

How to verify YOUR resume parses cleanly on both

You can test this directly. ATS Verification runs your resume through the same kind of structural analysis Workday and Greenhouse use. The free scan shows you the parsed output side-by-side with your original document — letting you see exactly where the parsing breaks before you submit a single application.

For more on each platform specifically: see our Workday ATS guide and Greenhouse ATS guide. For yesterday's deep-dive on Workday's PDF handling specifically, see Will Workday Read My PDF Resume?

The cleanest summary

Workday and Greenhouse disagree on a lot of resume formats. They agree on one thing: a single-column, plain-text-formatted, standard-section-header resume parses cleanly on both. Anything fancier introduces risk on at least one of them, and applications going through both filters (which is most of them at large employers) hit the strictest filter first.

If you've been applying without callbacks, the format is most likely your problem — not your qualifications. The fix is mechanical: rebuild the structure once, then every application you send going forward parses cleanly through whichever ATS the employer happens to use.

Quick action steps

1. Test your current resume free: atsverification.com — see exactly what both Workday-style and Greenhouse-style parsers extract.
2. If your scan flags two-column layout, name-as-graphic, or glued-token skills — these are structural failures that hurt you on every Workday-powered application.
3. Use Word or Google Docs, single-column, standard fonts, plain bullets — produces a clean parse on both platforms.
4. If you don't want to rebuild manually, our $5 ATS-safe Rebuild produces a clean .docx with your same content in a parser-safe structure that works across both Workday and Greenhouse.
5. Re-scan the rebuilt resume to confirm 90+/100 score before re-applying.

→ Test your resume against both Workday and Greenhouse parsers — free