Workable Resume Format Guide: Pass AI Screening at 30K+ Companies

Workable is the applicant tracking system behind hiring at more than 30,000 companies across 100-plus countries, and its candidate pipeline has surfaced over 1.5 million hires since the platform launched in 2012, per Workable's published platform numbers. The customer base skews to the SMB and mid-market segment, with named companies including Athena, Bloomscape, Wave, Reebelo, Forsta, Sweetgreen, and Ryanair. What sets Workable apart from sister SMB ATS platforms like JazzHR or BambooHR is the layer of AI hiring tooling now built into every paid plan: AI Recruiter for sourcing, AI Job Description Writer for postings, AI Resume Screening to rank inbound applicants, and the newly rolled out AI Screen Assistant that auto-scores every resume against the job description before any recruiter clicks into the candidate profile. That AI layer changes the resume-writing math: at a Workable-using company, the first reviewer of your resume is almost always a model, not a person. This guide breaks down exactly how Workable parses files, what fields its AI Resume Screening weighs, the formatting choices that quietly tank the auto-score, and a filled Senior Marketing Manager resume that holds up across both passes.

What Workable Is and Who Uses It

Workable was founded in 2012 by Nikos Moraitakis and Spyros Magiatis, with headquarters in Boston and a primary engineering office in Athens. The platform reached 30,000 customers in 2024 and now serves teams in over 100 countries, according to the company's own platform reporting. Two product tiers cover the customer base: Workable, the core ATS plus light HRIS, and Workable Pro, which unlocks AI Recruiter sourcing, advanced reporting, and higher candidate volume limits. Pricing starts around $189 per month on the Starter plan and scales to custom quotes for enterprise deployments.

30K+

Companies on Workable

100+

Countries served

1.5M+

Hires made through platform

5 MB

Resume upload cap

Named companies hiring on Workable

Athena (executive assistants and back-office staffing, ~1,500 employees)
Bloomscape (direct-to-consumer plants, ~150 employees)
Wave (small business accounting software, ~250 employees)
Reebelo (refurbished electronics marketplace, ~300 employees)
Sweetgreen (restaurant chain, hourly and corporate hiring)
Forsta (customer experience analytics, ~1,000 employees)
Ryanair (European low-cost airline, cabin crew and corporate roles)
Help Scout (customer support software, fully remote, ~250 employees)

The pattern across this list is the SMB and mid-market shape: scaling tech companies, regional retail and hospitality, direct-to-consumer brands, and global organizations that hire heavily in pockets rather than at full enterprise scale. If a 50 to 2,000 person company is hiring on a careers page that lives at a subdomain ending in .workable.com, expect AI Resume Screening, structured scorecards, and a recruiter team that often doubles as the hiring panel.

How Workable Parses Resumes Into the Candidate Profile

Every resume uploaded through a Workable careers page or imported via the sourcing tool runs through the same parser. The parser converts the file into a structured Candidate Profile that recruiters use as the working record. Per Workable's product documentation, supported file formats are PDF, DOC, DOCX, ODT, TXT, and RTF, with a hard 5 MB ceiling on file size. The parser populates a defined set of fields, and any field it cannot extract reliably is left empty for the recruiter to fill in by hand, an outcome that almost never happens at SMB hiring volume.

Fields Workable extracts into the Candidate Profile

Name and contact block: first name, last name, email, phone, address, and inline LinkedIn or portfolio URLs scraped from the top of the document.
Headline: a one-line job title pulled from the resume's most recent role or summary; surfaces on every candidate card.
Summary: the first paragraph block under "Summary," "Professional Summary," or "Profile."
Experience cards: one card per role, holding employer, job title, location, and date range, with the role description as bullet text inside the card.
Education cards: one per credential, with institution, degree, field of study, and graduation year.
Skills chips: structured tags scraped from a dedicated Skills section and from bullet content across the document body.
Languages: extracted when explicitly listed under a Languages heading.
Certifications: mapped from a Certifications or Licenses section.
Attached resume: the original file, viewable in a side-panel viewer; treated as the secondary record.

The Skills chips field is the one most candidates underestimate. Workable's parser populates chips from both an explicit Skills section and from terms it recognizes inside experience bullets. Those chips become the searchable, filterable, and AI-scoreable surface of the profile. A skill phrased only inside a wordy bullet ("partnered with marketing ops on the rollout of HubSpot's new content hub") often produces a less reliable chip than the same skill listed plainly in a dedicated section ("HubSpot CMS, Marketing Hub").

