What Greenhouse actually stores about your application

Short answer: Greenhouse stores a candidate profile - your contact details, your uploaded files, every answer you typed into the form, the source you came from, and later the notes and scorecards the hiring team writes about you. It does not store a score, a rank, or a rejection decision made by software. The interesting part is which of those fields are searchable, which are visible to whom, and how long any of it survives after a rejection.

What Greenhouse pulls out of your resume

When you attach a file, the parser fills the fields on the form so you do not have to type them: name, address, phone number, company names, job titles, school names. That is the ordinary parse, and it exists to save you keystrokes. If the file is over 2.5MB it does not happen at all - Greenhouse says plainly that it "can't parse resumes larger than 2.5MB," and the usual cause of a file that size is an embedded portfolio image.

A second, different extraction happens if the employer uses Talent Matching, which is the feature that lets a recruiter search past applicants. That one takes six things: skills, years of experience, job titles, start and end dates of employment, company names from employment history, and a derived industry classification for your past employers. Greenhouse states the system "does not process special category data" such as racial or ethnic origin, health information, or political opinions. It also caches what it read: "Parsed resumes are cached for 30 days."

Read the list again and notice which one you are least likely to have thought about. Your job titles.

If you are a Product Designer whose business card says "Experience Architect," the searchable record says Experience Architect, and you will not appear when a recruiter types the title the market actually posts. This is the single most consequential field in the whole profile, and it is the one candidates edit least, because the parser filled it in and it looked correct. The Bureau of Labor Statistics counted 132,700 web and digital interface designers in 2025, with a median annual wage of $104,000 in May 2025 - the closest federal occupation to what companies advertise as product design. Nobody is searching a database for architects of experience.

The form fields, and what each one becomes

Every question on a Greenhouse application ends up somewhere specific. Here is the map for the boxes you actually fill in.

The 500 MB ceiling on attachments to custom application questions is not a typo, and it is two hundred times the resume parse limit. Those are different pipelines: one is a file a human will open, the other is a file a parser must read. A Product Designer can upload a genuinely large case-study PDF as a portfolio answer and still keep the resume itself small enough to parse.

What happens between Submit and a recruiter opening your file

The path a single Product Designer application takes through Greenhouse Recruiting.

1. 1
   You upload the resume

   The parser reads the file and fills the form fields - name, contact details, current title, company, education. Files over 2.5MB are not parsed.

2. 2
   You answer the questions

   Typed answers are saved in the Application section of your profile, exactly as written. Nothing rewrites or summarises them.

3. 3
   Demographic answers split off

   If the job post collects them, they are stored anonymized and are not attached to your name inside the account.

4. 4
   A profile is created

   Contact details, source, attachments and answers sit on one record, joined later by notes, emails, scorecards and rejection reasons.

5. 5
   People with different permissions open it

   An interviewer sees public contact information, the resume and cover letter, and their own assigned scorecard - not the rest.

6. 6
   The record outlives the decision

   On hire or rejection a retention timer may start, if the employer configured one. Deletion is a manual step someone has to take.

Who can actually see what you wrote

Not everyone in the account sees the whole profile. Greenhouse's permission documentation draws two lines through it, and the person sitting across from you in the interview is on the wrong side of both.

A private custom application question shows the question to every user but routes the answer to the two admin rows only. Those answers are excluded from a candidate export from the All Candidates tab, although they do flow into the Business Intelligence Connector. So if a form asks your current compensation behind a private question, the panel interviewing you for the Product Designer role genuinely does not see the number - but the reporting layer does.

What happens to the demographic questions

They are stored anonymized. Greenhouse's wording is that "All in-app candidate demographic data is anonymized for compliance purposes and to reduce unconscious bias in the hiring process," and that "Your organization's Greenhouse Recruiting users can only view anonymized candidate demographic data." Greenhouse Technical Support cannot access it either, "unless given explicit consent by your organization."

Anonymized is not the same as stored somewhere else, and it is worth being precise about the difference. Greenhouse's anonymization glossary lists "custom demographic question responses" and "EEOC question responses" under Personal info, alongside your phone number and address - the same group, the same record. What the anonymization does is strip the link before any user in the account can view them.

Which is why declining to answer those questions buys you nothing. The answer is not on the record the hiring manager reads.

Everything else the profile collects after you apply

The most useful document Greenhouse publishes for a candidate is one written for administrators: the glossary of data that gets wiped when a profile is anonymized. It is, in effect, the field list. It is longer than most applicants expect, and about half of it is generated by the company rather than by you.

The fields on a Greenhouse candidate profile

Groups and field names as Greenhouse's own anonymization glossary lists them. The middle column is who puts the text there.

