Why ATS Screening Rejects 75% of UK CVs (And the 6 Fixes That Actually Work in 2026)

75% of UK CVs never reach a human recruiter. They are rejected by Applicant Tracking Systems (ATS) in under 5 seconds, based on rules most candidates have never been told about.

This is not a mystery, and it is not permanent. ATS rejection follows specific patterns. Once you know what the scanner actually looks for, you can fix your CV in an afternoon and move into the top 25 percent that reaches a human.

This post covers six specific fixes that move the score. Not "use keywords". Not "keep it simple". Specific, tactical changes you can make today.

What ATS Actually Does to Your CV

The software most UK employers use (Workday, Greenhouse, Oleeo, Taleo, iCIMS) runs the same basic pipeline on every CV it receives:

1. Parse the file. Convert the PDF or DOCX into structured text.
2. Identify sections. Experience, education, skills, contact.
3. Match keywords. Score against the job description's must-have and nice-to-have terms.
4. Flag anomalies. Weird formatting, missing sections, suspicious patterns.
5. Assign a score. Candidates below the threshold get filtered out before a human sees them.

Each of the six fixes below addresses a specific point where CVs fail this pipeline.

Key Takeaway: ATS is not a mystery box. It is a five-stage pipeline with known failure modes. Your CV gets rejected at one of those five stages, not because of some opaque algorithmic judgment.

Fix 1: Kill the Two-Column Layout

Two-column CVs look modern. Most parsers cannot read them properly.

When the parser converts a two-column PDF into structured text, it reads left-to-right across columns, not top-to-bottom down each. Your job titles land in the middle of someone else's skills. Your dates land between bullet points. The parser outputs garbage, and the scoring fails.

Test this yourself: copy the text from your CV PDF using Ctrl-A and Ctrl-C, then paste into a plain text document. If the result reads in the correct order, your CV is safe. If it jumbles, the ATS is getting the same jumble.

Fix: single-column layout. Boring is safer than beautiful. The visual polish of a two-column design costs you the chance to be scored accurately.

Fix 2: Remove Tables, Text Boxes, and Graphics

Tables confuse parsers. Text boxes get skipped entirely. Graphics (including skill-bar charts) get ignored.

If your skills are in a decorative bar chart with percentages, the ATS does not see them. If your contact details are in a side text box, they might be missing from the parsed output entirely. If your timeline is a table, date matching breaks.

Common culprits we see in rejected CVs:

• Skill-level bar charts ("React: 90% proficient")
• Contact info in a coloured sidebar text box
• Experience presented in two-column tables (role on left, bullets on right)
• Icons next to section headings
• Decorative lines that the parser reads as section breaks in weird places

Fix: plain text structure throughout. Headings as text. Bullet lists as real bullet lists. Contact info as a normal paragraph at the top. CVPilot's template parseability audit flags each of these automatically and scores the CV on all seven technical factors before it scores on content.

Key Takeaway: 43 percent of CV rejections in 2026 come from formatting issues, not qualifications. A perfectly qualified candidate with a fancy design can lose to a less-qualified candidate with a boring template.

Fix 3: Use the JD's Exact Verbs and Phrases

ATS scores heavily on exact-match. If the job description says "project management", the system gives more weight to that exact phrase than to "managing projects" or "project leadership".

This is not laziness on the scanner's part. It is how ATS has been tuned. Recruiters often do an exact-phrase search on top of the algorithmic score to filter down further. Close approximations are close misses.

Specific examples we see weekly:

• JD says "stakeholder management". CV says "worked with stakeholders". Miss.
• JD says "architect scalable systems". CV says "built scalable systems". Miss.
• JD says "agile methodologies". CV says "agile workflows". Partial miss.
• JD says "data-driven decision making". CV says "used data to make decisions". Miss.

Fix: for every must-have requirement in the JD, find the exact phrase, and use it verbatim somewhere in your CV where it is genuinely applicable. Not forced. Genuinely applicable.

This is the single highest-leverage change most candidates never make, because "re-word to match" feels unnatural. It is not. It is basic vocabulary alignment.

Fix 4: Order Skills by the JD's Priorities, Not Your Own Chronology

Most candidates list skills in the order they learned them. A React developer lists React first. A Python developer lists Python first.

