Oracle publishes no statement about columns, tables, text boxes or graphics in Taleo, which is most of what a typical Taleo format checklist is made of. What Oracle does document is narrower and far more useful: the supported file formats, an image-format trap, a size limit, a fixed list of extracted fields and one supported language. This page is that list, and it says so where Oracle is silent.
Quick answer
Upload a text-based file in a format Oracle lists as supported, written in English, under the 1,048,576 byte default attachment limit. Never upload a PNG or JPEG, because Oracle states a resume in an image format will not be parsed. Put your name, address, city, postal code, phone and email in the body, and give every role an employer, job function, responsibility line and start and end dates, and every qualification an education level, institution, program and dates, because those are the fields Oracle documents the parser extracting. Keep your metrics in the experience bullets, since no achievement field exists, and use the posting's exact words, because Oracle says keyword search provides exact matches. Oracle documents nothing about columns, tables, text boxes or graphics, so treat any confident layout rule you read as unsourced .
When you upload to a Taleo career section, the file is read by what Oracle calls a third party partner service, delivered as-is, and mapped to a fixed set of fields. Oracle publishes that set: first, middle and last name, address, city, postal code, home phone, email, place of residence and the pasted resume; education level, institution, program, start date and graduation date; employer, job function, responsibility and employment start and end dates. There is no achievement field, no bullet field and no metric field on that list.
Notice what else is missing. Oracle says nothing about multi-column layouts, tables, text boxes or graphics anywhere in the Taleo documentation. That silence is why this checklist looks different from every other Taleo format checklist you will find. The confident column rules circulating online are not sourced to Oracle, because there is no Oracle text to source them to, and a rule with no source cannot tell you whether following it changed anything.
What Oracle does state is concrete. Formatting is not carried through: resume parsing has no impact on the formatting of a text, bold, italics and bullets included. Headers and footers are readable, which is the exact opposite of the most repeated Taleo myth. Image-format resumes are accepted on upload and not parsed. Resume parsing is fully supported in English only. The default attachment maximum is 1,048,576 bytes across up to ten files.
So format for what is documented and stop there. The twelve rows below are the documented surface, each with the Oracle behavior behind it. The section order after them is where to put the content, and the read-back section after that is the step most candidates skip.
Each row is one formatting decision. The left side is what Oracle documents the Taleo parser handling; the right side is the version that fails, or the myth that sent you there. Row eight is the one other guides will not write: Oracle publishes no layout rule at all.
Save the resume in a format on Oracle's supported list: DOC, DOCX, RTF, TXT, HTML, PDF, ODT or WPD. Open the file and confirm you can select the text with your cursor before you upload it.
Anything outside that list, or a file with no real text layer. Oracle describes resume parsing as a third party partner service delivered as-is, so an unreadable file still produces an application record, just an empty one.
Upload the document, not a picture of the document. A text-based file is the only kind Oracle describes the parser reading.
A PNG or JPEG. Taleo Business Edition began accepting image uploads in 2024, and Oracle's release note states that the resume will not be parsed when it is in an image format. The upload succeeds, and nothing tells you the record it created has no searchable text in it.
Use whichever you already have. Oracle lists both among the supported formats and states no preference between them, so this is not the decision holding your application back.
Rebuilding the document because a guide told you Taleo requires DOCX. That rule has no Oracle source anywhere. The time is better spent on the language, the file size and the fields below.
Stay under 1,048,576 bytes, the default maximum for a candidate attachment, and under ten attachments in total. Some packages raise the per-file ceiling to 5 MB.
Stripping the resume down to fit under 100 KB. That number comes from the recruiter-side candidate file screen, not from the candidate upload, and a text-based resume is nowhere near the 1 MB candidate limit anyway.
Write the resume in English. Oracle lists resume parsing as fully supported for English and English U.K. only, and the career section setting that controls parsing scope defaults to English.
Submitting in one of the roughly twenty-five languages Oracle lists as not supported, or one of the seven it lists as limited, and assuming the parse worked. The file uploads cleanly either way.
Put your name, address, city, postal code, phone number and email in the body as plain lines. Those are the personal fields Oracle documents the parser extracting, alongside middle name and place of residence.
Rebuilding your layout because you read that Taleo cannot see headers. Oracle's own candidate management guide says resume parsing can detect text in the header and footer. Body placement is still the safer default, because the parser is third party and delivered as-is, but the myth itself is false.
