SS
Key Takeaways
- A low score on a well-written resume has three possible causes: the parser cannot read parts of it, your vocabulary differs from the posting's, or your evidence sits too far down the page.
- Design quality and parse quality are unrelated, and often inversely correlated - the most attractive resumes frequently parse worst.
- You can identify which cause applies in about two minutes using a copy-paste test and a vocabulary comparison.
- A parsing problem usually takes one edit to fix; a vocabulary problem takes three or four; a placement problem takes none at all, just reordering.
- A low score on a genuinely mismatched job is correct information, not a defect to engineer around.
A resume that reads well to you and scores badly in a scanner is not a contradiction. You and the scanner are evaluating different things.
You read meaning. It matches text. When those diverge, the score falls while the document still looks fine on your screen.
There are only three common reasons for that divergence. Each needs a different fix, and working out which one you have takes about two minutes.
Cause 1: The Parser Cannot Read Part of Your Resume
This is the most common cause and the easiest to miss, because nothing looks wrong.
Text inside a text box, a table cell, a header or a footer may render perfectly in your PDF viewer and arrive as nothing on the other side. The two-column layout is the usual culprit — a parser reading left-to-right across the page can interleave the columns into nonsense, or drop the narrow one entirely.
How to check it
Open your PDF, select all, copy, and paste into a plain text editor. What you see is roughly what the parser sees. The full method is in the 30-second copy-paste test.
Look for:
- Sections that vanished entirely
- Contact details missing, which happens when they sit in a page header
- Columns interleaved into scrambled lines
- Bullet characters becoming odd symbols
- Dates separated from the jobs they belong to
If your skills section is absent from that paste, no amount of keyword work will raise your score. The scanner never saw those words.
The structures that cause it
| Structure | Risk | Why it fails | Fix |
|---|---|---|---|
| Two columns | Critical | Reading order guessed from coordinates | Single column for experience |
| Text boxes | Critical | Separate objects, often skipped | Retype as body text |
| Tables | High | Cell order flattened inconsistently | Plain lines with spacing |
| Header/footer | High | Region often ignored entirely | Move into document body |
| Image-only export | Critical | No text layer at all | Re-export as real text |
| Custom bullet glyphs | Low | Arrive as stray symbols | Standard round bullet |
Our guide to ATS-friendly formatting covers the layout rules in more depth, and tables and columns explains exactly where parsers break.
The fix
Single column for your work history. Contact details in the body, not the header. Plain lines instead of tables. Standard fonts.
This is a structural fix, not a design surrender — a clean single-column resume looks entirely professional. See choosing a resume template for layouts that hold up.
Cause 2: You and the Posting Use Different Words
Your resume says "client relationship management." The posting says "stakeholder management." You have done the work. The scanner does not credit it, because it matches the posting's vocabulary rather than assessing your career.
This is not about adding skills you lack. It is about naming skills you have in the language the reader uses.
A worked rewrite
Your resume: Managed relationships with 12 enterprise clients, resolving escalations and running quarterly reviews.
The posting asks for: stakeholder management, account governance, QBRs
Rewritten: Owned stakeholder management for 12 enterprise accounts, running quarterly business reviews (QBRs) and resolving escalations.
Same job, same truth, same achievements. The second version is legible to a system matching against that posting, and it reads more professionally besides.
Two cautions
Only where it is true. If you have never run a QBR, do not write QBR. The line between framing and fabrication is the one rule that matters here.
Do not do it mechanically. A resume that reads like the job description pasted back at the employer passes the scanner and loses the human. Keyword stuffing versus coverage covers where that line sits.
Common vocabulary mismatches
| You probably wrote | The posting may say |
|---|---|
| Client relationships | Stakeholder management |
| Reporting | Business intelligence, analytics |
| Supplier contracts | Vendor management, procurement |
| Training new staff | Onboarding, enablement, mentoring |
| Fixing processes | Process improvement, operational excellence |
| Working with other teams | Cross-functional collaboration |
| Gathering requirements | Requirements elicitation, discovery |
The best resume keywords to beat ATS guide lists more of these by industry.
