Writing a CV That Gets Past Both AI and Human Screening

Modern UK applications go through two readers, and they could hardly be more different. The first is an algorithm that parses your CV in milliseconds, hunting for structure it can understand and terms it can match. The second is a tired human who gives the survivors somewhere between six and thirty seconds of attention. A CV that delights one can quietly fail the other, and the requirements differ in subtle ways.

The good news is that the two readers want mostly the same thing: a clear, well-organised account of what you have done and why it matters. Get the fundamentals right and you satisfy both at once. This post covers those fundamentals, in the order the two readers encounter them.

Reader One: What the Algorithm Looks For

An applicant tracking system (ATS) does two jobs. First it parses your CV, converting the document into structured data: name, contact details, job titles, employers, dates, skills, qualifications. Then it scores or filters that data against the vacancy.

Parsing is where most avoidable failures happen. If the software cannot read your CV cleanly, nothing that follows can rescue you. The main culprits are predictable.

• Tables and text boxes. Many parsers read tables left to right across cells, scrambling your carefully arranged two-column layout into nonsense.
• Headers and footers. Some systems ignore them entirely. If your phone number lives in the header, it may never reach the recruiter.
• Graphics, icons and skill bars. A five-star rating for “communication” reads as nothing at all to a parser.
• Non-standard section headings. “My Journey” might charm a human. Software is looking for “Work Experience”, “Education” and “Skills”.
• Unusual fonts and heavy design. Decorative templates, particularly those built in design tools and exported as flattened files, often parse poorly.

The safe format is plain: a single column, standard headings, a common font, dates in a consistent format (month and year), and a .docx or text-based PDF file. It looks unremarkable. That is the point. The design effort you save can go into the words instead.

Once parsed, the ATS looks for relevance: job titles that match or relate to the vacancy, skills and qualifications named in the advert, and recency of experience. Which terms to include, and how many, is its own subject; our guide to choosing CV keywords without triggering red flags covers it in depth.

Reader Two: What the Human Looks For

The human reader is not scoring you against a list. They are forming a judgement, quickly, about whether you look like someone worth a phone call. Research from UK job board audits consistently finds that recruiters spend well under a minute on a first read, and often under ten seconds.

In that window, they are scanning for a handful of signals. Does your most recent role make sense for this vacancy? Do your job titles show a coherent trajectory? Are there numbers that suggest real responsibility, such as budgets, team sizes, or results? Is the CV free of clutter, waffle and obvious errors?

Humans also notice what algorithms cannot: tone. A CV stuffed with imported buzzwords reads as evasive. Plain English, specific claims and short sentences read as confidence. If a bullet point could appear on anyone’s CV, it is doing nothing on yours.

Where the Two Readers Agree, and Where They Conflict

Mostly, the readers align. Clear structure helps the parser and the skimmer alike. Relevant terminology satisfies the matching engine and reassures the recruiter that you speak the industry’s language. Recent, relevant experience placed prominently serves both.

The conflicts are narrower than the folklore suggests, but they exist.

• Repetition. An ATS may reward a term appearing several times. A human finds the third repetition grating. Resolve this in the human’s favour: use the exact term once or twice, then vary it naturally.
• Length. Software does not care if your CV runs to four pages. UK recruiters do, and two pages remains the standard for most roles. If you are using AI to draft or rewrite sections, read our guide to the two-page CV rule when using AI, because AI tools expand by default.
• Design. A touch of visual polish helps a human but risks the parser. Keep any design minimal and structural: clear headings, sensible spacing, nothing that moves text out of the main column.

The Top of Page One Rule

Both readers concentrate their attention at the top of the first page. The ATS typically gives extra weight to your current or most recent role. The human may read nothing else before deciding whether to continue.

That top third should therefore carry your strongest, most relevant material: your name and contact details in the body of the document, a short professional summary aligned to the role, and your current position with two or three achievement-led bullet points. Anything generic, historic or marginal belongs further down, or nowhere.

This is also why a single CV rarely performs well across different vacancies. What deserves the top spot changes with each role. The process of adapting an existing CV to a specific advert is covered in our guide to mirroring job description language with AI.

How to Test Your CV Before Submitting

You do not need to guess whether your CV will survive screening. A few checks take minutes.

• The copy-paste test. Select the entire document and paste it into a plain text editor. If sections arrive out of order, contact details vanish, or bullet text merges into a block, a parser will struggle too.
• The heading check. Confirm your section headings are conventional and your dates follow one consistent format throughout.
• The six-second skim. Hand the CV to someone who does not know your work and give them six seconds. Ask what job you are applying for and what your biggest achievement is. If they cannot answer, the top of page one needs work.
• An ATS scan. Dedicated tools will parse your CV the way screening software does and report what was read, what was missed and how well it matches a given vacancy.

If you are hearing nothing at all from applications, screening failure is the most likely explanation, and it is fixable. We have written a fuller diagnosis in why your CV isn’t getting past initial screening.

A Faster Way

Everything above can be done by hand. Plenty of candidates rebuild their CV in a plain format, run the copy-paste test, and rewrite their top third for every application. It works; it just costs an evening per vacancy.

Zappli was built to do this work in one pass. The free Diagnose tier runs an ATS scan on your CV and returns a score plus one rewritten section, with no card required. If the diagnosis shows deeper problems, a £7.99 seven-day Pro-Pass unlocks full rewriting, or Pro-Monthly at £11.99 a month covers unlimited bespoke CVs, cover letters and interview prep. The Agent tier at £24.99 a month goes further and drafts tailored applications overnight while you sleep. Zappli is paid by candidates, never by employers, so the advice runs in your interest.

If a free diagnosis sounds more useful than pasting your CV into a text editor and squinting at the result, the link below is the place to start.