ATS & Screening

Demystifying the ATS: How Workday, Greenhouse, and Taleo Actually Read Your Resume

Most advice about Applicant Tracking Systems is either too vague to act on or based on myths that circulate on LinkedIn without any evidence. The reality is more mechanical and more fixable than the hype suggests: an ATS is primarily a database with a parsing layer. Your resume's job is to parse cleanly and contain the right text. That is it.

This guide explains exactly how the three most widely deployed ATS platforms — Workday, Greenhouse, and Taleo — handle your resume, and what formatting and language choices will help you clear the first filter reliably.

1. What an ATS Actually Does

When you submit a resume through a company's job portal, the ATS does several things in rapid sequence. First, it runs a parser — a piece of software that attempts to extract structured data from your document. The parser tries to identify fields like your name, contact information, current employer, job titles, dates of employment, education, and skills. This extracted data is stored in the database, not your raw document.

When a recruiter searches the system, they are querying that structured database — not reading your PDF. If the parser extracted your job title incorrectly, or missed a skill because it was in a text box the parser could not read, that information simply does not exist in the database. You are invisible for that search term.

The second function is scoring or ranking. Most platforms allow recruiters to define "must have" and "nice to have" criteria when they post a job. The system then scores incoming applications against those criteria and presents candidates in ranked order. Candidates who cleared the knockout questions and match the most criteria appear at the top. Candidates who do not clear a knockout question are often filtered out entirely before a human ever sees the application.

2. PDF vs DOCX — Which Wins

The honest answer is: it depends on the specific ATS and its version, and this changes over time. But the practical guidance based on 2025-2026 parser behavior is this:

• Workday handles both formats reasonably well but parses standard DOCX files more consistently than complex PDFs
• Greenhouse has strong PDF parsing and is generally reliable with either format
• Taleo (now Oracle Recruiting) has historically had weaker PDF parsing; DOCX is safer
• iCIMS, Lever, and SmartRecruiters generally handle PDF well when the document is simple

The format is rarely the deciding variable. What matters far more is document complexity. A clean, single-column PDF with standard fonts parses better than a visually complex DOCX with tables, headers, and footers. If you are unsure, submit DOCX to be safe — but strip it of any formatting that a parser might choke on.

3. Why Columns and Tables Fail

Resume parsers read documents in a linear sequence, typically top to bottom and left to right. When your resume uses a two-column layout, the parser often reads across both columns simultaneously rather than down each column separately. The result is that your text is extracted in the wrong order, jumbling job titles with dates, skills with employer names, or contact information with experience entries.

Tables have a related problem. The text inside a table cell is often parsed correctly, but the structural relationship between cells — the thing that makes a table readable to a human — is lost completely. A table used to line up your skill names with proficiency ratings becomes a flat list of text that the parser cannot interpret.

Other formatting elements that cause parsing failures:

• Text boxes and shapes — content inside these is frequently skipped entirely
• Headers and footers — critical contact information placed here is often not extracted
• Images and icons — completely invisible to parsers; do not put contact info inside them
• Non-standard fonts — can cause character encoding errors in extracted text
• Embedded hyperlinks displayed as icons — the URL is often lost, leaving a blank

The skills section example above also illustrates the keyword density point: spelling out the specific AWS services you know rather than just writing "AWS" gives you more parseable keyword surface area and more specificity for a recruiter reading your resume by eye.

4. How Workday, Greenhouse, and Taleo Differ

These platforms are not interchangeable. Each has distinct parsing behavior, different recruiter workflows, and different screening logic. Knowing which system you are applying through changes how you should optimize.

The single most consistent piece of advice across all platforms: fill out every optional field in the application form, even if you already uploaded a resume. Recruiters search by field value, not by parsing your document. If you leave the "Skills" field blank because it felt redundant, you disappear from skill-based searches entirely.

5. Keyword Matching vs Semantic Search

Older ATS platforms — and Taleo is the most prominent example — use exact or near-exact keyword matching. If the job description says "project management" and your resume says "program management," these systems may not connect the two. This is why mirroring the exact language from a job description has real value beyond just signaling alignment to a human reader.

Newer platforms, particularly Greenhouse and the more recent Workday versions, have started incorporating semantic matching — the ability to recognize that "built CI/CD pipelines" and "continuous integration experience" are related concepts. However, semantic search in ATS contexts is still far less sophisticated than it is in general-purpose search engines. Do not assume it will bridge large terminology gaps for you.

A practical approach that covers both scenarios:

• Use the job description's exact phrasing for the most important two or three skills or qualifications
• For secondary skills, use your natural phrasing but include common synonyms where space permits (e.g., "machine learning / ML")
• Never stuff keywords in white text or hidden sections — this was detectable in 2015 and is trivially caught today

6. The Formatting Rules That Matter

After stripping away the myths, the formatting rules that actually affect parsing come down to a short list:

• Single-column layout only. No sidebar, no two-column design, no skill bars in a second column.
• Standard section headers. Use "Experience," "Education," "Skills" — not "Where I've Been," "My Toolkit," or custom section names the parser has no mapping for.
• Dates in a recognizable format. "January 2023 – March 2025" or "Jan 2023 – Mar 2025" parse reliably. Ranges like "2023-25" or "Recent" often do not.
• No contact info in the header/footer area of the document. Put your name, email, phone, and LinkedIn URL in the body of the document, at the top.
• Bullet points using standard characters. The standard hyphen (-) or a simple bullet (•) parses correctly everywhere. Arrows, custom glyphs, and emoji bullets may render as garbled characters.
• File name matters slightly. Use FirstnameLastname-Resume.pdf or similar. Recruiters see the file name when they download it. "Resume_final_v3_USE THIS ONE.docx" signals disorganization.
• Keep it to one or two pages. Not because parsers care, but because every recruiter reading the parsed output in the ATS interface views a summary card — and a five-page resume does not display better than a two-page one in that interface.

7. ATS Checklist Before You Apply

The ATS is not your enemy. It is a database that a recruiter uses to find candidates. Your goal is to make sure your information is in that database correctly, and that the right keywords are present to surface you in the searches that matter. The formatting rules above get you there — the rest is making sure your experience and skills are genuinely strong for the role.