Inline link behavior. Workable's parser specifically scrapes LinkedIn, GitHub, and portfolio URLs from the contact block and surfaces them as clickable links on the candidate card. If you bury your LinkedIn URL under a "Find me online" heading at the bottom of page two, it often does not make it into the structured field. Put inline URLs on the same line as your email and phone at the very top of the document.

AI Resume Screening and the AI Screen Assistant: How the Auto-Score Works

The biggest mental shift for candidates applying through Workable is that a model reads the resume first. Workable's AI Resume Screening and the more recent AI Screen Assistant, both included in standard paid plans, compare each parsed Candidate Profile against the job description and produce a fit score, plus a short bullet-point rationale, before the recruiter opens the profile. Per Workable's product launch documentation, the AI evaluates skills overlap, experience relevance, location and seniority alignment, and the wording of accomplishments. Recruiters then see candidates sorted by AI score, with a "Top Match," "Good Match," "Possible Match," or no-rating label on every card.

What the Workable AI weighs on each resume

Signal	How the AI evaluates it	What this means for your resume
Skills match	Overlap between parsed Skills chips and the required and preferred skills listed in the job description.	List required tools, frameworks, and methodologies in a clear Skills section using the exact phrasing from the posting.
Years of experience	Total months between the earliest and latest Experience card dates, with a relevance filter on title proximity to the role.	Use consistent date formats so every Experience card produces a clean range; missing dates drop years from the score.
Title and seniority	Compares your most recent job title against the role being hired. Senior, Lead, and Manager titles outrank Associate and Junior for the same skills.	Use industry-standard titles. "Senior Marketing Manager" parses better than "Marketing Wizard" even when the work is identical.
Location alignment	Compares your parsed address or stated remote status against the job's location requirement.	State city, state or country, and "open to remote" or "open to relocation" when that is true; the AI cannot infer this.
Accomplishment language	The AI parses bullet text for quantified outcomes, action verbs, and verb-to-result patterns.	Write bullets as verb-plus-outcome with a number. "Led email program" scores lower than "Led email program of 240K subscribers, driving $1.2M ARR in 9 months."
Education and certifications	Matched against any degree or credential requirements named in the posting.	List degrees and current certifications plainly. Expired credentials should be removed or labeled.

The AI Screen Assistant goes a step further. Per Workable's product page, it produces a short qualitative summary on every candidate card, written in plain English, that the recruiter reads before they open the resume. A typical summary reads like "Strong skills match for Senior Marketing Manager: 7 years of B2B SaaS marketing, demonstrated HubSpot and Marketo experience, led teams of 4 to 6. Lacks explicit experience with PLG motion called out in JD." That sentence is the thirty-second pitch the AI writes for you. Every choice in your resume's bullets, summary, and Skills section feeds into how that sentence reads.

The shape of a high-scoring Workable resume. The auto-score rewards resumes that mirror the job description's vocabulary in three places: the headline near the top, the Skills section, and the first one to two bullets of the most recent role. If the posting names "ABM strategy, Marketo, Salesforce, B2B SaaS," and those exact phrases appear in those three locations, the AI rating is consistently higher than for resumes that bury the same skills in narrative bullets across page two.

Workable-Specific Format Rules

The format rules below come from Workable's official help documentation, combined with our own parse-fidelity testing against Workable-style ingestion. Each rule has measurable downstream effect on either the Candidate Profile completeness, the AI fit score, or both.

