What Is ATS Scoring? A Complete Guide for 2026

You spend an hour tailoring a candidate’s CV. You match their experience to the role. You highlight the right skills. You hit send. And then… nothing. No acknowledgement, no interview, no feedback. The CV disappeared into a black hole.

That black hole has a name: Applicant Tracking System, or ATS. And the reason the CV vanished is almost certainly not the content — it’s because the ATS couldn’t read it properly.

This guide explains how ATS scoring works, what determines whether a CV passes or fails the automated screening, and the eight specific factors you can check — and fix — before you submit.

What is an ATS?

An Applicant Tracking System is software that companies and recruitment agencies use to manage the hiring process. Think of it as a database that receives, stores, and filters job applications. Popular systems include Greenhouse, Lever, Workday, Bullhorn, and iCIMS.

When a candidate submits a CV, the ATS doesn’t just store the file — it parses it. The system attempts to extract text from the document, identify structured fields (name, job titles, dates, skills, education), and index that data so recruiters can search and filter candidates later.

Here’s the problem: if the ATS can’t parse the CV correctly, the extracted data is garbled or incomplete. The candidate’s skills don’t appear in search results. Their experience doesn’t map to the right fields. Effectively, the candidate becomes invisible — not because they’re unqualified, but because the system couldn’t read their document.

What is ATS scoring?

ATS scoring is a way to measure how well a CV will perform when processed by an Applicant Tracking System. But there are actually two different types of scoring, and most people conflate them:

1. Keyword relevance scoring — does the CV’s content match a specific job description? This is about whether the right skills, job titles, and industry terms appear in the document.
2. Structural compatibility scoring — can the ATS actually read the CV at all? This is about whether the document format, layout, fonts, and encoding allow the system to extract text accurately.

Most advice online focuses on type 1 — stuffing keywords into your CV to match a job description. That matters, but it’s pointless if type 2 fails. If the ATS can’t extract text from the document, it doesn’t matter how many keywords you’ve included. The system literally cannot see them.

Structural compatibility comes first. Keyword relevance comes second.

The 8 factors that determine ATS compatibility

ATS compatibility isn’t a single yes-or-no check. It’s a combination of factors, each carrying different weight depending on how severely it affects the system’s ability to parse the document. Here are the eight factors that matter most, ordered by importance.

1. Text extractability (most important)

This is the single most important factor. Can meaningful text actually be extracted from the PDF? Some CVs look beautiful on screen but contain barely any extractable text — because the content is embedded in images, rendered as vector graphics, or uses encoding that produces gibberish when copied.

The simple test: open the CV in a PDF viewer, press Ctrl+A to select all, then Ctrl+C to copy, and paste into a plain text editor. If what comes out is readable — names, dates, paragraphs of text — you’re in good shape. If you see garbled characters, empty space, or nothing at all, the ATS will have the same experience.

A good CV should have at least 50 extractable words, with a high ratio of meaningful characters (letters and digits) to total characters.

2. Reading order

Even when text can be extracted, it might come out in the wrong order. This is especially common with multi-column layouts. A two-column CV might look logical to a human reader, but when the ATS extracts text line by line, it can interleave content from both columns — producing a jumbled mess where a job title from the left column appears next to a skill from the right column.

Signs of reading order problems include excessive short text fragments, large gaps of whitespace between lines of content, and very low average line lengths in the extracted text. Single-column layouts almost always produce better reading order.

3. Font embedding

Every font used in a PDF should be embedded — meaning the font data is included inside the PDF file itself, rather than relying on the viewer’s system to provide the font. When fonts aren’t embedded, the ATS may substitute a different font, which can cause character mapping errors. A common symptom: the letter “fi” (a ligature in many fonts) becomes a blank space or a replacement character.

You can check font embedding by opening PDF properties in most viewers (in Adobe Reader: File → Properties → Fonts). Every font should show “Embedded” or “Embedded Subset”. If any font shows “Not Embedded”, there’s a risk.

4. PDF metadata

PDF files contain metadata fields: Title, Author, Subject, and Creator. Many ATS systems read these fields to supplement what they extract from the visible content. A PDF with a meaningful Title (e.g. “John Smith — Senior Developer”) and Author field gives the ATS additional structured data to work with.

This is a relatively minor factor, but it’s also the easiest to fix. Most PDF editors and export tools allow you to set these fields before saving.

5. Section headings

ATS systems look for standard section headings to understand the structure of a CV. Headings like Experience, Education, Skills, Certifications, and Professional Summary act as signposts that help the parser map content to the right fields.

Creative heading alternatives (“My Journey”, “What I Bring”, “The Story So Far”) might look distinctive, but they’re invisible to most ATS parsers. Stick to the conventional terms. There are over 30 widely recognised heading variations that ATS systems understand — including Work Experience, Employment History, Technical Skills, Core Competencies, Academic Background, and Professional Development.

The most critical heading to include is some variant of “Experience” or “Work Experience”. Without it, most ATS systems struggle to identify where the employment history begins.

6. Special characters and icons

Many modern CV templates use icon fonts (like FontAwesome) for visual flair — a phone icon next to a phone number, an envelope icon next to an email address. These look polished to humans, but ATS systems see something very different.

Icon fonts use Unicode’s Private Use Area (PUA) — a range of character codes that have no standard meaning. When an ATS extracts text containing PUA characters, it sees meaningless symbols, or worse, replacement characters (the “?” in a diamond you sometimes see). A handful of these is tolerable; dozens of them can significantly degrade parse quality.

