Why Lever Strips Symbols and Icons on US Resumes

11 min read

Short answer: Lever strips symbols and icons because its parser reads your resume as plain text, not as a designed layout. Phone icons, checkmark bullets, stars, and emoji usually delete or turn into garbage characters, so contact lines break and skill lists import out of order. Your pretty Canva resume can look empty in the recruiter view even when the PDF looks fine on your screen. That's why you shouldn't trust the design file alone.

That's why your Lever upload looked perfect in Preview but the parsed profile showed a mangled phone number. If you're applying through Lever tonight, check your resume for free on a plain export before you trust the design template. You'll catch missing digits before a recruiter does.

Below: why the strip happens, what still breaks after you remove icons, exceptions worth knowing, and exact fixes with before/after lines you can copy.

Why Lever's parser removes symbols

Applicant tracking systems do not render your resume like a browser. They extract character streams and map them to fields: name, email, employer, dates, bullets. Icons and symbol fonts are not reliable characters in that pipeline. Lever's upload flow follows the same pattern most US startups use: text first, layout discarded.

When you use a phone glyph before your number, the parser may drop the glyph and keep digits, drop both, or attach stray bytes that break search. Same for envelope icons before email, LinkedIn logos inline, or checkmarks as bullet prefixes. The system is built to avoid storing unknown unicode, not to preserve your visual brand.

Screening reality: I've opened Lever profiles where the candidate's email was blank because an icon sat between label and address. They never knew.

This is not Lever punishing design skills. It is text extraction doing what text extraction does. PDFs that embed icons as vector art fare worse than simple typed text. DOCX with Wingdings-style bullets fare worse than standard list formatting.

Symbols also interfere with keyword matching. A skills row that reads "checkmark Python checkmark SQL" after parse is worse than "Python, SQL" typed plainly. Recruiters searching Python may miss a corrupted token.

Read Lever ATS formatting fixes for US tech roles for the wider parsing checklist beyond icons alone.

Unicode math symbols and fancy dashes cause the same strip behavior. Resume templates that use en-dash or em-dash characters between dates may import as blanks. Use a simple hyphen between month and year. Degree symbols and trademark glyphs in company names can split tokens recruiters search later.

Hyperlinks hidden behind icon buttons do not export URL text. Always show the full URL string for LinkedIn, GitHub, and portfolio links. Recruiters copy-paste from parsed view more often than you expect.

Lever applications at US startups often include custom questions. Your plain resume still matters because recruiters compare parsed history to your answers. If icons mangled your title, you may look overqualified or underqualified versus what you typed in the form.

Mobile-first resume templates with large icon headers are popular on template sites. They fail on Lever more often than desktop-first layouts because glyph density is higher above the fold. Strip the header first when troubleshooting.

If you work in design or marketing, keep a portfolio PDF with full visuals. Your Lever upload should still be plain. Interviewers who care about taste will ask for the portfolio link after the screen.

What breaks when symbols strip, and how to fix it

Contact block corruption

Icon-heavy headers are the top failure. Parsers read left to right. A row of glyphs plus text often becomes empty or merges fields.

Before: Header row with phone icon, email icon, map pin, and LinkedIn logo inline before each value.
After: Four plain lines: full name, city and state, phone digits, email, full LinkedIn URL on its own line.

Bullet lists that reorder or vanish

Custom checkmarks and arrows act like random characters. Some imports skip the line. Others prepend junk that pushes the first word of your bullet out of the experience block.

Before: Each bullet starts with a green checkmark character from a template pack.
After: Standard round bullet or hyphen; each line starts with a past-tense verb; no leading symbols.

Section headers with emoji or dividers

Emoji section titles like star EXPERIENCE star may parse as empty headings. Horizontal rules made of repeated dashes or icons can split employer names across blocks.

Before: "⚡ SKILLS ⚡" as a section title with decorative line art.
After: Plain heading "Skills" in bold, 12pt, same as other sections.

Rating stars and skill bars

Visual skill meters use repeated star or dot characters. After strip, you may get "Python" on one line and five blank squares on the next. Recruiters cannot tell your level and the ATS may not link skill to proficiency.

Before: Python followed by five filled star icons, SQL with four stars.
After: Skills line: "Python, SQL, AWS" in comma text, with proficiency shown in bullets where you used each tool.

Exceptions: what usually survives

Plain ASCII punctuation generally survives: hyphens, parentheses, commas, periods in URLs. Standard ampersand in company names usually survives. Bullet characters from Word's built-in list styles usually survive better than pasted unicode icons. Simple one-column layout without text boxes survives better than everything.

