What Is an ATS? How Applicant Tracking Systems Actually Work

This guide pulls back the curtain on how ATS systems actually work. We're not talking surface-level explanations. We're going deep into the parsing engines, scoring algorithms, and ranking systems that determine whether your resume gets a 30-second human review or gets rejected in milliseconds. If you've ever felt like your resume disappeared into a black hole after applying online, this is why — and more importantly, how to fix it.

By the end of this guide, you'll understand exactly what an ATS does, which major platforms dominate the market, what these systems actually check, and how to optimize your resume so it survives every stage of automated screening.

Think of an ATS as a combination of three things: (1) a document parser that converts your PDF or Word file into machine-readable data, (2) a keyword matching engine that measures how closely your background aligns with the job description, and (3) a ranking system that puts applications into tiers: shortlist, maybe-pile, or reject.

The critical point: ATS systems are not intelligent the way you might think. They don't understand the nuance of your career transitions or the impact of your accomplishments. They match keywords, verify job titles and dates, and tally skill mentions. A human recruiter reads context. An ATS counts occurrences.

Every major company now uses one. LinkedIn's 2024 Recruiting Report found that organizations with 500+ employees almost universally deploy ATS technology. Even mid-size startups use simpler versions. If you've applied online in the last decade, your resume has been screened by an ATS.

When you click "Apply," your resume enters a five-stage pipeline. Understanding each stage is the key to surviving automated screening.

Stage 1: Resume Submission & Document Ingestion

You upload a PDF, Word document, or sometimes a plain text file. The ATS immediately stores it in a cloud database (usually AWS S3 or Google Cloud Storage) and begins preprocessing. The goal: convert your nicely formatted resume into machine-readable data.

Here's where the first critical failure point emerges: file format matters enormously. PDFs sometimes confuse parsing engines because they don't contain semantic information — just rendered pixels and text layers. Complex formatting, tables, and graphics can cause the parser to scramble your data. A section header might be misaligned with its content. Your bullet points might be extracted in the wrong order or merged together.

The ATS creates two versions of your resume: a "raw extract" (the text it actually manages) and the original file (which recruiters see if you advance). This is why you can have a beautifully designed resume visible to humans but garbage data in the ATS database — they're literally different objects.

Stage 2: Parsing Engine — Converting Pixels to Structured Data

This is where the heavy lifting happens. The parsing engine attempts to extract structured fields from unstructured resume text. It's trying to answer questions like: What is this person's name? What are their email and phone number? Where did they work, and when? What are their skills?

Modern ATS systems use a combination of:

• Regular expressions (regex): Pattern matching for emails, phone numbers, dates, and URLs. Fast and reliable, but brittle if you deviate from expected formats.
• Machine learning models: Named entity recognition (NER) and trained classifiers that identify job titles, company names, education, and skills even if formatted irregularly.
• Heuristic rules: Hard-coded logic (e.g., "text between 'EXPERIENCE' and 'EDUCATION' is likely work history") that handles common resume structures.

The parsing quality varies wildly between ATS systems. Premium platforms like Workday and Greenhouse have better ML models and handle edge cases more gracefully. Older systems like Taleo struggle with unconventional formatting.

What gets extracted? At minimum: name, contact information, work history (company, job title, dates, description), education, and skills. Advanced systems also extract certifications, languages, volunteer experience, and technical proficiencies.

Stage 3: Keyword Scoring — The Filtering Engine

Once your resume is parsed, the ATS compares your extracted data against the job description. This is a keyword matching process — not semantic understanding, but literal string matching with some sophistication.

The ATS is asking: Does this resume contain the skills, job titles, and experience we're looking for? The scoring usually works like this:

1. Required keywords: The system identifies must-have skills or experience from the job description. If you don't have these, your score plummets.
2. Preferred keywords: Secondary skills, certifications, or nice-to-have qualifications. Each match adds points.
3. Keyword frequency: If the job description mentions "Python" five times and your resume mentions it twice, the score reflects that mismatch.
4. Proximity matching: Some systems reward keywords that appear together (e.g., "Python development" beats "Python" and "development" appearing separately).
5. Recency weighting: Recent experience is typically valued more than old experience. A job title you held three years ago is worth less than one you hold today.

The result: a score, usually on a 0–100 scale. In many systems, anything under 40 is auto-rejected. 40–60 is "maybe" (goes to a recruiter). 70+ is "shortlist" (fast-tracked for review).

Here's the catch: the job description is the template. If the job description uses the term "data scientist" and you've listed yourself as "machine learning engineer," you might miss that match — depending on the ATS's intelligence. Synonyms are sometimes recognized, sometimes not. Different terminology for the same skill can be catastrophic.