Cause 3: Your Evidence Sits Too Low
Some systems weight the top of the document more heavily, and every human reader does. If the skill the posting cares about appears once, in the fourth bullet of your third job, it is technically present and practically invisible.
The fix is placement, not addition. Move the relevant evidence up — into your summary, your skills section, or the first two bullets of your most recent role.
This is the cheapest fix available. You are not writing anything new; you are reordering what is already there. Since recruiters decide in about six seconds, what sits in the top third of page one does most of the work.
Working Out Which One You Have
Run these in order. The first is the most common and the cheapest to check.
Step 1: Run the copy-paste test
If content is missing or scrambled, that is your problem. Stop here and fix the layout. Nothing else matters until the document arrives intact.
Step 2: Compare vocabulary
Put the posting beside your resume. Take its five most-repeated terms and check whether those exact words appear on your resume. If you do the work but use different words, rewrite those lines.
Step 3: Check placement
For the two or three things the posting cares about most, ask how far down the page the evidence sits. If it is past the first half, move it up.
Step 4: Accept the honest answer
If none of the three applies and the score is still low, the match is probably genuinely weak. That is useful — it tells you to spend your time on better-fitting postings rather than on this document.
Diagnostic Summary
| Symptom | Likely cause | Fix effort | Score impact |
|---|---|---|---|
| Sections missing from paste | Parsing | One layout edit | Large |
| Paste clean, terms absent | Vocabulary | 3–4 rewrites | Moderate |
| Terms present but late | Placement | Reordering only | Small on score, large on humans |
| Everything clean, still low | Genuine mismatch | None available | — |
What This Is Not
It is not a reason to strip all formatting. A plain single-column resume in a standard font parses reliably and still looks good. You are removing structural traps, not design.
It is not a reason to chase a perfect score. Past a reasonable threshold, extra keyword work produces diminishing returns and starts to damage readability. What counts as a good score covers where to stop.
It is not a reason to add skills you do not have. A scanner can be satisfied by a word. An interview cannot. If a posting names something you lack, there are honest options.
A Worked Diagnosis
An operations analyst scoring 54% on a role she was well qualified for.
Step 1 — copy-paste test. Her skills section was missing entirely. It had been built as a shaded text box in the template she bought. That alone explained a large part of the gap, because every technical term she had was invisible.
Step 2 — vocabulary. With the skills section retyped as body text, the score moved to 68%. Comparing terms, the posting asked repeatedly for "process improvement" and "stakeholder management." Her resume said "made things more efficient" and "worked with other departments."
Step 3 — placement. Rewriting those two lines took the score to 79%. Her strongest evidence — a project that cut a month-end close from nine days to five — was the fourth bullet of her current role. Moving it to first changed nothing on the score and changed the first thing a recruiter reads.
Total time: about twenty minutes. Two of the three fixes were structural or lexical, not content. She did not add a single skill she did not have.
How This Differs by Role Type
The three causes are universal, but which one bites you depends on the field you are in. Knowing your field's typical failure saves a diagnostic round.
Design and creative roles
Parsing is almost always the cause. Designers build resumes in design tools, and those tools produce exactly the structures parsers cannot read — frames, grids, custom glyphs, occasionally flattened images.
The tension is real: your resume doubles as a work sample. The workable compromise is two files — a designed PDF for direct applications and human review, and a clean single-column version for portals. Your portfolio carries the visual case, so the resume does not have to.
Engineering and data roles
Vocabulary is the usual cause, because tool names are unforgiving. A posting asking for PostgreSQL and a resume saying "SQL databases" will not match, even though you obviously qualify.
Name specific technologies and versions where the posting names them. How specific your skills should be covers where the line sits.
Management and operations roles
Placement is the usual cause. Senior work is described in long paragraphs about scope, and the concrete evidence ends up in the middle of a dense block.
Lead with scale — budget, headcount, remit — in the first bullet of your current role.
Healthcare, legal and regulated fields
Credentials are the binding constraint, and they are frequently mis-parsed because they sit in a right-aligned column or a table. A licence number that does not arrive is functionally a missing licence.
Put registrations in plain body text near the top, not in a sidebar.
Career changers
Often all three at once, which is why the diagnosis feels hopeless. Work the order strictly — parsing, then vocabulary, then placement — rather than rewriting everything. Translating experience into a new industry's language is the vocabulary step for this case specifically.
