Greenhouse is the rare vendor that publishes its own list of what makes a resume fail to parse. It also publishes a 100 MB upload limit and, in a different article, a 2.5 MB parse limit. The gap between those two numbers is where applications quietly disappear, and almost no guide mentions either.
Quick answer
Get the file under 2.5 MB in doc, docx, pdf, rtf or txt, because Greenhouse accepts uploads up to 100 MB but states separately that it cannot parse anything larger than 2.5 MB. Then work through Greenhouse's own published failure list: one column, no tables, no graphics or photos or word art, contact details in the body rather than a header or footer or text box, and never an image of the document. Write employer names with Inc., Co., Ltd or LLC and job titles in full, both of which Greenhouse names specifically. Finally, answer the screening questions carefully, because Auto-Reject fires on those answers and never on your resume text .
Greenhouse is built around what it calls structured hiring, and the part you meet as an applicant is a job board, a resume upload and a set of questions the employer wrote. What makes it unusual among the systems in this cluster is candour: Greenhouse publishes a support article listing the specific things that make a resume fail to parse, and it is the same list of layout sins every other vendor leaves you to infer.
Two numbers matter more than any formatting tip, and they live in two different articles. Candidate uploads can be up to 100 MB. Greenhouse Recruiting cannot parse resumes larger than 2.5 MB. Nothing in the application flow tells you which side of that line your file landed on, so a heavy resume submits cleanly, sends you a confirmation, and arrives as an attachment nobody parsed. Meanwhile the guides ranking for this topic quote a 5 MB limit and a 2 MB limit, neither of which appears anywhere in Greenhouse's documentation.
On automation, both halves of the public argument are wrong. Greenhouse does document an Auto-Reject feature, which screens and rejects candidates based on their responses to custom application questions, alongside an Auto-Advance that can move you straight to a later interview stage. So the claim that every Greenhouse rejection is a human decision is false. But nothing in Greenhouse rejects you for keyword density either, so the opposite claim is equally false. The automated gate is the dropdown, not the document.
One caveat worth stating plainly, because nobody else does. The failure-list article sits in Greenhouse's recruiter-facing documentation, and the recovery it describes is a recruiter action. Greenhouse does not publish that the identical parser governs a candidate self-service application on a public board. It is the same product and one parser is the reasonable reading, but the eight rules below are Greenhouse's documented parser behaviour rather than a promise about your specific application.
Greenhouse announced on May 5, 2026 that it had agreed to acquire Ezra AI Labs, which runs structured, AI-led voice interviews with candidates on demand, producing role-specific evaluations with structured scores and transcripts for each candidate. For an applicant this is the most consequential thing in Greenhouse's 2026 roadmap: an AI that interviews you and scores the conversation, rather than one that reads your resume. Chief product officer Meredith Johnson framed it as every evaluation being explainable, with every hiring decision staying with the team.
These are not generic resume tips, and none of them is inferred. Each rule maps to something Greenhouse states in its own support documentation, which is why you will find a rule about columns here when the other guides in this cluster deliberately refuse to write one. For the full do-this and avoid-this checklist with a parser-safe template, see the Greenhouse resume format guide.
Greenhouse documents two different numbers in two different places, and the gap between them is where applications disappear. Candidate uploads can be up to 100 MB. But Greenhouse Recruiting cannot parse resumes larger than 2.5 MB. So a heavy file uploads successfully, confirms successfully, and arrives as an unparsed attachment. The 5 MB and 2 MB limits quoted in other guides appear nowhere in Greenhouse's documentation.
Its supported list is short and explicit: doc, docx, pdf, rtf and txt. Anything outside that is not a format Greenhouse says it accepts for a candidate upload, and an image of a resume is listed separately as a parse-failure cause rather than a format choice.
This is the most specific and least repeated rule Greenhouse publishes. Its own failure list names company names that do not include identifying words such as Inc., Co., LTD or LLC. The parser is trying to recognise an organisation, and a bare word it cannot confirm as a company is a field it may leave empty.
Greenhouse gives the example itself: Sr. Account Exec instead of Senior Account Executive is listed as a cause of an unsuccessful parse. Abbreviated titles read fine to a person and cost you the structured title field. Spell out Senior, Vice President, Engineer and the rest.