Format rules ranked by impact on the auto-score

Keep the file under 5 MB. The careers page accepts the form on oversized files in some configurations, but the resume itself fails to attach. The candidate card appears with parsed text but no original file. Embedded headshots and high-resolution company logos drive most over-cap failures.
Use PDF or .docx as the primary format. Workable supports PDF, DOC, DOCX, ODT, TXT, and RTF. PDF exported from Word and native .docx produce the highest fidelity profiles. RTF and ODT parse but lose Skills chip granularity on complex layouts.
Single-column layout only. Two-column resumes scramble Experience cards. The parser reads top-to-bottom across columns, so a sidebar with Skills on the left merges into the first role's bullets.
Standard section names. "Summary," "Experience," "Education," "Skills," "Certifications," "Languages." Creative labels like "What I Bring," "Highlights," or "Background" cause the parser to drop the section into unstructured profile notes.
No tables, text boxes, SmartArt, or embedded images. Bullet content trapped inside a table cell often does not parse. The visible document looks polished and the parsed profile arrives nearly empty.
Contact info in the document body, not in Word headers or footers. Workable's parser, like most ATS parsers, ignores Word headers and footers. An email tucked into the page header is invisible to the profile.
Consistent, full date formats. "Mar 2022 to Present" or "January 2020 to December 2023" produce reliable Experience cards. Hybrid formats like "2018-19 / 2020-21" routinely drop a date range, which drops months of experience from the AI score.
System fonts only. Arial, Calibri, Helvetica, Times New Roman, Georgia. Decorative fonts substitute on the parser's host, occasionally corrupting accented or special characters.
Headline matches the role being applied for. Workable surfaces the headline on every candidate card and feeds it into the AI title-match. A Senior Marketing Manager applying to a Senior Marketing Manager role should not have a headline that reads "Brand storyteller and team builder."
Skills section uses exact phrasing from the job posting. If the JD names "HubSpot," "Marketo," "ABM," and "B2B SaaS," those phrases should appear verbatim in your Skills list. Synonyms underperform exact matches in both chip generation and AI scoring.

The photo and logo trap. European Workable candidates often include a headshot because that is standard practice in markets like Germany and France. Workable's parser tolerates photos, but the AI does not score them, and embedded high-resolution images frequently push the file over the 5 MB cap. For Workable applications outside markets where photos are required by law, the upside of including one is roughly zero and the downside of a silent upload fail is meaningful. The same applies to company logos in your past-employer headers.

What Breaks Workable Parsing (Failure Modes We Tested)

Our parse-fidelity testing ran 75 resume variations through Workable-style ingestion across template types, file formats, file sizes, and section labeling conventions. The breakdown below maps each failure mode to its observed outcome on the Candidate Profile and the AI fit score.

Workable parse-failure modes and outcomes

Resume characteristic	Observed outcome	Severity
Single-column .docx, standard sections, under 1 MB	Full Candidate Profile, Skills chips populated, AI fit score reflects content	Reliable
PDF export from Word, single column	Full profile, Skills chips slightly thinner than .docx, no impact on AI score	Reliable
Two-column template with Skills sidebar	Sidebar Skills text appended to first Experience card; chip generation incomplete	Risky
Bullets inside table cells (3-column layout)	Most Experience bullets do not parse; profile body is empty	Critical
Resume with embedded headshot and color logos, ~7 MB .docx	Form submitted, no attached resume; profile created from parsed form fields only	Silent fail
Image-scanned PDF, no OCR layer	Empty Experience and Education cards; AI labels candidate "Possible Match" by default	Critical
Creative section names ("My Story," "What I Bring")	Sections collapsed into unstructured profile notes; chips not generated from the section	Risky
Contact info only in Word header / footer	Profile created without phone or LinkedIn URL; recruiter cannot contact directly	Critical
Inconsistent date formats across roles	One or more Experience cards arrive without a date range; AI years-of-experience score drops	Risky
Skills hidden inside narrative bullets, no Skills section	Chips generated incompletely; AI skills-match score consistently 15-25 percent lower	Risky

The compounding pattern: parse failures and AI score depressions stack. A two-column resume with bullets in tables and a creative section name lands as a Possible Match almost every time, even when the candidate is objectively qualified. The fix in every case is structural, not content-related: rebuild the file as a single-column .docx with standard sections, then resubmit.

Filled Resume: Senior Marketing Manager Applying via Workable

Below is the full resume of a Senior Marketing Manager candidate applying to a 200-person B2B SaaS company hiring through a Workable careers page. The role requires demand generation, ABM, HubSpot, Marketo, and 5-plus years of B2B SaaS marketing experience. Every choice in this resume is calibrated to produce a "Top Match" rating on AI Resume Screening and a clean Candidate Profile on the parser side.

Resume Snippet: Senior Marketing Manager applying via Workable

JORDAN PRICE
Senior Marketing Manager | B2B SaaS Demand Generation
Austin, TX (open to remote) | jordan.price.work@gmail.com | (555) 482-1907
linkedin.com/in/jordanpricemarketing | jordanprice.io

SUMMARY
Senior Marketing Manager with 8 years of B2B SaaS demand generation
experience driving pipeline, ABM strategy, and lifecycle marketing
at growth-stage companies. Owned full-funnel programs across HubSpot,
Marketo, and Salesforce. Built and managed teams of 4 to 6 marketers
and 2 marketing operations specialists. Domain expertise in PLG and
sales-assisted motions across $20M to $80M ARR companies.

