ATS Filters: What Software Does to Your CV Before Any Recruiter in Switzerland

An ATS does not read a CV like a person: it extracts plain text, compares it to criteria defined by the job posting, and assigns a score. This score determines whether the application will be presented to a human recruiter or rejected outright. The actual quality of the profile is only evaluated after this first automated screening, and only if the CV passed the filter.

According to several European HR studies, more than 70% of CVs are eliminated before any human review in organizations deploying these systems. In Suisse Romande, a single posting at a major Geneva bank or cantonal hospital can receive 150 to 300 applications: the automated filter is not optional, it is a logistical necessity.

Parsing, Extraction, Score: What Happens in the 30 Seconds After Submission

Upon receiving a CV, an ATS executes three successive operations. The first is parsing: the software automatically extracts text from the file and attempts to identify structural blocks (experience, education, skills, contact info). This step fails partially or completely on a large number of CVs, particularly those created with graphical templates or software that encodes text as images rather than readable characters.

The second operation is classification: the software associates each extracted segment with a category in its data model. The most recent job title becomes the "current position." Listed companies populate the work history. Tools and technologies mentioned form the skills profile. If parsing failed on a section, that section is absent from the profile, invisible for downstream processing.

The third operation is scoring: the ATS compares the extracted profile with the job posting criteria and calculates a compatibility score. This score determines ranking within the application pool. A recruiter opening their ATS interface doesn't see 200 CVs in order of arrival: they see applications ranked by descending score, with "unqualified" applications often hidden by default.

Keywords That Separate a Retained CV from a Filtered One

ATS scoring relies primarily on lexical match between the CV and the job posting. The ATS doesn't understand synonyms: it recognizes character strings. A profile mentioning "software development" doesn't automatically match a posting requesting "software engineering." A CV discussing "risk management" will score lower than one using exactly "risk management" if the posting is in that language.

The most weighted terms in scoring are generally the exact job title, technical skills explicitly named in the posting, tools and software cited, mentioned certifications, and sector acronyms. In Geneva finance: FINMA, LSFin, MiFID II. In Vaud pharma: GMP, ICH Q10, FDA. In tech: programming languages, frameworks, cloud platforms. A CV that names these terms in the right linguistic register automatically receives a better score.

Keyword density matters, but not without limits. Some ATS systems detect keyword stuffing (artificial accumulation of terms out of context) and penalize CVs that list terms without grounding in concrete achievement. The best practice is to integrate key terms into descriptions of real work experience, not in a "skills" section that verbatim repeats the posting.

Formats That Make a CV Invisible to an ATS

File format is as important as content. A perfectly written CV can be entirely ignored by an ATS if its format prevents parsing. Several layout types produce empty or incomplete profiles in the system.

Multi-column CVs created with Canva, Adobe InDesign, or complex Word templates create a structural problem. The ATS parser reads text as a linear flow, left to right and top to bottom. A two-column layout produces incoherent sequence: "Finance Director CFA Lausanne Education HEC Paris 2014" on one line, making classification impossible.

PDFs exported from PowerPoint or Canva contain text rendered as image. The parser cannot extract the text, and the resulting profile is empty. The ATS receives the file, but sees nothing.

Microsoft Word headers and footers are ignored by some ATS systems. A candidate placing contact info in the document header may find themselves without name or email in the company system. HTML or Word tables pose the same problem: linear reading produces unreadable content, and adjacent cells are often concatenated without separators.

The recommended format is a PDF generated from standard word processing software (Word, Google Docs, LibreOffice), with linear layout in a single column, sections clearly marked by headings, and common fonts (Arial, Calibri, Georgia). This produces a "text-layer PDF" whose extraction is reliable across all ATS systems on the market. For more details on structure and expected conventions, the guide on the Swiss Romande CV template covers best practices for layout.

What the Recruiter Sees After the Filter

When a Swiss recruiter opens the application file in their ATS, the interface presents a list sorted by score. Applications below a certain threshold (often set between 50 and 70% match) are hidden or moved to a "not retained" folder rarely consulted. The recruiter begins human review from the highest-scored applications, not in order of arrival.

This ranking is not final. In companies using their ATS well, an experienced recruiter can review rejected applications if the pipeline is insufficient. But this review only happens if the volume of qualified applications is too low. In practice, for an IT posting in Geneva receiving 180 applications, the recruiter will never revisit the 150th application.

The direct consequence is that two candidates with equivalent profiles can have radically different outcomes depending on the lexical quality of their CV. A CV well-calibrated for the specific posting reaches a human recruiter. A generic CV, even if excellent substantively, stays in the pile. This is the mechanism of silent rejection: no response, no feedback, no explanation. To understand the full process of how applications are read and sorted in Switzerland, the guide on how a CV is actually read details the steps after the ATS filter.

An ATS does not evaluate a candidate: it evaluates a document. Qualifications, potential, career trajectory:none of these exist for the software if the CV doesn't bear their exact lexical traces. Passing the filter is not a matter of luck or modesty: it is a matter of precision. Candidates who adapt their vocabulary to each posting aren't cheating. They are translating their background into the language the system can read.

Sending a generic CV to a Workday posting at Nestlé or a Taleo posting at UBS resembles submitting a Vaud building permit application according to Geneva cantonal forms: the substance may be impeccable, but the format prevents administrative processing from starting. Adapting a CV's vocabulary to each posting is not cosmetic: it is the condition for the file to enter the right workflow.

---

Frequently Asked Questions

How can you tell if a Swiss company uses an ATS?

Several signals indicate ATS presence: an online application form on the company website (not just email), a candidate portal with "my account" space (Workday, Taleo, SAP SuccessFactors), or a reference number in the posting. In Suisse Romande, all companies with more than 200 employees, cantonal hospitals, banks, and multinationals almost certainly use an ATS. Local SMEs and independent recruitment firms still often receive CVs via email.

What CV format passes an ATS filter best: PDF or Word?

A "text-layer" PDF generated from Word, Google Docs, or LibreOffice is the most reliable format: readable by all ATS systems on the market, it preserves formatting and cannot be accidentally modified. The Word file (.docx) is an acceptable alternative if the document contains no complex formatting. Absolutely avoid: PDFs exported from Canva, PowerPoint, or Keynote, which contain text rendered as image and produce an empty profile in the ATS.

Do ATS systems read cover letters?

Rarely automatically. Most ATS systems treat the cover letter as an attachment reviewed by the human recruiter, not as a scored document. Some newer systems apply textual analysis to the letter, but primary scoring is based on the CV. In practice, a well-adapted cover letter strengthens the application after the ATS filter, not before.

Can a candidate know if their CV passed the ATS filter?

Not directly. ATS systems send automatic confirmation but do not communicate the score obtained. Some portals display a status "application forwarded to hiring manager" indicating the threshold was passed. In the absence of feedback within two to three weeks of submission, it is reasonable to assume the CV did not reach the human review stage.