Unlike most vendors, Greenhouse states this rather than leaving it to inference. Its failure list names resumes with a columned layout, complex resumes with tables, headers and footers, and specifically resumes with the name and contact information in the header, footer or text box. Put contact details in the first lines of the body.
Greenhouse names all four: graphics, photos, word art, and resumes uploaded as an image rather than a document. It also notes that externally linked images are not loaded when it converts a resume to PDF for preview, though images embedded in the file itself are unaffected.
Resumes without clear sections and differing formats throughout is a documented failure cause. Greenhouse does not publish a required section order or heading vocabulary, so this is about internal consistency rather than matching a prescribed template: one heading style, one bullet style, one date style, applied throughout.
Greenhouse documents that the parser skips data it cannot verify as authentic, and names the pattern precisely: placeholder values like First Last, Company 1 or Client 1. Anonymised client names in a consulting history are the common real-world version of this, and they parse as nothing.
Sources: Greenhouse Support, Unsuccessful resume parse; Greenhouse Support, Supported formats for resumes, cover letters and other candidate uploads; Greenhouse Support, Resume parsing with non-English languages; Greenhouse Support, Application rules overview; Greenhouse Support, Talent Matching; Greenhouse Newsroom, Greenhouse Has Entered into a Definitive Agreement to Acquire Ezra AI Labs, 5 May 2026. Accessed 2026-09-03.
The scanner reads your resume the way an ATS parser does, then shows the fields that come back empty, the keywords the posting wants, and the bullets that need a number. Free, no signup to see the score.
Greenhouse boards can live at job-boards.greenhouse.io, but most large employers turn on the redirect that sends that URL to their own careers domain instead. Of seven confirmed Greenhouse employers we checked, six redirected away. The reliable check is the public board API, which Greenhouse documents as needing no authentication: boards-api.greenhouse.io/v1/boards/company/jobs. These two are verified that way:
SpaceX is the one employer here with a full company hub, in the SpaceX resume hub. Two traps to know about. The board token is often not the company name, so DoorDash is doordashusa and Anduril is andurilindustries. And a live board is not the same as an active one: at least one well-known company returns a valid board with zero postings. For the full verified list and the employers commonly mistaken for Greenhouse, read the companies that use Greenhouse guide.
The minimum-viable process for any Greenhouse application. The full HowTo is published in the page schema; the visible steps below are byte-aligned with it.
Greenhouse accepts doc, docx, pdf, rtf and txt, and allows candidate uploads up to 100 MB. Its parser stops at 2.5 MB. Check the actual file size before you apply, because a file between those two numbers uploads and confirms without ever being parsed.
One column, no tables, no graphics or photos or word art, contact details in the body rather than a header, footer or text box, and never an image of the document. Give it clear sections in one consistent style, and use your real employer and client names rather than placeholders.
Greenhouse lists company names without an identifying word such as Inc., Co., Ltd or LLC as a parse-failure cause, and gives Sr. Account Exec instead of Senior Account Executive as its own example of an incomplete title. Both are quick edits that most applicants never make.
Greenhouse documents Auto-Reject, which screens and rejects candidates based on their responses to custom application questions. That is the only automated rejection it publishes, and it fires on your answers rather than your resume text. Answer them accurately, then upload your resume to ResumeAdapter against the target posting to see your score and the terms the posting uses that your resume never says.
This is the check most guides get wrong about Greenhouse, and it takes ten seconds to run yourself. SpaceX is the clearest example: the human-facing board URL bounces you to spacex.com and mentions Greenhouse nowhere, while the public board API returns the entire live requisition set. Same employer, same day, two completely different answers.
GET job-boards.greenhouse.io/spacex 302 Found Location: www.spacex.com/careers (no mention of Greenhouse)
Six of the seven boards we checked behaved this way. Judged on the visible URL alone, you would conclude SpaceX does not use Greenhouse at all.
GET boards-api.greenhouse.io /v1/boards/spacex/jobs 200 OK "meta": { "total": 2287 } "Aerodynamics Engineer (Starship)"
Public, unauthenticated, and unambiguous. Greenhouse documents that job board data needs no authentication, so anyone can run this check on any employer in seconds.
FAQ
The questions candidates ask when a Greenhouse application goes quiet. Answers are byte-identical to the FAQPage JSON-LD, because AI engines that extract HTML and AI engines that extract JSON-LD should not see different text.