Use them for the person who eventually reads the attachment. Oracle states that resume parsing has no impact on the formatting of a text, and names bold, italics and bullets specifically.
Expecting emphasis to carry meaning into the record. What Taleo stores is a set of plain field values plus searchable text, so a bolded skill and an unbolded one arrive identical.
Make your own layout call, and know you are making it without a source. Oracle's Taleo documentation says nothing about multi-column layouts, tables, text boxes or graphics, in either direction.
Treating a confident column rule as a Taleo fact. Every source that states one cites nothing, because there is nothing in Oracle's guides to cite. If a single column makes you feel safer, use one; just do not expect it to be the thing that rescues a failed application.
Give every role an employer, a job function, a responsibility line, and a start and end date. Those five are exactly the work experience fields Oracle documents the parser extracting.
Leaving the employer implied, the title invented, or the dates off one role. A field the parser cannot fill is a field a recruiter cannot filter you into.
State the education level, the institution, the program, and the start and graduation dates. Those five are the documented education fields, and they are separate fields.
A single compressed line such as BSc, 2019. Four of the five documented education fields have nothing to take from it.
Keep your numbers in the document. Oracle's search documentation lists the pasted resume and all file attachments among the content a candidate keyword search covers, so the text stays reachable even when no field holds it.
Assuming a quantified result lands in a structured field somewhere. Oracle's extracted field list has no achievement, bullet or metric field at all, and the older Legacy search setting narrows coverage further, to the three most recent attachments per candidate.
Use the posting's exact words, in the posting's own form. Oracle states that keyword search is considered to provide exact matches, with terms joined by AND unless the recruiter changes it.
Relying on the system to bridge a variant for you. A recruiter can add a trailing wildcard, or switch on a related-terms option that matches words sharing the first six letters, but neither is on by default and neither is your choice to make.
Sources: Oracle Using Legacy Recruiting 21B, Candidate Management; Oracle Implementing Career Section 25B; Oracle Implementing Recruiting 25A, Attachment Permissions and Settings; Oracle Implementing Recruiting 20A, Candidate and Requisition Search; Oracle Taleo Business Edition 24A What's New. Accessed 2026-09-03.
Oracle documents no heading detection and no section-order behavior, so nothing here claims the parser prefers this order. Each section is on the list because it feeds a documented extracted field, or because Oracle lists it among the content a recruiter's keyword search covers.
Name, address, city, postal code, phone and email as plain lines at the top. These are the personal fields Oracle documents the parser extracting, and body placement costs nothing even though Oracle says headers and footers can be detected.
Two or three lines carrying the terms the posting itself uses. Oracle documents no position weighting anywhere, so this is not a parser trick: it is the block a recruiter reads first, and text an exact-match keyword search can actually hit.
A plain list using the posting's own form of each term. The default search is exact match with terms joined by AND, so a variant you invented is a term the recruiter's search will not find.
Reverse-chronological roles, each carrying an employer, job function, responsibility and start and end date, because those are the five documented work experience fields. Your metrics belong in these bullets, since Oracle documents no achievement field to hold them.
Education level, institution, program, start date and graduation date, written out rather than abbreviated, because Oracle documents all five as separate extracted fields.
Named credentials and anything left over. Oracle lists career objectives and additional information among the content a candidate keyword search covers, so text here stays searchable even though no structured field holds it.
The scanner reads your resume the way a Taleo parser does, then shows the fields that come back empty, the exact terms the posting uses that your resume never says, and the bullets that need a number. Free, no signup to see the score.
Oracle specifies that the resume upload block must sit alone on its page, before the attachments, personal information, education, work experience and resume blocks. The order is deliberate: the upload pre-populates the form, and the form is still what you submit. Oracle also states that candidates are always shown the resume upload block when the education or experience blocks of a submission are empty, so this path is hard to avoid.
Here is the part almost nobody explains. Oracle documents that all fields mapped by the profile import service are populated whether or not those fields are displayed in the flow, and that where a field is not configured in the flow, the candidate will not have the opportunity to correct or edit that specific information. A mis-parsed employer, job function or date can therefore sit on your candidate record, reachable by a recruiter's search, with no screen that ever offers it to you for correction.
Quitting halfway does not undo it either. Oracle states that if candidates end the application process part way through, all valid data from the import is saved to their record whether or not they reached that page in the flow, though the application itself is still flagged as incomplete. The partial record stays behind. Note that the duplicate check runs before the candidate file is created, so a second attempt is matched against what the first one left.