	Filled in by	Fields Greenhouse lists in the group
Personal info	You	Name and pronunciation recording, pronouns, current company, current title, headline, phone numbers, email, social media, websites, address, education, employment, device information, custom…
Other details	Both	Candidate custom fields, location, tags, referral questions, follow-up reminders, attachments
Source	Your answer, then the system	Source, and who gets credit for it
Responsibility	The company	Recruiter, coordinator
Candidate / prospect process	Both	Job post questions, application custom fields, candidate stage data, scorecards and interviews, offer custom fields and all offer data, rejection reason, rejection notes, rejection questions,…
Real Talent	The company's tools	Fraud report, match score and reasoning, identity verification, AI opt out requests
Greenhouse Notetaker	The company's tools	Candidate opt out/in details, recording link, interview transcript link, AI interview notes
Activity feed	Both	Activity notes, emails, notes, prospect touchpoints, LinkedIn InNotes and InMail
Integrations	The company's tools	Third-party integrations

Two rows there deserve a second look. "Rejection reason" and "rejection notes" are fields, which means the reason someone passed on you is written down and kept - you will never see it, but it exists as text on your record. And the activity feed stores the emails, so a curt reply you sent about scheduling is part of the same profile as your portfolio.

Why nobody at Greenhouse can tell you where your application went

Because they are not allowed to look. Greenhouse's own support page for candidates says it "is solely a processor of candidate applications on behalf of our customers and will not share the status of any candidate application," and directs you to "reach out to the company to which you have applied."

The privacy policy makes the same point from the legal side: the employer is the controller, Greenhouse is the processor, and its customer contracts "require us to redirect inquiries relating to the Personal Information we process on behalf of a customer back to that customer without taking any action on the request." That is not an evasion. It means your record for a Product Designer role at one company is walled off from your record at another, even though the same vendor stores both. There is no cross-company file on you.

If you want the record deleted

You ask the employer, and you ask the specific employer, one at a time. There is no central switch.

What the company can do on their side is configure a data retention rule. The timer "starts when a candidate is hired or rejected on all job applications" and is "applied retroactively to all hired or rejected candidates." It resets if a rejected candidate is converted into a prospect for a different job. When the period lapses, Greenhouse does not erase anything automatically - "a specified user is notified that they should manually delete the data," and "Deleting a candidate's personal data must be done manually."

Here is what the documentation does not tell you. Greenhouse publishes what the fields are and what the retention machinery can do; it publishes no default number of days, and no indication of how many of its customers configure a rule at all. So the honest answer to "how long does my application sit there" is: as long as one company decided, and you cannot find that out from any page Greenhouse hosts. The vendor documentation tells you the shape of the system. It tells you nothing about the habits of the employer using it.

If someone inside the company referred you

The referral is a stored field, not a favour that happens in a hallway. Greenhouse's glossary lists "Source" and "who gets credit" as two separate items on the profile, alongside "referral questions" - so the system records both where the application came from and which employee is attached to it. That attribution usually drives a referral bonus, which is why it is tracked as its own field rather than folded into a note.

What this means for you is narrow and worth knowing. If a designer at the company told you to apply and you found the posting yourself on a job board, the credited source is the job board, and their referral may not exist on your record at all. Ask them to submit you through the internal referral form before you apply, or name them in the box that asks how you heard about the role. The field is filled once, at submission, and nobody goes back to amend it.

If the parser filled a field in wrong

Fix it before you submit, not after. The auto-filled fields are editable on the form, and whatever is sitting in them when you press submit is what gets stored - the parser does not get a second pass, and a recruiter is not going to correct your title for you.

The three worth checking every time, in order: current title, current company, and the employment dates. Talent Matching derives years of experience from those start and end dates, so a role with a missing end date or a format like "Spring '24" can contribute nothing to the total. A Product Designer with nine years of work can be invisible in a search for eight-plus because of how three dates were typed.

Here is the difference in practice. The parse of a two-column resume with the name in the header, for someone applying to a Product Designer opening:

Name:            (blank)
Current title:   Experience Architect II Skills Figma
Company:         (blank)
Dates:           Spring '24 – Present

And the same person, same career, after thirty seconds of editing the form fields by hand:

Name:            Dana Whitfield
Current title:   Senior Product Designer
Company:         Northgate Software
Dates:           03/2024 – Present

Nothing about the candidate changed between those two records. Only the second one comes back when a recruiter searches the account eight months later for a designer with five-plus years.

The one thing to get right

Edit the job title the parser filled in, and make it the title employers post rather than the title your company invented. It is the field Talent Matching indexes, the field a recruiter types into the search box, and the one that the auto-fill gets wrong most often - because it copies your company's language faithfully.

Then leave the rest alone. A clean single-column resume under 2.5MB, real answers in the custom question boxes, and the demographic questions answered or skipped as you prefer, since neither choice reaches the hiring team. If retyping the same handful of fields into every Greenhouse form is the part wearing you down, the platform fills them from your resume and answers the screening questions so the record that gets stored is the one you meant to store.