What to do right now on a Lever application

Upload a plain backup file. Open parsed preview when Lever shows it. Manually correct employer names and dates if fields shifted. Do not assume the PDF attachment saves you; many recruiters work from parsed view first.

Copy-paste plain resume pass

"Remove header icons; replace emoji headings; standard bullets only; no skill stars; one column; Arial or Calibri 11pt; export PDF from Word; upload; open parsed preview; fix phone and email fields; resubmit if mangled."

Before/after: two-column template with icons

Before: Left sidebar with icon list for skills, contact icons stacked, right column experience with arrow bullets.
After: Single column, contact on top in text only, skills as comma line or short list, experience with standard bullets.

Social and web icons in footers

Before: Footer row of Twitter, GitHub, and portfolio icons without spelled URLs.
After: "github.com/handle" and "portfolio.example.com" on separate lines in 11pt plain text.

International characters and accents

Accented names usually survive when typed as UTF-8 in Word. Problems spike when accents are rendered as special font glyphs or images. Type your legal name as text. Do not use stylized letterforms for the first heading.

When stripping helps you

Plain text forces clarity. Candidates who rebuild without icons often discover duplicate phone lines, broken date order, and skills listed twice. Treat Lever as a formatting audit, not an insult to your design sense.

Lever vs attachment: what recruiters open first

Some recruiters click the PDF immediately. Others sort and filter on parsed fields only. You must survive both paths. Plain upload improves parsed fields; clean layout still helps human readers who open the attachment from email forwards.

Referrals do not bypass parsing. Your contact may vouch for you, but the profile still lands in Lever with the same extraction rules. Plain file protects the referral path too.

Icon mistakes that still cost interviews

Trusting the PDF attachment alone. Recruiters often screen parsed data. Fix the plain version even if the PDF looks beautiful.

Using icon fonts from design marketplaces. They look crisp on screen and die in extraction. Use text.

Putting LinkedIn as an icon without URL text. Always spell the full https link. Icons do not click in parsed view.

Decorative vertical bars between sections. Pipes and bars split lines unpredictably. Blank space and bold headings are enough.

Re-uploading the same template with tweaks. If parse failed once, rebuild in Word plain, do not patch the Canva export again.

Assuming US rules differ on Greenhouse or Workday. Symbol strip is vendor-agnostic. Plain text wins everywhere.

See how resume styling impacts screening for layout issues beyond symbols.

Watermark backgrounds behind text. Low-contrast watermarks strip unevenly and leave noise characters in body text.

Infographic resume hybrids. Charts and icons as PNG blocks inside Word often OCR as garbage. Keep charts for the interview deck.

Multiple font families in one line. Stick to one sans font for body. Icon fonts masquerading as letters break extraction.

Scanned paper CVs uploaded as PDF. OCR misreads icons as random letters. Retype and export from Word instead of scanning paper.

Using special bullet characters from PDF printers. Print-to-PDF from browsers often embeds icon bullets. Save from Word or Google Docs with standard lists.

Verify the plain file before Lever submit

Run your stripped-down export through a parser check the same way you would for keyword match. You want to see contact fields populate and bullets stay under the right employer.

Build two exports from the same content: Lever-plain.pdf for upload, designed.pdf for email intros when a human asks directly. Never swap them by accident at 11pm.

If a friend's resume parses clean and yours does not, diff the files. Usually the difference is icons, columns, or text boxes, not writing quality.

Run a free ATS check on the plain PDF. Pair with a plain-text cover letter when the req allows upload; avoid icon headers there too.

Plain file wins on Lever

Why Lever strips symbols and icons comes down to text extraction, not aesthetics. Your fix is boring on purpose: plain contact lines, standard bullets, one column, preview before submit.

Save the designed PDF for humans you email directly. Upload the plain version to Lever every time. When parsed preview still breaks, strip one more layer of decoration and re-upload before you blame the market.

Symbols are optional for humans who already know you. They are liabilities for parsers that decide whether a stranger sees your file at all.

Run the same plain pass before every Lever batch apply session. Icons creep back in when you update templates from marketplace downloads. A quarterly audit costs ten minutes and prevents silent parsing failures.

Teach this rule to anyone reviewing your resume for free. Friends mean well when they say "make it pop." Pop is what parsers delete. Plain is what gets you searchable.

Bookmark this page before your next template purchase. If the preview thumbnail shows icons in the header, assume Lever will strip them and plan the plain export from day one.