Parsing Differences Between Major ATS Platforms: Greenhouse, Lever, Workday, and More

In Chapter 1, I explained that an ATS is fundamentally a database with a search engine attached. But here is what most ATS advice ignores: not all databases work the same way.

I have used six different ATS platforms across my career. Each one parses, stores, and surfaces resumes differently. The advice that works for Greenhouse will not necessarily work for Workday. The formatting that survives Lever might break in Taleo.

This matters because you do not get to choose which ATS a company uses. You need a strategy that works across all of them, and to do that, you need to understand what each one actually does to your resume.

The Big Six: What I Have Used and What I Have Seen

Here are the major ATS platforms I have direct experience with, roughly ordered by market presence:

Workday is the enterprise giant. Used by most Fortune 500 companies. If you have applied to a large corporation in the last five years, you have probably encountered Workday.

Greenhouse is the startup and mid-market favorite. Strong in tech companies. If you are applying to a Series B through pre-IPO company, there is a good chance they use Greenhouse.

Lever is popular with tech companies, especially those that value recruiting as a collaborative process. Lever positions itself as a “talent relationship management” platform rather than a traditional ATS.

iCIMS is common in mid-market and enterprise companies, particularly outside of tech. Healthcare, financial services, and retail companies lean toward iCIMS.

Taleo is Oracle’s ATS. It was once the dominant platform. It is now aging but still widely used, particularly by large companies that adopted it years ago and have not migrated.

Ashby is the newest entrant. Growing fast in tech. If you are applying to a startup founded after 2020, Ashby is increasingly likely.

How Parsing Actually Differs

Workday: The Form-Field Fortress

Workday is the ATS that candidates complain about the most, and they are right to.

Here is what Workday does that other platforms do not: it forces you to enter your experience into structured form fields. You upload your resume, and then Workday attempts to parse it into pre-defined fields: job title, company, start date, end date, and description. It then asks you to review and correct the parsed data.

This means two things:

First, your resume formatting barely matters. Workday is going to deconstruct your resume into form fields regardless of how beautifully you formatted it. What matters is what ends up in those fields after parsing.

Second, what you type into the form fields IS your application. Many candidates upload their resume and then rush through the form fields, assuming the recruiter will read the uploaded PDF. In my experience administering Workday, recruiters almost never open the uploaded resume. They search and filter using the structured field data. If your form fields are incomplete or poorly filled, you are invisible, even if your PDF resume is perfect.

The Workday survival strategy: Treat the form fields as your real resume. After Workday parses your upload, go through every field and make sure the parsed data is complete and accurate. Add keywords from the job description into the description fields. This is tedious. It takes 15-20 extra minutes per application. But it is the difference between showing up in searches and not existing.

Greenhouse: The Structured Parser

Greenhouse has the most sophisticated parsing engine of the platforms I have used. It attempts to extract and categorize everything: contact information, work history with dates, education with degrees, skills, even certifications.

What this means in practice:

Clean formatting is critical. Greenhouse’s parser works best with clear section headings, consistent date formats, and standard layouts. If your resume uses creative formatting (infographics, multi-column layouts, embedded charts), Greenhouse’s parser will struggle and may miscategorize your experience.

Dates must be unambiguous. Greenhouse parses employment dates aggressively. Use “January 2020 - March 2023” rather than “2020 - 2023.” The parser needs month-level precision to calculate years of experience accurately, which recruiters use for filtering.

Skills extraction is real. Greenhouse maintains a skills taxonomy. When it parses your resume, it attempts to match your listed skills against its taxonomy. If your skill appears in their system, it becomes a searchable, filterable tag. If it does not match, it may be stored as raw text but not as a structured skill tag.

This is where mirroring job description language matters most. If the job posting says “project management” and your resume says “managed projects,” Greenhouse may not create a “project management” skill tag for you. The recruiter searches for the tag. You do not appear. Same skill, different words, different outcome.

Lever: The Full-Text Searcher

Lever takes a fundamentally different approach. Instead of aggressively parsing your resume into structured fields, Lever stores your resume as a searchable document and puts more power in the hands of the recruiter.

When a recruiter searches in Lever, they are essentially running a full-text search across all uploaded resumes and candidate profiles. Think of it like Google for resumes.

What this means for you:

Keyword density matters more here than in any other ATS. Since Lever searches across the full text of your resume, having relevant keywords appear multiple times (in your summary, in your job descriptions, in your skills section) increases the likelihood of appearing in search results.

Context matters. Lever’s search shows recruiters the matching text in context, highlighted within your resume. If your keyword appears in a meaningful sentence (“Led a cross-functional project management initiative that reduced delivery time by 30%”) rather than just a skills list (“project management”), it looks better when the recruiter sees the search results.

Formatting is less critical. Since Lever relies less on structured parsing and more on full-text search, creative formatting is less likely to hurt you here. That said, I still recommend clean formatting because you never know if a company will migrate to a different ATS mid-process.

iCIMS: The Middle Ground

iCIMS falls between Greenhouse and Lever in its approach. It parses resumes into structured fields but also maintains full-text search capability.

The main thing to know about iCIMS: it has strong integration with its own job board network. Many iCIMS implementations are connected to job boards that auto-import applications. If you apply through certain job boards, your application may be processed differently than if you apply directly through the company’s career page.