Common Mistakes When Diagnosing This
Rewriting content before checking structure. If the parser cannot read your skills section, rewriting it changes nothing. Always test parsing first.
Adding keywords to the skills list only. A term that appears in a list and nowhere in your described work is weak evidence to a human, even when it satisfies a scanner.
Assuming a professionally designed resume is safe. Designers optimise for visual impact, which means columns, text boxes and custom bullets — the exact structures that parse worst.
Testing the wrong file. Your source document may be fine while the export is not. Test what you actually send.
Treating the score as a verdict. It approximates keyword relevance for one posting. It does not predict whether you will be interviewed.
Fixing all three causes at once. Change one variable, re-scan, and you learn which one mattered. Change everything simultaneously and you learn nothing — and you will repeat the whole exercise on your next application because you still do not know what was wrong.
Assuming a clean paste means a good resume. Parsing is a floor, not a standard. A perfectly readable resume full of duty statements will still be ignored by the human who reads it next.
Frequently Asked Questions
My resume was designed by a professional and still scores badly. Why?
Design quality and parse quality are unrelated, and often inversely correlated. Designers optimise for how the page looks, which leads to columns, text boxes and custom bullets — the structures parsers handle worst.
Does a low score mean a human never sees my resume?
Not usually. Most systems rank rather than reject outright, so a low-scoring resume sits further down a long list rather than being deleted. On a posting with hundreds of applicants the practical effect can be similar.
Should I use the exact words from the posting even if they sound clumsy?
Use them where they are accurate and can be written naturally. If a phrase cannot be written without sounding forced, the cost to the human reader usually outweighs the gain with the scanner.
Will converting my resume to .docx fix a low score?
Only if your problem was a PDF exported as an image or with unselectable text. If the copy-paste test returns clean text, the file format is not your issue. See the ATS file format guide.
How much should my score move after fixing these?
A parsing fix often moves it substantially, because whole sections become visible for the first time. A vocabulary fix moves it moderately. A placement fix moves it least with scanners but most with human readers.
Can the same resume score differently on two jobs?
Yes, and it should. The score is a comparison against one posting, so a different posting is a different comparison rather than a change in your resume's quality. A twenty-point spread across two roles with the same job title is entirely normal, and usually reflects how differently the two were written.
Is there a length that scores better?
Not directly, though very short resumes have less evidence to match and very long ones dilute relevance. The resume length guide covers what actually matters.
How long should the whole diagnosis take?
About fifteen minutes. Two minutes for the copy-paste test, five comparing vocabulary against the posting, and the rest reordering. If you are spending an hour, you have skipped the structural check and gone straight to rewriting.
Should I fix the base resume or tailor per application?
Both, in that order. Parsing and bullet quality are base-resume problems — fix them once and every future application benefits. Vocabulary and placement are per-application, and take about five minutes each time.
My score is low on every posting, not just one. What does that mean?
A consistent pattern across ten or more well-matched postings points at the base resume rather than any individual application. Run the parsing check first; if it is clean, the issue is usually that your bullets describe duties rather than outcomes.
What if my score is fine but I get no interviews?
Then the scanner stage is solved and the problem is downstream — usually bullet quality or targeting. High score, no interviews covers what to check next.
What to Do Next
Work in the order above: parsing, then vocabulary, then placement. Most people skip straight to keywords, which is why they rewrite for an hour and the score barely moves.
Once you know the document parses cleanly, the remaining question is narrow and answerable: which of this posting's terms are genuinely absent, and which of those can you honestly claim?
Scan your resume against the posting — the found and missing lists separate a parsing problem (almost nothing found) from a vocabulary problem (found, but not the posting's terms) immediately.
Free, about a minute. If your layout turned out to be the issue, the resume templates are built single-column for exactly this reason.
Find out why your score is low
Scan your resume against the job description and see exactly which checks fail.
Drop your resume here or choose a file
PDF only. Max 2 MB.
We never share your data or use it to train AI models.
Templates that keep this structure intact
This template is ATS-tested — start from it and the formatting rules in this guide are already handled.
Was this guide useful?
Be the first to rate it.