Stage 4: Ranking Algorithm — The Triage System

Once every application has a score, the ATS ranks them. Most systems use simple logic: sort by score, descending. But sophisticated platforms (Workday, Greenhouse, Lever) apply secondary ranking criteria:

• Recency of application: Earlier applications get a slight advantage (encourages faster hiring).
• Geographic proximity: Candidates near the office location rank higher.
• Employment history length: Longer tenure at previous roles might signal stability.
• Education prestige: Some systems code for university rankings (though this is increasingly criticized).
• Experience level: Senior candidates ranked above junior for senior roles, and vice versa.

The top 5–10% of applications move to active review. Everything else sits in a pool that recruiters might dig through if they don't find enough qualified candidates.

Stage 5: Human Review — The Final Gate

If you survived stages 1–4, a recruiter or hiring manager now sees your application. But here's the thing: they see both your original resume (the formatted PDF) AND the ATS's parsed data. If the ATS made extraction errors, the recruiter might see garbled information and make a bad first impression — even if your actual resume is excellent.

At this stage, humans apply context, assess cultural fit, verify employment history, and evaluate soft skills — all things the ATS cannot do. But the damage is already done if you were ranked poorly.

Not all ATS systems are created equal. Let's break down the market leaders and their characteristics.

Workday (Enterprise Standard)

Workday dominates the enterprise space — used by most Fortune 500 companies and government agencies. It's excellent at parsing complex resume formats and has sophisticated ML-based keyword matching. Workday systems handle job title variations well and recognize industry-standard terminology.

The downside: Workday is ruthlessly efficient at filtering. If you don't match the stated requirements, you're unlikely to advance. There's less room for context to save you.

Greenhouse & Lever (Tech-Focused, Candidate-Friendly)

These platforms dominate tech hiring and are generally considered more sophisticated than their enterprise competitors. They have excellent parsing and recognize non-traditional backgrounds better (e.g., bootcamp graduates, self-taught engineers).

Both emphasize transparency — some companies using Greenhouse or Lever will actually show you what the ATS extracted from your resume, which is rare and valuable.

Taleo (Older, Less Forgiving)

Oracle Taleo powers a huge volume of large corporation hiring, particularly government contractors and financial institutions. Unfortunately, Taleo's parsing is weaker than modern platforms. It struggles with PDFs, non-standard formatting, and unconventional section headers.

If you're applying to a Taleo system, simplicity is survival. Use a plain, straightforward format. Avoid creative design.

iCIMS (Mid-Market Workhorse)

iCIMS is widely used by mid-size companies and hospitality, retail, and manufacturing sectors. Its keyword matching is solid but less sophisticated than Workday. It handles most standard resume formats well.

Now that you understand the pipeline, let's get specific about what ATS systems evaluate.

1. Core Contact Information

The ATS must extract your name, email, phone number, and ideally a city/state. If it can't find these — usually because they're formatted in an unusual way — you're likely filtered out immediately.

2. Work History (Company, Title, Dates, Description)

The ATS extracts your employment history and analyzes it against job requirements. It looks for:

• Job title match (exact or synonymous)
• Duration of tenure (longer is often better)
• Company reputation (some systems weight FAANG companies higher)
• Keywords in job descriptions (specific accomplishments, skills used)

Format matters: use a clear structure like "Job Title | Company Name | Month Year – Present." Avoid unusual date formats. The ATS needs to parse "Jan 2022 – Present" but might struggle with "January, the year of our Lord two thousand twenty-two – Now."

3. Education & Credentials

The system extracts:

• Degree type (Bachelor, Master, PhD, diploma)
• Field of study (Computer Science, Business Administration, etc.)
• University name (and sometimes cross-references university rankings)
• Graduation date
• GPA (if you include it, which is generally only recommended for fresh graduates)

If a job requires a specific degree (e.g., "Bachelor's in Engineering required"), the ATS will check for it. If you have alternative credentials (bootcamp, certifications), some systems recognize them — others don't.

4. Skills & Keywords

This is the heavyweight scoring category. The ATS scans your entire resume for skill keywords mentioned in the job description. It tallies:

• Technical skills (programming languages, frameworks, tools)
• Soft skills (leadership, communication, project management)
• Industry-specific knowledge (regulatory compliance, industry tools)
• Certifications (AWS Certified, Google Cloud, PMP, etc.)

Critically: the ATS doesn't understand that you're an expert. It counts occurrences. If the job description emphasizes "Python" and you mention Python twice, that's better than once — even if your mention was in a bullet point, not a dedicated skills section.

