Greenhouse vs Lever vs Ashby: How Each ATS Parses Your Resume in 2026

TL;DR. Greenhouse, Lever, and Ashby all combine resume parsing with recruiter-configurable keyword filters, but they handle structure, custom fields, and screening differently. Optimize your resume for the strictest of the three (Greenhouse), and the same file will pass all three platforms without any per-ATS rewrites.

Why this comparison matters in 2026

If you apply to a tech role this year, there is roughly an 80% chance the first reader of your resume is one of these three Applicant Tracking Systems. Greenhouse alone powers Airbnb, Stripe, Coinbase, Shopify, Pinterest, and most YC-funded companies. Lever powers Netflix, Eventbrite, and Reddit. Ashby is the fastest-growing ATS in AI-native and developer-tools companies (Linear, Vercel, Replit, Notion).

These platforms all sound similar in marketing copy ("AI-powered talent acquisition"), and they all show recruiters the same kind of dashboard. But the way each one ingests your PDF, extracts structured data, scores it against the role, and surfaces it to a human is meaningfully different. Optimizing blindly for "ATS in general" usually means optimizing for none of them.

Side-by-side comparison

Greenhouse: the strictest parser

Greenhouse is the oldest and most conservative parser in the group. It ships with a battle-tested set of extraction rules tuned for the kinds of resumes recruiters were uploading in 2015, which means anything that looks designerly tends to flatten into a single text blob. Once your resume is a blob, downstream keyword matching collapses, and your application starts the funnel with a 0% match score regardless of how qualified you are.

How Greenhouse parsing actually works

1. The PDF is converted to text using a fixed reading order (top to bottom, left to right).

2. The text is segmented into sections by matching standard English headings: Experience, Education, Skills, Projects.

3. Each section is mapped to Greenhouse profile fields. Anything that does not match a known section becomes "Notes" and is ignored by keyword filters.

4. Recruiters configure required keywords per role. Your file is scored on presence and proximity of those keywords inside the mapped sections.

Optimize for Greenhouse by

• Strict single-column layout. Two columns produce interleaved garbage.
• Standard headings with the exact words above. No "Where I have built things."
• Real text, never images. If you cannot highlight and copy your name from your PDF, neither can the parser.
• Exact phrase match to the JD vocabulary. If the posting says "React.js," do not write "React" or "React 18."
• PDF exported from Word or Google Docs. Pages, Notion, and Canva exports often produce non-standard PDF structures.

Lever: more lenient layout, more screening questions

Lever is a younger ATS and its parser is noticeably more forgiving on layout. The catch is that Lever roles often include extensive screening questions ("Why are you interested in this role?", "Tell us about a time you shipped under a tight deadline"). These questions are scored, and incomplete or one-line answers commonly trigger auto-rejection long before a recruiter looks at your resume.

How Lever scoring actually works

Lever lets recruiters define a per-role scorecard with a mix of resume keywords and screening-question requirements. The scorecard then scores each candidate. Your resume might be parsed perfectly, but if your screening answers are short or off-topic, you will not be advanced. Lever also tracks every interaction in the funnel and surfaces drop-off rates back to the recruiter, which means your screening answers are read more often than you would expect.

Optimize for Lever by

• Apply the same structural rules as Greenhouse defensively.
• Take the screening questions seriously. Write 50 to 150 words per question, not one-line answers.
• Add portfolio, LinkedIn, and GitHub links explicitly in your resume header. Lever extracts these well and they show up in the recruiter view.
• Mirror the JD's vocabulary in your screening answers, not just your resume.

Ashby: newest, smartest, hardest on data quality

Ashby was founded in 2018 and is the youngest of the three by a wide margin. Its parser is ML-augmented and handles modern resume designs (icon rows, two-column profile sections, color blocks) better than the other two. The trade-off is that Ashby validates structured fields more aggressively and surfaces AI-suggested screening keywords directly to recruiters, which means a sloppy keyword match is more visible and more damaging.

How Ashby differs

• Ashby asks for structured fields (location, current title, years of experience, work authorization) and validates them against your resume. Mismatches show up to the recruiter.
• The platform suggests keywords to recruiters during scorecard setup, and surfaces near-misses ("candidate mentioned 'Postgres' but the role asks for 'PostgreSQL'") instead of silently dropping you.
• Modern resume designs are tolerated, but text must remain selectable. Image-based or heavily-stylized PDFs still fail.

Optimize for Ashby by

• Modern design is OK, but keep all text selectable and standard fonts.
• Be specific in your structured fields. Ashby validates these against your resume.
• Match the JD's exact terminology. Ashby's AI surfaces near-misses to recruiters and they read like keyword mistakes.

The universal rule

Build your resume for Greenhouse first. The same file that passes Greenhouse will sail through Lever and Ashby. The reverse is not true: a beautiful, two-column, icon-heavy resume that looks great in Ashby will be rejected outright by Greenhouse before a human ever sees it.

If you have only time to optimize once, optimize for the strictest parser. The cost of a clean single-column layout is purely visual; the cost of being filtered out by Greenhouse is the entire opportunity.

How Fursa handles all three

Fursa auto-submits to all three platforms (plus Workday, iCIMS, and Workable) using Playwright browser automation. The AURA pipeline tailors a resume per role and targets 90%+ ATS compatibility against a parser-aware scoring heuristic, so the same draft passes all three platforms without any per-ATS rewrites. You upload your resume once, and AURA produces a parser-clean tailored version for every job you apply to.

Quick decision table

If you are not sure which ATS a company uses, check the URL of the application page. Greenhouse hosts on boards.greenhouse.io, Lever on jobs.lever.co, and Ashby on jobs.ashbyhq.com. The URL gives you the answer in two seconds.