If the JD weights cloud platforms as must-have and frontend frameworks as nice-to-have, your skill order should follow. AWS, Azure, GCP before React, Vue, Angular. Even if you have spent more time in React.

ATS scans skills in order. Skills in the first 20 tokens weight more heavily than skills later in the list. Position matters as much as presence.

Fix: rewrite the skills section for each job application. Read the JD. List its must-haves in order. Then paste the relevant ones from your own skill set, in that order. Secondary skills follow.

Key Takeaway: Your skills section is ordered by the job, not by you. Reordering takes 3 minutes per application and is the second-highest leverage change you can make.

Fix 5: Standard Section Headings Only

"My Journey" instead of "Experience". "What I Know" instead of "Skills". "The Story So Far" instead of "Summary".

These creative section titles signal personality. To ATS, they signal "unrecognised section, cannot score". Your whole experience section might be invisible to the scanner because the heading does not match the expected list.

The scanner is looking for the standard headings: Experience, Work Experience, Employment History, Education, Skills, Professional Summary, Certifications, Projects. Anything outside this vocabulary risks being parsed into the wrong section or ignored entirely.

Fix: keep the personality in the content, not the headings. Your writing voice shows through in your bullets, not in your section titles. A section labelled "Experience" with strong content beats a section labelled "Adventures" with the same content.

Fix 6: Quantify Every Bullet Where Possible

ATS ranks quantified achievements higher than general ones. A bullet with a number or a percentage scores better than a bullet without.

This is partly algorithmic (numerical tokens get weight) and partly recruiter behaviour (humans reviewing the top of the filtered pile look for evidence of impact).

Weak:

• Led marketing campaigns for B2B clients.
• Improved team productivity through new processes.
• Built scalable applications used by many users.

Strong:

• Led 12 B2B marketing campaigns, generating £2.4m in pipeline over 18 months.
• Improved team productivity by 28 percent after implementing a new sprint review process in Q2.
• Built scalable applications used by 40,000 monthly active users across three product lines.

Not every bullet can have a number. Most can. If a bullet cannot be quantified, it can usually be specified instead: named project, named tool, named client (if not under NDA), specific timeframe.

Fix: audit your CV and find every bullet without a number or named entity. Rewrite at least 60 percent of them with specifics. The remainder can stay descriptive.

The Order to Apply These Fixes

If you have a limited amount of time this week, apply the fixes in this order:

Fixes 1, 2, 5 and 6 are one-off investments. You fix the CV once and reuse the structure. Fixes 3 and 4 are per-application changes, 30 minutes of tailoring per job you apply to.

Candidates who skip the per-application tailoring almost never beat ATS. The tailoring is not optional. It is the scoring mechanism.

How to Test Whether Your CV Would Pass

Two ways to check without waiting for a rejection:

Method 1: the copy-paste test. Open your CV PDF. Press Ctrl-A, Ctrl-C. Paste into a plain text document. Does it read in the correct order? Does every section appear? Are your contact details included? If yes, the basic parseability is sound.

Method 2: automated parseability audit. CVPilot's template parseability score runs seven checks automatically: single-column layout, no tables, standard headings, reasonable word count, extractable text, no glyph corruption, contact info visibility. A passing score is 100. Most candidate CVs score 60 to 80 on the first scan.

Both methods catch 80 percent of parseability failures. Neither catches keyword misalignment, which is where the real score is won.

The Deeper Issue Most Advice Misses

Most "beat the ATS" advice focuses on the scanner itself. In 2026, that is only half the problem.

Even CVs that pass the ATS get filtered further. Recruiters typically review the top 30 of 200 filtered candidates. They spend 6 to 15 seconds on each CV. What gets them to open your document is the ATS pass. What gets them to shortlist you is quantified impact, specific verbs, and evidence of the exact skills the role needs.

Both layers matter. A CV that passes ATS but fails recruiter review is still a rejection, just a slightly later one.

Key Takeaway: ATS rejection is the obvious failure. Passing ATS and failing recruiter review is the invisible one. Both are solved by the same six fixes, applied consistently.

Your Next Step

Apply the six fixes to your current CV this week. Test it with the copy-paste method first, then with an automated audit.

If you want the automated audit run for you in 60 seconds, with a score on all seven parseability checks plus keyword coverage against a specific job description, CVPilot is free to try.

Ready to optimise your CV for 2026 ATS? Try CVPilot free and see your ATS score in under 60 seconds.