The rule: if information is important, convey it with real text, not icons. Write “Email:” instead of using an envelope icon. Write “Phone:” instead of using a phone icon.

7. File size

Most ATS systems impose upload size limits, commonly between 2MB and 5MB. A CV under 2MB is ideal. Between 2MB and 5MB is usually acceptable but may trigger slower processing. Over 5MB risks outright rejection by some systems.

Oversized CVs are typically caused by embedded high-resolution images (profile photos, company logos) or uncompressed graphics. If your CV is over 2MB, look at image compression or removal as the first fix.

8. Page count

This isn’t strictly an ATS parsing issue, but page count affects both ATS processing and human review. A 1–2 page CV is ideal for most roles. 3–4 pages is acceptable for senior positions with extensive experience. Beyond 4 pages, some ATS systems truncate the document, and even those that don’t will present so much data that key information gets buried.

How these factors are weighted

Not all factors are equally important. If we assign weights based on how severely each factor affects ATS parsing, the breakdown looks roughly like this:

• Text extractability: 25% — without this, nothing else matters.
• Reading order: 20% — extractable but scrambled text is nearly as bad as no text.
• Font embedding: 15% — missing fonts cause character-level corruption.
• PDF metadata: 10% — helpful for field mapping but not critical.
• Section headings: 10% — help the parser understand document structure.
• Special characters: 10% — icons and PUA chars add noise to parsed output.
• File size: 5% — mainly affects upload acceptance.
• Page count: 5% — affects processing and human review.

A weighted score across these factors gives a reliable 0–100 compatibility rating. Scores above 90 indicate excellent ATS compatibility. Between 70 and 90 is good with room for improvement. Below 70 means the CV has structural issues that are likely to cause parsing failures in some systems.

Common mistakes that tank your ATS score

Knowing what the factors are is half the battle. Here are the most common mistakes we see in CVs that score poorly:

• Multi-column layouts — they look great on paper but destroy reading order. Sidebar designs, two-column skill grids, and parallel-layout work histories all cause interleaving problems.
• Header images and logos — a candidate photo, a decorative header graphic, or a company logo adds file size while contributing zero parseable content. If the candidate’s name is part of a header image, the ATS won’t even know who the CV belongs to.
• Icon-heavy designs — skill bars, star ratings, progress circles, and icon fonts are invisible to ATS parsers. If a candidate’s top skill is represented by 5 filled stars and no text, the ATS sees five meaningless characters.
• Tables for layout — using HTML or Word tables to control spacing and alignment often scrambles the reading order. ATS parsers may read across rows instead of down columns, or skip table content entirely.
• Non-standard fonts without embedding — creative typefaces are fine as long as they’re embedded in the PDF. Without embedding, character mapping breaks silently.
• Creative section headings — “Where I’ve Made an Impact” instead of “Experience”. The human reader understands the intent; the ATS does not.

ATS compatibility vs. keyword matching

It’s worth restating the distinction because they’re frequently confused:

• ATS compatibility answers: “Can the system read this document?” It’s about format, structure, and encoding. It’s independent of any specific job.
• Keyword matching answers: “Does this CV match this job description?” It’s about content relevance. It’s specific to each application.

You need both, but they happen in sequence. If compatibility fails, keyword matching never gets a fair chance. Think of it as two gates: gate one checks whether the document is readable, gate two checks whether the content is relevant. Most rejected CVs fail at gate one — before the content is even evaluated.

This is why optimising for ATS compatibility is the higher-leverage activity. You do it once and it benefits every application. Keyword optimisation, by contrast, needs to be tailored for each role.

A practical ATS compatibility checklist

Before submitting any CV, run through these checks:

1. Copy-paste test — select all text in the PDF and paste into a plain text editor. Is the output readable and in the right order?
2. Single-column layout — is the CV structured in a single column, or does it use sidebars and multi-column sections?
3. Fonts embedded — check PDF properties to confirm all fonts are embedded or embedded as subsets.
4. Standard headings — does the CV use recognisable section headings like Experience, Education, and Skills?
5. No icon fonts — are all important details conveyed in text, not through icons or visual indicators?
6. Metadata set — does the PDF have a meaningful Title and Author in its document properties?
7. Under 2MB — is the file size reasonable? No high-resolution images inflating the document?
8. 1–2 pages — is the content concise enough for the role level?

Automating ATS compatibility checks

Running through this checklist manually for every CV is possible but tedious — especially for recruitment agencies processing dozens of candidates daily. This is where automated ATS scoring tools add value. They analyse the rendered PDF and check every factor programmatically, producing an instant compatibility score with specific suggestions for improvement.

At Vitae, we built ATS compatibility scoring directly into the CV rendering pipeline. Every time a CV is generated, the system automatically analyses the output PDF across all eight factors and produces a score from 0 to 100 with a letter grade. If there are issues — a missing section heading, a non-embedded font, PUA characters from icon fonts — the report flags them with specific, actionable suggestions.

The scoring is entirely structural and deterministic — no AI black box, no keyword guessing. It tells you exactly what’s wrong with the document format and exactly how to fix it. Because the analysis runs on the final PDF (not the source content), it catches issues that only manifest after rendering: font embedding failures, reading order problems from complex layouts, and file size bloat from image compression.

The result is confidence. When you see a score of 90+ before sending a candidate CV to a client, you know it will parse correctly in whatever ATS the client uses. No more invisible candidates. No more CVs lost to formatting issues.