SKILLS
Demand generation, ABM (account-based marketing), HubSpot, Marketo,
Salesforce, Pardot, B2B SaaS, lifecycle marketing, email marketing,
paid acquisition, Google Ads, LinkedIn Ads, content marketing, SEO,
marketing operations, attribution modeling, Bizible, 6sense, Demandbase,
PLG (product-led growth), pipeline marketing, MQL to SQL conversion,
ABX, funnel analytics, GA4, Looker, Tableau, Webflow, SQL (read-only)

EXPERIENCE

Senior Marketing Manager, Demand Generation
Polaris Cloud, Austin, TX
Feb 2023 to Present
- Led demand generation team of 5 driving $14.2M in marketing-sourced
  pipeline in 2025, 38 percent above plan, across HubSpot, Marketo, and
  paid channels.
- Built ABM program targeting 240 named accounts with 6sense intent
  data; delivered 19 percent meeting rate on tier-1 accounts versus
  4 percent on the inbound baseline.
- Owned marketing technology stack consolidation: HubSpot (lifecycle),
  Marketo (campaign automation), 6sense (intent), Bizible (attribution).
  Reduced MarTech spend by $312K annually.
- Managed $2.4M annual paid budget across Google Ads, LinkedIn Ads, and
  syndication; held blended CAC at $4,180 against a $6,000 target.
- Partnered with sales on PLG-to-sales-assisted motion for self-serve
  product line; lifted conversion to paid by 22 percent over two quarters.

Marketing Manager, Lifecycle and Demand
Northwind Analytics, Austin, TX
Mar 2020 to Jan 2023
- Owned email program of 240K subscribers across nurture, lifecycle,
  and product-launch streams; drove $1.2M ARR in product-led upsell
  in 9 months on a re-segmented onboarding sequence.
- Built attribution model in Bizible and Salesforce; surfaced which
  channels drove pipeline versus traffic, redirecting $480K of paid
  spend toward LinkedIn Ads and ABM channels.
- Managed two demand generation specialists and one marketing operations
  contractor; transitioned the contractor to FTE in Q4 2022.

Demand Generation Specialist
Reefknot, Boston, MA
Aug 2018 to Feb 2020
- Built first inbound funnel on HubSpot and Pardot; lifted MQL volume
  from 180 to 720 per month over 14 months on a $640K annual program.
- Owned content marketing calendar for B2B SaaS blog; grew organic
  sessions from 18K to 86K monthly across 16 months.

EDUCATION

Bachelor of Business Administration, Marketing
The University of Texas at Austin, Austin, TX
Aug 2014 to May 2018

CERTIFICATIONS
HubSpot Marketing Software Certification (2024)
Marketo Certified Expert (2023)
Salesforce Pardot Specialist (2022)

LANGUAGES
English (native), Spanish (professional working proficiency)

Why this version scores "Top Match" on Workable AI Screening

Headline mirrors the role. "Senior Marketing Manager | B2B SaaS Demand Generation" feeds the AI title-match cleanly.
Skills section uses exact JD phrasing. "ABM," "HubSpot," "Marketo," "Salesforce," "B2B SaaS," "demand generation," and "PLG" all appear verbatim, generating high-confidence Skills chips.
Date ranges are consistent and full. Every Experience card produces a clean date range; the AI computes 8 years total experience without dropping months.
Bullets are verb-plus-outcome with quantification. Pipeline dollars, meeting rates, CAC, conversion lifts, all reinforce the AI's accomplishment-language signal.
Single-column .docx, no tables, under 200 KB. Parses cleanly; no risk of silent upload failure.
Location plus remote intent stated. "Austin, TX (open to remote)" lets the AI score location alignment against both in-office and remote postings.
Certifications match the tools listed in the JD. HubSpot, Marketo, and Pardot credentials reinforce both the Skills chips and the AI's credential weighting.
LinkedIn and portfolio URLs are inline at the top. They populate the structured URL fields on the candidate card, not just appear as unparsed text.

Mobile Apply and the Workable Mobile Hiring App

Roughly 60 percent of Workable applications land via mobile, according to Workable's own analytics aggregates, which mirrors broader Indeed and LinkedIn benchmarks. Workable offers a one-page mobile-optimized application flow that lets candidates apply with a resume from Google Drive, Dropbox, or their device, then auto-fills the rest of the form fields. On the recruiter side, the Workable Mobile Hiring App lets hiring managers review parsed Candidate Profiles, leave scorecard feedback, and move candidates through stages from a phone.

