Plenty of half-knowledge circulates around the ATS application: invisible white text, keyword quotas, "ATS-optimised" templates for €29. Most of it solves a problem that, in Germany, looks different from what the guides claim.
What is an ATS, and what does it do to your application?
An applicant tracking system is the database your application lands in. It ingests your files, extracts fields like name, work history, education and skills, stores them in structured form and makes them searchable.
What an ATS normally does not do is decide on its own that you're unsuitable. It sorts, it displays, it searches. The filtering happens at a screen with a person sitting in front of it.
That's the crucial difference from the popular story. Your opponent isn't an algorithm. Your opponent is an incompletely parsed data row and a few seconds of attention.
Does an ATS reject applications automatically?
In Germany, almost never. The representative Bitkom survey published 10 March 2025 (852 companies with three or more employees, fieldwork autumn 2024) found that just 1% of companies screen applications with AI support. For 74%, AI-driven pre-selection is out of the question entirely.
At the same time the process is fully digitised: 100% of companies surveyed accept electronic submissions, and 88% file documents in a talent pool for later selection rounds.
Which leads to an uncomfortable clarification. There is no automatic no. There is a failure to be found. If your qualification sits in a field the recruiter isn't searching, it never appears in the result list, and nobody knows you're there.
What an applicant tracking system extracts
The system breaks your document into fields. Anything it can't confidently assign ends up, at best, in a free-text field, where a skills search won't surface it.
| Field in the system | Where it's pulled from | Typical mistake in the application |
|---|---|---|
| Contact details | Top area of the CV | Details in the page header, often not parsed |
| Work history | Section with roles and date ranges | Date range given only as a year, or omitted |
| Job title | Line above employer and dates | Internal title with no market-standard equivalent |
| Education | Section with degree and institution | Degree given only as an abbreviation |
| Skills | Dedicated block or body text | Skills only inside a graphic or a rating bar |
| Certificates | Dedicated block | Attached as a file but never named in the CV |
The consequence is mundane and still missed daily. If a skill exists only in a graphic, a text box or a header, it doesn't exist for the search.
Why qualified applications get stuck anyway
Two causes, both documented. The first is criteria that are too literal. Hidden Workers: Untapped Talent, from Harvard Business School and Accenture (September 2021), found that 88% of employers surveyed agree qualified high-skilled candidates get vetted out because they don't meet the exact criteria in the job description. The filter asks about phrasing, not capability.
The second is speed. The Ladders eye-tracking study (2018, 30 recruiters over ten weeks) measured an average of 7.4 seconds for the initial scan of a CV. What worked there were simple layouts with clear sections and headings; cluttered design, missing white space, multi-column layouts and long sentences performed badly.
Together they set the actual task: quick for the machine to parse, quick for the human to grasp.
How to build an ATS-ready application
Keep the file simple. One column. PDF unless told otherwise, with selectable text rather than a scan or a screenshot. Standard fonts. No text boxes, no graphics carrying load-bearing content, no tables for your work history. Headers and footers frequently get lost in parsing, so contact details belong in the body. Use conventional section headings while you're at it: Work Experience, Education, Skills, Certifications. A creative title like "My Journey" is useless to a parser.
Write dates as month and year in one consistent format (03/2021 – 08/2024), and explain gaps rather than hiding them. For job titles, give your real position and add the market-standard term in brackets: "Referentin Marktkommunikation (Marketing Manager)". That serves the search without inventing anything.
Give skills their own block, with each term spelled out and abbreviated: "Structured Query Language (SQL)". Systems match strings, not meanings.
The top third should say what you deliver rather than what you were responsible for:
- Marketing: conversion movement, testing cadence, reporting, CAC/ROAS fundamentals
- Sales: pipeline build, discovery, qualification, objection handling
- IT: tested delivery, incident reduction, performance improvements
- Electrical engineering: commissioning, fault isolation, standards-compliant repair
Then add one line reading "Work sample:" with exactly one link behind it. Offer five and a reader with 7.4 seconds opens none of them.
What ATS optimisation is not
White text on a white background gets parsed and displayed in the preview anyway. Recruiters see it, and it is a documented attempt to deceive. A block of 40 keywords without context gets found and then discarded, because none of the terms recurs in the experience section. Copying the job ad is transparent to recruiters, who wrote the text themselves.
A bought "ATS template" is worth a second look before you pay for it. Many of the designs on sale are two-column with icons, which is precisely what parses badly.
The line is simple: anything that makes your real qualification easier to find is legitimate. Anything that fakes a qualification is not.
What counts beyond the form
A cleanly parsed CV gets you into the result list. It doesn't answer the question that comes next: is any of this true?
That's why a second layer pays off, evidence that comes from you plus a check that doesn't. How to build the first part in seven days is in Portfolio for your application.
The second part is what a structured skills check covers. SkillStamp's assessment consists of a technical test tailored to your profession, a behavioural assessment that AI analyses and trained assessors interpret, a video introduction or a written motivation letter as an alternative, a check of your certificates and references, and a 15-minute live interview with a consistent rating framework. The result is a report that belongs to you, and you decide which employers see it. For qualifications earned abroad the verification step matters particularly, because it establishes authenticity before anyone has to ask.
For employers: your filters are costing you candidates
The 88% from the Hidden Workers study is not a candidate problem, it's a criteria problem. Two changes help immediately:
- Split the requirements. What must be present on day one, and what is learnable within 60 to 90 days? Only the first group is allowed to filter.
- Search on activities, not titles. Point your search terms at what people do and the tools they use. Titles don't compare across industries; activities do.
Then validate the must-haves through evidence rather than phrasing. Stop filtering on vocabulary and you start finding the people who can actually do the work.
Common questions
Cover letter or not? If the field exists, fill it. The career change article has a skeleton for the case where your path needs explaining.
Photo? Still customary in Germany, not legally required. Include it as a separate image so it doesn't interfere with text parsing.
Word or PDF? PDF unless told otherwise. If the ad explicitly asks for .doc or .docx, follow that.
Your next step
Take your current CV, copy the text into an empty plain-text file and read the result. Whatever shifts, disappears or turns unreadable there will do the same in the applicant tracking system. Fix that first, then the content.
How structured evidence works: SkillStamp for talent. And if you want to know how much AI is actually in German selection processes: AI in recruiting.