Lever powers hiring at Shopify, Reddit, Zendesk, and 18,000+ other companies under the Employ Inc. umbrella. It has a specific weakness with acronyms and a heavy recency weighting that most candidates never account for.
Lever, now operating as LeverTRM under Employ Inc., combines applicant tracking with candidate relationship management. This means it tracks candidate interactions, email threads, and recruiter notes alongside application data β making it a richer system than pure ATS tools. Lever is popular among mid-market tech companies and fast-growing startups that want collaborative hiring with structured pipelines. Its 18,000+ combined customer base under Employ Inc. makes it one of the more widely encountered systems in the technology sector.
Lever's parser has two behaviors that most candidates overlook. First, it does not expand abbreviations β 'ML' is a different string from 'Machine Learning', and both must appear explicitly to match job descriptions that use either form. Second, Lever applies significantly heavier weighting to experience from the past five years. A strong background from six or more years ago contributes much less to your match score than recent equivalent experience. If your most relevant work is not in your recent roles, Lever may rank you low despite a genuinely strong background.
DOCX produces more consistent section identification in Lever. Text-based single-column PDFs parse adequately. The critical requirement is single-column layout regardless of format β Lever reads content linearly and two-column designs cause job descriptions to merge with unrelated fields.
Part of Employ Inc. with 18,000+ combined customers
Abbreviation handling, recency weighting, and layout behaviors specific to Lever TRM
Lever's matching engine uses exact string matching without an abbreviation expansion dictionary. This creates a specific problem for technical roles where both forms of a term commonly appear: 'ML' in job descriptions and 'Machine Learning' in resumes, or 'PM' vs 'Product Manager', or 'SWE' vs 'Software Engineer'. Your resume must contain both the spelled-out form and the common abbreviation to match job descriptions that use either. Write 'Machine Learning (ML)' the first time you use each term.
Lever's relevance scoring applies a time decay function to work history. Experience from within the last 5 years contributes fully to your match score; experience from 5-10 years ago contributes at a reduced rate; older experience contributes minimally. This means a candidate with strong recent experience in adjacent skills can outscore someone with directly relevant experience from 8 years ago. Lead with your most recent roles and ensure recent positions describe the skills the job requires.
Any text that exists as part of an image β logos with company names, scanned certificates, skill icons with text labels, or decorative elements β is not processed by Lever's text extraction layer. This is a common failure point for candidates who include their university logo, a certification badge image, or a profile photo with their name overlaid. Lever reads only machine-readable text characters, not image content.
Lever reads document content in linear order from top to bottom. A two-column layout in Word places the left column's content first in document order, then the right column's content β regardless of visual alignment. When the parser encounters a date from column two immediately after a job title from column one, it cannot tell which title the date belongs to. Job descriptions end up associated with the wrong positions, and employment history becomes unreliable.
Abbreviation strategy, recency optimization, and formatting rules for Lever TRM
On the first occurrence of any acronym or abbreviation, write the full form followed by the short form in parentheses: 'Machine Learning (ML)', 'Product Manager (PM)', 'Search Engine Optimization (SEO)'. This ensures your resume matches job descriptions that use either the spelled-out form or the abbreviation, doubling your keyword coverage without any content repetition.
Lever's recency weighting means your last 5 years of experience carry substantially more scoring weight. Allocate more space, more bullet points, and more keyword-rich descriptions to your recent roles. If your most relevant experience is older, find ways to reinforce those skills in recent roles β even if you used them less frequently, mention them explicitly in recent job descriptions.
Use 'Month YYYY' format for all employment dates β 'March 2021', 'Jan 2023'. Inconsistency between date formats in different sections confuses Lever's date parser and creates incorrect employment duration calculations. Accurate duration data directly affects how long your experience in each skill is counted for matching purposes.
Build your resume as a single-column document with content flowing from top to bottom. Do not use Word tables to create side-by-side content, text boxes for skill callouts, or any design element that places text in non-linear document order. This is the foundational requirement for Lever to correctly associate job descriptions with the right positions.
Submit in DOCX format when possible. Lever uses document structure from DOCX files to identify sections more reliably than PDF text extraction. Use standard headings: 'Experience', 'Education', 'Skills'. DOCX heading styles (Heading 1, Heading 2) give the parser additional structural signals that PDF text cannot provide.
Lever includes candidate matching that scores applications against job descriptions based on keyword overlap and experience relevance. The scoring weights recent experience more heavily than older history and requires exact string matches β abbreviations do not automatically match their full forms. Recruiters see a match score alongside each application, which influences review order. A well-formatted DOCX with explicit abbreviations and strong recent experience sections scores noticeably higher.
Many companies using Lever allow applications through LinkedIn Easy Apply or direct LinkedIn apply, which passes your LinkedIn profile data into Lever. This bypasses the resume upload parser entirely and uses structured LinkedIn profile data instead. LinkedIn application data is generally more reliably structured than parsed resume files. If applying through LinkedIn is available, it often produces a cleaner candidate profile than uploading a file.
Lever applies reduced scoring weight to experience older than 5 years, but this does not mean excluding it. List older relevant positions with concise descriptions. More importantly, find legitimate ways to reference relevant older skills in recent job descriptions β even brief references to long-standing proficiencies count as recent mentions. A 'Core Competencies' section can also list skills without tied to dated positions, keeping them visible to the matcher.