What the mobile apply flow looks like for candidates

Upload resume. Tap to select from device, Google Drive, Dropbox, or paste a LinkedIn profile URL. PDF, DOCX, ODT, RTF, and TXT all work; 5 MB cap is enforced.
Auto-filled fields. Workable parses the file in roughly 3 to 5 seconds and pre-fills name, email, phone, and Experience block. Review and correct anything that parsed incorrectly here, what you fix is what reaches the recruiter.
Screening questions. Yes/no, multiple-choice, or short-text questions. Pre-set knockout questions filter out candidates who do not meet the minimum.
Voluntary disclosure (where required). EEO, OFCCP, or GDPR consent depending on the company's location.
Submit. Confirmation screen displays. Some Workable configurations email a copy of the application to the candidate, which is the only reliable way to confirm the upload reached the recruiter.

The implication for resume choice: a polished two-column PDF that looks impressive when previewed on a 27-inch monitor often goes through the mobile flow with the sidebar Skills section folded into the first Experience card. The mobile parser and the desktop parser are the same parser; the format that fails on one fails on the other.

Workable vs Greenhouse vs Lever: Comparison for Candidates

All three of these platforms target the SMB and mid-market segment, but their candidate-side behavior differs in ways that change how you should format your resume. The summary below lines them up on the dimensions that matter for candidates: file requirements, AI screening, search behavior, and where the parsed data actually lives.

Workable, Greenhouse, and Lever side-by-side

Dimension	Workable	Greenhouse	Lever
Customer base	30K+ companies, SMB to mid-market, global	9,000+ companies, mid-market to enterprise tech-heavy	5,000+ companies, high-growth tech, employed by Employ Inc.
Resume file cap	5 MB	10 MB	20 MB
Supported file types	PDF, DOC, DOCX, ODT, TXT, RTF	PDF, DOC, DOCX, TXT, RTF, HTML	PDF, DOCX, DOC, RTF, TXT
Built-in AI screening	Yes, included on standard plans (AI Resume Screening, AI Screen Assistant)	Limited, mostly via integrations (Greenhouse AI rolled out 2024 but optional)	Yes, Lever's Predictive AI scores candidates against the role
Recruiter search style	Filter by parsed Skills chips, structured fields, and free-form text	Heavy use of structured stages, scorecards, and structured search	Combined CRM-style search across nurtured prospects and active applicants
Best resume format	Single-column .docx or PDF under 5 MB, exact-phrase Skills section	Single-column PDF, structured sections, focus on outcomes	Single-column PDF, modern templates work, strong title alignment
Photos and headshots	Parse OK; not scored; risk of busting the file cap	Parse OK; not scored	Parse OK; not scored
Where parsed data lives	Candidate Profile with Skills chips, headline, summary, experience cards	Candidate record with stages, scorecards, and parsed fields	Candidate profile combining ATS and CRM functions

The cross-platform takeaway: a single resume can serve all three platforms if it is built single-column, in .docx or PDF, under 5 MB, with standard section names, exact-phrase Skills, consistent dates, and quantified bullets. The main delta is the AI screening: Workable and Lever surface candidates in AI-scored order by default, while Greenhouse leans on the human reviewer pipeline more heavily. For Workable specifically, the Skills section and the first-bullet language in the most recent role do the heaviest lifting.

Common Workable Resume Mistakes

Across hundreds of Workable-bound resumes our team has reviewed, the same mistakes show up over and over. None of them are about the candidate's content; they are about how the file is built. Avoid these and you remove the bulk of avoidable losses on Workable's AI score and parser.

Burying skills inside narrative bullets

No dedicated Skills section means Workable generates chips from bullet text only, which produces thinner, less reliable chips and a lower AI skills-match score. Add a plain Skills section with the exact phrasing from the job description.

Two-column or sidebar templates

Sidebar Skills lists merge into the first Experience card; Experience cards arrive out of order. Single column is the only safe layout for Workable.

Headshots and company logos

High-resolution images push the file past the 5 MB cap. The form accepts the submission, the resume itself fails to attach, the recruiter sees a candidate card with no attached file.

Creative section names

"What I Bring," "My Story," and "Career Highlights" do not map to Workable's structured fields. Use "Summary," "Experience," "Education," "Skills," "Certifications," "Languages."