Profile imports from LinkedIn or Indeed ride on the same block, so the same read-back applies to them. The practical discipline is short: upload a file the parser can read, then walk every visible field and correct it on the page where it appears, because that page is the only place you will get.
FAQ
The formatting questions candidates ask before an application goes into a Taleo career section, answered from Oracle's documentation and marked as undocumented where Oracle is silent. 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.
A text-based file in one of the formats Oracle lists as supported, written in English, under the 1 MB default attachment limit. Give every role an employer, a job function, a responsibility line and start and end dates, and every qualification an education level, institution, program and dates, because those are the fields Oracle documents the parser extracting. Then use the posting's exact words, since Oracle states that keyword search provides exact matches. Oracle documents nothing about columns, tables or graphics in Taleo, so no honest source can hand you a layout rule for it.
Oracle does not say. There is no statement anywhere in Oracle's Taleo documentation about multi-column layouts, tables, text boxes or graphics, in either direction. Guides that assert a column rule for Taleo are not citing Oracle, because there is nothing to cite. A single column is a reasonable default for any ATS and costs you nothing, but it is not a documented Taleo requirement, and it will not rescue an application that failed on file format, on language, or on the prescreening answers.
Any format on Oracle's supported list, as long as it has a real text layer you can select with your cursor: DOC, DOCX, RTF, TXT, HTML, PDF, ODT or WPD. Oracle states no preference between PDF and DOCX, so advice that Taleo requires one of them is unsourced. The genuine trap is an image. Taleo Business Edition began accepting PNG and JPEG uploads in 2024, and Oracle's release note says the resume will not be parsed when it is in an image format, so a photo or screenshot of your resume uploads successfully and extracts nothing.
Put them in the body, but not for the reason you have probably read. The common claim that Taleo cannot read headers and footers is false: Oracle's candidate management guide states that resume parsing can detect text in the header and footer. The reason to use the body anyway is that Oracle also describes resume parsing as a third party partner service delivered as-is, so detection is documented while correct field mapping is not promised. Body placement costs you nothing and removes the variable.
Oracle states that resume parsing has no impact on the formatting of a text, and names bold, italics and bullets specifically. In practice that means formatting is neither required nor harmful: what Taleo stores is a set of plain field values plus the searchable text of your pasted resume and your attachments. Format for the human who eventually opens the attachment, and do not expect a bolded skill to outweigh an unbolded one, because nothing in Oracle's documentation gives emphasis any weight.
Oracle publishes no length rule for Taleo, so any page count you read is a general resume convention rather than a Taleo behavior. The only documented size constraint is the file itself: the default maximum for a candidate attachment is 1,048,576 bytes, with up to ten attachments per candidate, and some packages allow up to 5 MB per file. A text-based resume of any sensible length sits far inside that. Write for the recruiter who reads the attachment, not for a limit Oracle never set.
Partly, and it is worth being precise about which part. Formatting decides whether your content is extracted at all: an image-format resume yields nothing, and a language Oracle does not support for parsing yields nothing. Past that floor, no scanner score comes from Taleo itself, because Taleo generates no percentage and Oracle documents ranking through the ACE, Required and Asset prescreening criteria instead. What a scanner score does measure is how completely your resume covers the posting, and that is moved by vocabulary rather than layout. Fix the file so it parses, then spend the remaining effort on the words the posting actually uses.
Because the upload seeds the form rather than replacing it, and Oracle delivers the parser as a third party service, as-is. Oracle specifies that the resume upload block must sit alone on its page before the attachments, personal information, education and work experience blocks, so whatever the parse produced is what the rest of the flow starts from. The part worth knowing is that Oracle also documents that all fields mapped by the import service are populated whether or not those fields are displayed to you, and that where a field is not configured in the flow, you will not have the opportunity to correct or edit that information. Fix every field you can see before you advance.
The pillar: the eight rules Oracle documents, why ACE prescreening criteria rank you instead of keyword density, and how Taleo differs from Oracle Recruiting Cloud.
The employer list verified from live apply URLs, the big names commonly mislabelled as Taleo, and how to check any company's ATS yourself in one click.
The fields Oracle documents the parser extracting, how the recruiter keyword search reads your attachment, and why prescreening decides the outcome.
A plain, text-first resume scored against a job description, with the keyword and metric fixes that moved the score, annotated line by line.
Get the parser-readable check, the fields that come back empty, the exact terms the posting uses that your resume never says, and a rewrite plan. Free to scan; no signup to see the score.
