What Is ATS? How It Actually Reads Your Cover Letter

So what is ATS, and why does it decide whether a human ever opens your cover letter? An Applicant Tracking System is the software recruiters use to collect, sort, and rank every application before a person reads a single line. More than 98% of Fortune 500 companies run one. If you have applied to more than a handful of jobs this year, an ATS has already touched your file.

Here is the part most explainers skip: the ATS is not one villain with one rulebook. It is a database with a search bar, and the recruiter typing into that search bar decides what gets found.

What Does an ATS Actually Do?

An ATS stores your resume and cover letter, extracts the text into structured fields (name, dates, job titles, skills), and makes that text searchable. Recruiters then filter or search by keyword, job title, or years of experience.

The parsing itself runs on natural language processing, not a rigid list of exact-match rules. Older, rule-based parsers topped out around 65% accuracy and choked on anything but a plain single-column layout. The transformer-based engines running inside Greenhouse, Lever and Workday today hit closer to 97%, which is why some of the "never use two columns, never use a table" advice from five years ago is now outdated for the newest platforms, even if it still holds for older installs.

That distinction matters because it changes the fix. On an older ATS, a formatting mistake can genuinely scramble your dates and job titles beyond recognition. On a modern one, the bigger risk is not formatting at all: it is simply not using the words the recruiter searched for.

It is not a black box that auto-rejects you the moment you hit submit. In a Jobscan survey of recruiters, 99.7% said they use keyword filters inside their ATS to find candidates. That is a human decision layered on top of the software, not the software deciding alone.

Three numbers worth remembering:

• 76.4% of recruiters filter by resume skills first

• 59.7% filter by education

• 55.3% filter by job title

Miss all three and your application sits in the database, technically "received," practically invisible.

The Myth: "The ATS Rejects Most Resumes Automatically"

This one is everywhere, and it is mostly wrong for the platforms used by mid-size and large employers today. Modern engines (Greenhouse, Lever, Workday, Oracle Taleo) parse and rank. They rarely auto-reject on formatting alone unless a recruiter has set a hard filter, like a specific certification or a location radius.

What actually happens: your application gets a lower rank, sits on page four of the recruiter's search results, and never gets opened. Functionally similar outcome to a rejection. Mechanically, a very different problem, and one with a very different fix.

Does an ATS Even Read Your Cover Letter?

This is the question this site's readers actually need answered, and most ATS explainers do not bother. The honest answer: it depends on the platform, and treatment is inconsistent.

Some systems parse cover letters exactly like resumes, extracting text and making it searchable by keyword. Others store the cover letter as an attachment and never index it at all, meaning a recruiter has to manually open the PDF for your carefully matched company tone to count for anything.

Two practical rules follow from that:

• Never put information in your cover letter that is missing from your resume. If the ATS only indexes one of the two documents, the other one carries the weight alone.

• Repeat your three or four highest-priority keywords in both documents. Redundancy is not laziness here, it is insurance against a parser that ignores half your application.

The Formatting Choices That Actually Break Parsing

Forget vague advice like "keep it simple." Here is what breaks parsing, with numbers attached.

Two-column layouts fail on seven of eight ATS platforms tested by parsing researchers, because the engine reads left to right across the page and stitches unrelated lines from both columns together. A .docx file outperforms a PDF on six of eight systems, which contradicts the common advice to always submit a PDF. Headers and footers are a blind spot too: roughly a quarter of ATS platforms fail to correctly pull contact information stored there.

Formatting mistakes are the fixable half of this problem. The other half, keyword coverage, takes more judgment.

A smaller but common trap: date-format inconsistency. One documented parsing test found a resume where mixed date formats caused eight years of continuous experience to calculate as three, because the parser could not read "Jan '18" the same way it read "01/2018." Pick one date format, apply it everywhere on both documents, and move on. It is a two-minute fix that a keyword-perfect letter cannot compensate for if the underlying dates read wrong to the machine doing the ranking.

Do Small Companies and Startups Use an ATS Too?