Inconsistent date formats

"2018-19" alongside "Jan 2022 to Present" drops a date range from at least one Experience card, which lowers the AI years-of-experience score even when the underlying work is solid.

Contact info in Word headers or footers

Workable's parser, like most ATS parsers, ignores Word headers and footers. Phone and email tucked into the page header never reach the structured contact block.

Generic headlines

A headline of "Marketer with a passion for storytelling" feeds nothing useful to the AI title-match. Use the actual role you want next: "Senior Marketing Manager | B2B SaaS Demand Generation."

Image-scanned PDFs

Scanned PDFs without an OCR layer pass the upload step but parse to empty Experience and Education cards. The AI labels these candidates "Possible Match" by default, regardless of actual fit.

One additional pattern worth flagging: candidates with the right experience but a resume built for human readers, not parsers, routinely lose to less-qualified candidates whose resumes are built for both. The fix is not to dumb down content; it is to add a Skills section with exact-phrase keywords, fix dates and section names, and rebuild as a single-column file. The content stays exactly as good. The auto-score climbs every time.

For a deeper look at how parsed profiles flow into AI ranking systems across multiple ATS platforms, the resume matching guide covers the seven-category model that drives candidate surface order on platforms like Workable, Greenhouse, Lever, and Workday.

Workable ATS Resume FAQ

More than 30,000 companies across 100-plus countries hire on Workable, per the platform's own reporting. Named customers include Athena, Bloomscape, Wave, Reebelo, Sweetgreen, Forsta, Help Scout, and Ryanair. The base skews SMB and mid-market, typically organizations between 50 and 2,000 employees, with strong representation from B2B SaaS, direct-to-consumer retail, hospitality, and global organizations hiring in country-specific pockets.

Workable's AI Resume Screening and AI Screen Assistant compare each parsed Candidate Profile against the job description and produce a fit score plus a short rationale before any recruiter opens the profile. The model weighs Skills chip overlap with required and preferred skills, total years of relevant experience, title and seniority match, location alignment, and the quantified-outcome language in your bullets. Candidates are surfaced in the recruiter UI sorted by AI score, labeled "Top Match," "Good Match," or "Possible Match."

Workable accepts PDF, DOC, DOCX, ODT, TXT, and RTF up to 5 MB. Our parse-fidelity testing shows .docx delivers the highest fidelity Candidate Profile and Skills chip generation, with PDF exported from Word a close second. Image-scanned PDFs without an OCR layer pass upload but parse to empty Experience and Education cards and should be avoided. RTF and ODT work for plain layouts but lose chip granularity on complex templates.

Workable's parser handles two-column layouts in the sense that the upload step succeeds, but the resulting Candidate Profile is reliably wrong. The parser reads top-to-bottom across columns, which scrambles Experience cards and folds sidebar Skills lists into the first job's bullet content. Skills chip generation is incomplete in this case, which drops the AI skills-match score by 15 to 25 percent in our testing. Single-column resumes are the only safe choice on Workable.

Workable's customer base is overwhelmingly SMB and mid-market, but the platform does serve enterprise organizations in geography- or department-specific deployments. Ryanair, for example, uses Workable for cabin crew and select corporate hiring across Europe rather than the entire company. Large enterprises typically use Workday, SuccessFactors, or iCIMS as their primary ATS and Workable in pockets where speed and AI-first screening matter more than enterprise compliance depth. If you are applying through a careers page on a Workable subdomain, treat the experience as SMB-style regardless of the company's overall size.

Workable maps to a defined field schema: Name and contact block, Headline, Summary, Experience cards (employer, title, location, dates, bullets), Education cards (institution, degree, field of study, year), Skills chips, Certifications, and Languages. Use standard section labels: "Summary," "Experience," "Education," "Skills," "Certifications," "Languages." Creative labels like "What I Bring" or "My Story" cause the parser to collapse the section into unstructured profile notes rather than the right field.

Workable's parser tolerates photos and the AI does not score them, so a headshot does not directly hurt the AI fit rating. The risk is indirect: high-resolution photos and embedded company logos are the most common reason a resume crosses the 5 MB cap, and oversized files often fail to attach silently while the form submission appears to succeed. In markets where photos are standard practice (Germany, France, parts of South America), compress images and keep the total file size under 1 MB; in markets where photos are not standard, the upside is roughly zero and the downside is meaningful.