One feature scores candidates and it is off unless the employer turns it on. Talent Matching reviews your resume and your answers to the application questions, then sorts candidates into five named bands: strong match, good match, partial match, limited match, and needs manual review. It is not a percentage, it is available only on the Plus and Pro tiers, and a Site Admin has to enable it. Greenhouse states plainly that Talent Matching is assistive AI rather than automated decision-making, and that it does not automatically advance or reject candidates. Greenhouse also has its AI independently bias-audited every month by a third party, and publishes the results.
Yes, and this is where almost every guide gets it wrong in one direction or the other. Greenhouse documents a feature called Auto-Reject that screens and rejects candidates based on their responses to custom application questions, and a matching Auto-Advance that can move you straight to a later interview stage. Employers configure the rules, joining up to five questions with and-or conditions. What Greenhouse does not do is reject you for keyword density in your resume text. So the automated gate is real, and it sits on your dropdown answers about work authorisation, location, salary or years of experience, not on your bullet points.
Greenhouse publishes two numbers and the gap between them matters more than either one. Candidate uploads can be up to 100 MB, but Greenhouse Recruiting cannot parse resumes larger than 2.5 MB. A file in between uploads cleanly, gives you a confirmation, and lands as an attachment nobody parsed. Aim comfortably under 2.5 MB. The 5 MB and 2 MB limits you may have read are not in Greenhouse's documentation at all.
Yes, and unusually for an ATS vendor Greenhouse says so itself rather than leaving you to guess. Its published list of parse-failure causes names resumes with a columned layout, complex resumes with tables, headers and footers, and resumes with the name and contact information in the header, footer or text box. It also names graphics, photos, word art and resumes uploaded as an image. This is the one major system where the layout advice you have heard is actually vendor-documented.
There is nothing in its documentation about dates. Greenhouse publishes an explicit list of what makes a resume fail to parse, and no date-format item appears on it, nor any font requirement, nor any parse-accuracy percentage. Guides prescribing a particular date style for Greenhouse specifically are filling a gap the vendor left empty. Be internally consistent, because inconsistent formatting throughout is on the list, and spend the saved effort on the things that are.
No Greenhouse document claims it detects AI-written resumes, and its identity product does something different. Real Talent verifies who you are, not who wrote your application: candidates confirm identity through CLEAR using government ID authentication, a selfie check with liveness detection, and device metadata, on the Plus and Pro tiers. That is fraud prevention rather than authorship detection. Greenhouse's chief executive has publicly criticised the volume of AI-generated applications, describing an AI doom loop, but criticism is not detection.
It depends entirely on which companies you count, and the honest answer is more interesting than the headline. In Jobscan's 2026 sample of 13,130 companies where an applicant tracking system could be detected, Greenhouse appeared most often, at 19.0%, ahead of Workday at 17.5%, Lever at 15.6% and iCIMS at 14.4%. But that sample skews toward the employers job seekers actually scan postings for. In the separately collected Fortune 500 set, Workday leads at over 40% and no Greenhouse figure is reported. Greenhouse dominates among newer, smaller, faster companies rather than the largest ones.
Greenhouse is the applicant tracking system of the challengers rather than the incumbents. Employers confirmed on Greenhouse through its public board API include SpaceX, Anduril Industries, Stripe, Coinbase, Robinhood, Airbnb, DoorDash, Reddit, Discord and Oscar Health. The pattern is sharpest in defense: SpaceX and Anduril run Greenhouse while not one legacy prime does, with Lockheed Martin on IBM BrassRing, Northrop Grumman on Eightfold, and RTX and Boeing on Workday. The same split holds in finance, where the challengers are on Greenhouse and the incumbent banks are on Workday or Oracle. Be careful with Lever, which is the system most often mistaken for Greenhouse.
The twelve format calls that have a Greenhouse source behind them, the numbers other guides invented, and the 100 MB against 2.5 MB trap in full.
The eight verified employers with their board-API proof, the defense-tier split nobody documents, and how to check any company's ATS yourself in one click.
What Greenhouse documents about extraction, how Auto-Reject and Auto-Advance actually fire, and where Talent Matching fits in.
See a single-column, parser-safe resume scored against a job description, with the keyword and metric fixes that moved the ATS score, annotated line by line.
Get the parser-readable check, the requisition keywords your resume is missing, the bullets that need a number, and a rewrite plan. Free to scan; no signup to see the score.