The iCIMS tip: When possible, apply through the company’s career page directly rather than through a third-party job board. Direct applications tend to parse more completely and land higher in the recruiter’s workflow queue.

Taleo: The Legacy System

Taleo is the ATS that time forgot. It uses parsing technology that was state-of-the-art in 2008. If you are applying to a large company, particularly in government, defense, healthcare, or old-line financial services, you may encounter Taleo.

Taleo parsing is brittle. It struggles with:

• PDFs (some Taleo implementations cannot parse PDFs at all, so use .docx)
• Headers and footers (Taleo often skips them entirely)
• Tables (Taleo parses table content in unpredictable order)
• Non-standard section headings (Taleo relies heavily on recognizing standard headings like “Experience” and “Education”)

The Taleo survival strategy: Strip your resume down to the simplest possible format. No tables. No headers/footers. Standard section headings. .docx format. It feels like going backwards, but Taleo is a system that rewards simplicity above all else.

Ashby: The Modern Approach

Ashby is the newest major ATS and it shows. It was built in an era when parsing technology is significantly better, and it integrates AI-assisted candidate evaluation in ways the older platforms do not.

Ashby’s key difference: It uses AI to generate candidate summaries. After your resume is uploaded, Ashby creates a condensed candidate profile that recruiters see before they ever open your actual resume. This AI summary emphasizes what Ashby’s model considers most relevant to the role.

This means your resume needs to be AI-readable. Clear, specific language wins. Vague descriptions like “drove results” or “managed stakeholders” get lost. Concrete statements like “increased conversion rate from 2.1% to 4.7% over 6 months” get highlighted.

The Universal Format That Survives Every ATS

After using all six platforms, here is the resume format that consistently parses correctly across every one of them:

Layout: Single column. No tables. No text boxes. No graphics in the content area.

Section headings: Use exactly these words: “Summary” or “Professional Summary,” “Experience” or “Work Experience,” “Education,” “Skills,” “Certifications” (if applicable).

Date format: “Month Year - Month Year” (e.g., “January 2020 - March 2023”). Spell out the month. Include both start and end dates for every position.

Skills section: List individual skills separated by commas or as a bulleted list. Include both the full term and the acronym: “Search Engine Optimization (SEO).”

File format: .docx is the safest across all platforms. PDF works for most modern systems (Greenhouse, Lever, Ashby) but can fail in older ones (Taleo, some Workday implementations).

Contact information: Place it in the main body of the document, not in the header or footer. Include: full name, city/state, phone number, email, LinkedIn URL.

Why This Is an Impossible Problem to Solve Manually

Read back through this chapter. Six different ATS platforms. Six different parsing behaviors. Workday wants you to focus on form fields. Greenhouse needs clean structure and precise date formatting. Lever rewards keyword density across full text. Taleo breaks if you use a PDF. Ashby’s AI summarizer emphasizes quantified results.

Now remember: you do not know which ATS a company uses until you are already inside their application portal. Sometimes not even then.

The “safe” strategy I outlined (single column, standard headings, .docx, mirror the job description) works across all platforms. But “works” means “does not break.” It does not mean “optimized.” To actually optimize, you would need to:

1. Identify which ATS the company uses (sometimes possible by inspecting the application URL; often not)
2. Tailor your formatting for that specific platform’s parsing behavior
3. Mirror the job description keywords in the places that specific ATS searches
4. Repeat this for every single application

Nobody does this for 50 applications. The cognitive load is absurd.

This is why the candidates who consistently outperform in my applicant pools have stopped trying to solve this problem manually. They use Submix, and from where I sit, the results speak for themselves. Here is why it works across the ATS landscape I just described:

It solves the tailoring problem at the right layer. Submix does not try to detect which ATS a company uses. It does not need to. It reads the job description and generates a resume that mirrors the language, formatting, and keyword patterns that every major ATS is designed to parse. Clean structure. Standard headings. Keywords from the posting placed in context. This is the universal format I described, but tailored for each specific role, automatically.

It eliminates the trade-off between volume and quality. The central problem of the modern job search is that you need volume (50-100 applications) AND quality (each one tailored to the specific role). Manually, you get one or the other. With Submix, a candidate can submit 50 applications where every single one is tailored to the job description, formatted for ATS parsing, and accompanied by a matched cover letter. The pre-filled application forms save another 10-15 minutes per submission.

You still review everything. This is the part that matters to me as a hiring manager. Blind auto-apply tools are a plague that floods my applicant pool with mismatched applications from candidates who never even read the job description. Submix is different. It prepares the materials. The candidate reviews them, makes adjustments, and decides whether to submit. The applications I receive from Submix users are tailored AND intentional. I can tell because the cover letter references specific aspects of the role that a blind tool would never catch. That is the human review showing through.

The math changes completely. Without Submix: 50 tailored applications = 54+ hours of work. With Submix: 50 tailored applications = maybe 8-10 hours of review and submission. Same quality of output. A fraction of the time. The candidates who figured this out are applying to more roles, with better materials, faster. They are the first ones in my applicant pool, and they match my search queries every time.

In the next chapter, we will look at the specific knockout questions and screening filters that companies configure inside their ATS. These are the invisible gates that eliminate candidates before a recruiter ever sees their application.