Smart ATS & CV Analysis
What an ATS actually does, and what it does not
An applicant tracking system is a database. It stores applications, parses them into fields such as name, employer, job title and dates, and lets a recruiter search and filter what came in. That is the whole job. It is closer to a filing cabinet with a search box than to a gatekeeper.
This matters because the most repeated claim about ATS software is that it automatically rejects most CVs before a human sees them. That claim usually travels with a precise sounding percentage and no source behind it. In practice, systems are configured by the employer, and the common setup is that a person runs a search or reads a ranked list. Rejection is a decision someone makes, not something the software does on its own.
So the real risk is not being rejected by a machine. It is being invisible in a search: your CV parsed into the wrong fields, the terms a recruiter searched for missing from your text, or your experience readable to a human eye but not to a parser. Those are fixable problems, and they are what this page and this checker are about.
What this ATS checker looks at
Upload a PDF or Word file and the analysis runs in two directions at once. The first is mechanical: can the document be read at all, does the text come out in a sensible order, do the sections and dates survive extraction. If a CV loses its structure here, nothing else on the page matters.
The second is editorial, and it is the part most tools skip. Rather than counting keywords, the analysis reads what your experience lines actually claim: whether a bullet names a result or only a duty, whether the summary says something specific about you, whether your seniority is supported by scope. You get an AI evaluation with a score and a separate recruiter style read of the same document.
Deliberately absent: there is no single mystical ATS compatibility percentage pretending to be an industry standard. No such standard exists, because every employer configures their system differently. A number invented to look authoritative would be the least honest thing this page could show you.
How to read the score you get
Treat the score as a reading of this document for this target role, not as a permanent property of your career. Change the target position and the same CV will score differently, which is the point rather than a flaw. A CV is an argument aimed at a specific job, and the same argument is stronger or weaker depending on what it is aimed at.
The comments are worth more than the number. A score tells you roughly where you stand; the written evaluation tells you which line is doing the damage and why. If you only have time to act on one thing, take the weakest experience bullet the evaluation names and rewrite it so it contains a decision you made and something that changed as a result.
The formatting choices that actually break parsing
Most formatting advice on this topic is superstition. A short list of choices genuinely causes trouble, and the rest is taste. These are the ones worth changing before you apply anywhere.
- Text inside an image. A logo is fine, but a name, a headline or contact details rendered as a graphic are invisible to text extraction. If you cannot select the text with your cursor, a parser cannot read it either.
- Multi column layouts. Two column designs look elegant and often extract in the wrong order, interleaving the sidebar with the main column so that a skill lands in the middle of a job description.
- Contact details in the header or footer. Some extraction tools drop those regions entirely. Your email and phone number belong in the body of the first page, where nothing can strip them.
- Tables used for layout. Simple tables usually survive, nested ones frequently do not, and a table used to fake columns has the same reordering problem as a column layout.
- Dates written inconsistently. Mixing 03/2021, March 2021 and 2021-03 across entries makes it harder to build a clean timeline, and a broken timeline reads as a gap that was never there.
File type matters less than people think. Modern systems read PDF and Word without difficulty, so send whatever the employer asks for, and choose Word only when the posting explicitly requires it. A single column PDF with selectable text and ordinary section headings clears almost every parser in use.
Why pasting the job posting changes the answer
Without a job posting, your CV is compared against a general picture of the role you typed. With one, it is compared against the specific text an employer wrote, including the words they chose and the requirements they put first. Those are usually the terms a recruiter will search on, so the gap that comes back is a concrete list rather than general advice.
Use the result as a checklist, not a script. If the posting asks for something you have done and your CV does not name it, that is a genuine omission worth fixing. If it asks for something you have not done, adding the phrase anyway will fail in the first interview, and it damages more than the missing keyword ever would.
What this analysis cannot tell you
Being clear about the limits is part of a useful tool. Four things this cannot do:
- It cannot see the specific system your employer uses, or how they configured it. Nobody outside the company can, and any tool claiming otherwise is guessing.
- It cannot predict whether you will be interviewed. Hiring depends on competition, timing, referrals and budget, none of which are in your document.
- It cannot verify your claims. A well written CV describing work you did not do will score well and collapse in a conversation.
- It cannot replace tailoring. A strong general CV still loses to one written for the specific posting, and no automated pass changes that.
Frequently asked questions
Is my CV really rejected automatically if it is not ATS friendly?
Usually no. The common failure is quieter: your CV is stored but does not surface in the searches a recruiter runs, or it parses into the wrong fields so your last job looks like something else. That is why the goal is being findable and readable rather than passing an imaginary test.
PDF or Word, which is safer?
Send what the employer asks for. Both are read reliably by current systems, so the file type is rarely the problem. What matters far more is that the text is selectable, the layout is a single column, and the section headings are ordinary words like Experience and Education.
Do I need to repeat keywords to rank higher?
No, and stuffing is easy to spot. Use the employer's own term once in the place it belongs, usually the experience line where you actually did that work. A term that appears in context reads as evidence; the same term repeated in a list at the bottom reads as padding.
Should my CV be one page?
Length follows content, not a rule. Early career work fits on one page and looks padded if stretched. Ten or more years usually needs two, and compressing it hides the very scope that justifies the level you are applying for. Beyond two pages, the case has to be unusual.
Is a photo a problem?
It depends on the country, not on the software. In the United States, the United Kingdom and Ireland a photo is normally left off and some employers discard CVs that include one. In much of continental Europe it is ordinary. Follow the norm of the country you are applying in.
How often should I re run the analysis?
Once per meaningful rewrite, not once per edit. Change the things the evaluation named, then check the new version against the specific posting you are applying to. Running it repeatedly on nearly identical text tells you little and is the least useful way to spend the attempt.
Job posting
If you paste a job ad, the ATS score and keyword match are calculated against it (like Jobscan). Leave empty to evaluate against the target role only.