5. Experience Level & Seniority

Some ATS systems infer your career stage based on:

• Years of total experience
• Job titles (Analyst vs. Senior Analyst vs. Director)
• Company progression (from startup to Fortune 500)

If you're applying for a "Senior" role but your resume shows 3 years of experience, the ATS might flag you as under-qualified. Conversely, if you're overqualified (15 years applying for an entry-level role), you might be ranked lower — hiring managers worry about retention and boredom.

6. Geographic Location

The ATS extracts your location and compares it to the job location. Remote roles ignore this. Local roles prioritize nearby candidates. Some systems even weight this — a local candidate might get a scoring boost.

If you're job hunting across regions, make sure your location is current and accurate on your resume.

There's a lot of misinformation about ATS systems online. Let's separate fact from fiction.

You might not know which ATS a company uses, so here's what formats generally work across the board.

Best practice: Unless you know the company uses a modern system (Greenhouse, Lever, Workday), submit in DOCX or PDF. Both have broad support. Avoid image files entirely — no major ATS system can parse them.

Understanding ATS systems is one thing. Beating them is another. Here's the complete optimization checklist based on everything we've covered.

Let's see what the ATS actually "sees" when it parses a resume. Here's a mock resume with annotations showing what gets extracted and how the parser interprets it:

Most ATS systems rank applications on a 0–100 scale. Here's how those scores typically translate:

That said, scoring thresholds vary by company and role. A highly competitive role (senior engineer, product manager at a top tech company) might have a cutoff of 75%. An entry-level administrative role might move to review at 50%.

The critical insight: You don't need a perfect 100. You need to be above the human review threshold. If you're a genuine fit for the role with honest keyword matches, you should score 60+. If you're scoring below 50, there's likely a mismatch between your background and the role — or your resume isn't communicating your fit clearly.

It's important to understand the limitations of ATS technology. These systems are powerful tools, but they have blind spots.

ATS systems cannot:

• Evaluate career transitions. If you're pivoting from finance to tech, the ATS sees skills mismatch. Humans understand the underlying value of your analytical abilities.
• Recognize non-traditional credentials. A bootcamp diploma or self-taught background might not map to "Bachelor's Degree" in the ATS logic, even if you're equally qualified.
• Understand impact without quantification. "Improved customer retention" is less valuable to an ATS than "Improved customer retention by 23%." Metrics matter.
• Recognize gaps or employment history anomalies. A 2-year gap without explanation will hurt your score, even if you have good reasons. Humans might understand; the ATS doesn't.
• Value soft skills equally to hard skills. "Leadership" is harder to verify and score than "Python." ATS systems lean heavily on technical skills.

This is why ATS should be thought of as a filter, not a judge. It's a necessary gate you must pass, but it's not evaluating your true potential. Your goal is to make sure you pass the gate so a human can assess you properly.

Understanding ATS theory is useful. Seeing it in practice is even better. GetNewResume's AI resume tailoring tool includes an ATS compatibility score that simulates what real ATS systems do. Here's how it maps to actual ATS behavior:

When you upload your resume and a job description into GetNewResume, the system:

1. Parses your resume the same way a real ATS does — extracting name, contact info, work history, skills, and education.
2. Analyzes the job description to identify must-have skills, preferred qualifications, and key terms.
3. Scores your resume based on keyword matches, job title alignment, and experience relevance — exactly like Workday or Greenhouse.
4. Shows you the gaps — which skills you're missing, which experience sections aren't communicating your fit, where to add keywords.

What makes this valuable: you're not guessing what an ATS will do. You're seeing the exact scoring logic and can tailor your resume accordingly. The score you see in GetNewResume is a realistic proxy for what you'd see in an actual applicant tracking system.

Most candidates don't have this visibility. They submit resumes and wait. With GetNewResume, you can optimize before you apply — and get concrete feedback on your ATS compatibility score.

Let's recap the essential truths about applicant tracking systems:

When you click "Apply" today, your resume doesn't go to a human. It goes to a sophisticated parsing and scoring engine that looks for specific keywords, validates your work history, and measures how closely you align with stated requirements.

ATS systems aren't malicious. They're not trying to reject good candidates. They're simply trying to sort thousands of applications into manageable piles — shortlist, maybe, reject. They do this using the only tools available to them: keyword matching, structured data extraction, and algorithmic ranking.

Your job, as a candidate, is to understand this system and work within it. Use job description keywords honestly. Format your resume for clarity and parsing accuracy. Quantify your achievements. Include a dedicated skills section. Get your contact information right.

Do this well, and you'll bypass the ATS filter and make it to human review. That's where the real evaluation happens. That's where context, nuance, and your true value as a candidate get recognized.

The ATS isn't your enemy. It's just a gate. And now you know exactly how to open it.