Not always, and this is worth knowing before you spend twenty minutes optimizing a cover letter for a company of twelve people. Fortune 500 adoption sits above 98%, but a five-person startup hiring through a shared inbox or a lightweight tool like Notion is often reading every application by hand, no parser involved.

The practical signal: if you applied through Greenhouse, Lever, Workday, iCIMS or a branded careers portal with a structured form, an ATS is involved. If you emailed a resume directly to a founder or replied to a LinkedIn post, you are likely writing straight to a human, and pure keyword-matching starts to matter less than it does at scale. Company size is a better predictor of ATS involvement than industry.

This is also where company tone matching earns its keep. A keyword-perfect letter that reads like it was written for a 10,000-person enterprise falls flat at a startup, and a casual, personality-forward letter can look underprepared at a bank. ATS compliance gets you past the parser. Matching how the company actually talks is what gets you past the recruiter reading the shortlist.

What Recruiters Search For Once You're Ranked

Getting parsed correctly gets you into the pool. Ranking well gets you found. Recruiters searching an ATS database behave a lot like people searching Google: specific phrases first, broad terms as a fallback.

A resume or cover letter that contains the exact job title from the posting receives 10.6 times more interview requests than one that does not, according to Jobscan's analysis of its own user data. Not a synonym, not "similar role." The literal words from the job description, worked naturally into a real sentence.

The target for keyword density is 25 to 35 role-specific terms pulled from the job posting. Below that range, you rank low. Above it, you risk tripping a keyword-stuffing flag, which now runs on 83% of companies using AI-assisted screening.

How to Write a Cover Letter an ATS (and a Human) Can Actually Read

Start with the job title, stated plainly, in your first two sentences. That single move accounts for the biggest jump in the 10.6x interview-request stat above.

Then match language, not just keywords. If the job description says "cross-functional," do not write "collaborative." The ATS is matching text strings, and a recruiter scanning results afterward is matching for the same phrase they searched. Synonyms lose on both counts.

Keep formatting boring on purpose: standard fonts, no tables, no text boxes, no graphics carrying information the parser needs. Save the personality for your sentences, not your layout. Resume parsing accuracy varies sharply by platform, from 88% on Workday to 93% on Lever, and none of that variance comes from how creative your template looks.

One structural note specific to cover letters: name the company by name, and reference something concrete about the role, not generic enthusiasm. "I want to bring my project management experience to Acme's supply chain team" parses and reads better than "I am excited about this opportunity," which says nothing an ATS or a recruiter can act on.

Is a Perfect ATS Score Worth Chasing?

No, and here is the honest ceiling. A 75% keyword match rate is the widely cited target, and plenty of candidates land interviews at 65%. Past roughly 85%, you are usually padding with repetition that reads badly to the human who eventually opens the file. ATS score is a floor to clear, not a summit to conquer.

Jobscan's own research on ATS adoption puts Fortune 500 usage at over 98%, which means avoiding these systems entirely is not a real strategy for most job seekers. Working with the mechanics, correctly, for about twenty minutes per application, is.

What Actually Changes Your Callback Rate

Three things move the needle, in order of impact: the exact job title in your first two sentences, 25 to 35 keywords pulled straight from the posting and spread naturally across both documents, and formatting plain enough that a six-year-old parser and a tired recruiter can both make sense of it in ten seconds. Everything else, including which font looks "more professional," moves the needle by roughly nothing.

Run the same job description through your resume and your cover letter side by side before you submit either one. Read the posting once, highlight the repeated nouns (the tools, the certifications, the exact team or department name), then check both documents contain them in a real sentence, not a keyword dump at the bottom of the page. That single pass, done consistently, closes most of the gap between a 40% match rate and a 75% one, and it takes less time than rewriting your opening paragraph for the fifth time.

None of this replaces a cover letter that actually sounds like you. ATS-ready and generic are not the same thing. The goal is a document a parser can read cleanly and a recruiter wants to keep reading once they open it, in that order.