35% of UK job applications miss interview threshold

JobSpace AI has published research showing that 35% of UK job applications fall below the threshold needed to progress to interview. The findings are based on an analysis of 5,782 CV scans matched against UK job descriptions.

The data challenges the long-circulated claim that 75% of CVs are rejected automatically before a recruiter reads them. Instead, the figures suggest most applications in the sample reached a level classed as interview-ready, while a sizeable minority did not.

Of the 5,782 CVs analysed, 64.5% scored 75 or above, which JobSpace AI classed as interview-ready. Another 22.4% scored between 50 and 74 and were deemed at risk of rejection, while 13.1% scored below 50 and were considered likely to be filtered out before reaching a recruiter.

The research drew on CV scans submitted by UK job seekers and assessed against job descriptions supplied by the same users. The sample covered submissions made over a four-month period and was based on actual candidate documents rather than recruiter surveys or modelled estimates.

Keyword gap

A smaller subset of 248 CVs received full keyword analysis. In that group, candidates matched an average of 48% of the keywords in the job descriptions they targeted and missed 9.1 keywords per application on average.

The missing terms were most often linked to process and governance rather than technical expertise. Phrases such as continuous improvement, compliance, customer service, SLA or service levels, change management, and stakeholder management appeared regularly in job adverts but were often absent from applicants’ CVs.

That pattern suggests the issue for many applicants lies less in their underlying experience than in how they describe it. Recruiters and screening systems often look for the language used in role specifications, especially in functions where process, oversight, and service delivery feature heavily.

“The gap most candidates don’t see isn’t a skills gap – it’s a language gap,” said Nicholas Barooah, Founder, JobSpace AI.

“Job adverts are written around frameworks and processes. Most CVs describe what someone achieved without using the governance and process terminology recruiters are screening for. Candidates who bridge that gap move from the 35% to the 65% – often with relatively small changes to how they describe existing experience,” Barooah said.

Myth questioned

The findings also cast doubt on one of the most frequently repeated claims in careers advice: that three quarters of CVs are screened out automatically. According to JobSpace AI, that figure has circulated for years across careers media, social media posts, and CV-writing services, but lacks a traceable primary source.

Its analysis points to a different picture. Automated filtering remains part of recruitment practice, but the results suggest the bigger issue is not universal exclusion by software. Instead, a notable share of applicants may be weakening their prospects by failing to reflect the wording and priorities set out in job adverts.

That distinction matters because it shifts attention away from the idea of a closed system and towards one in which many applications can be improved. For candidates whose CVs fall into the middle band or lower-scoring group, the data suggests relatively modest revisions in terminology and alignment may affect whether an application progresses.

How scoring worked

The scoring model assessed keyword alignment, formatting compatibility, and role-seniority match. Each CV was measured against a real job description, and the resulting score was used to place the application into one of three categories.

The research focused on UK users and was intended to reflect real-world submissions rather than hypothetical tests. Because job seekers provided the documents voluntarily, the dataset offers a snapshot of how candidates are currently presenting themselves in live applications.

The figures also underline the competitive nature of recruitment, even when most CVs are not screened out immediately. A document that reaches a recruiter is not necessarily a strong contender, particularly when employers compare applicants on closely matched wording, evidence of process knowledge, and relevance to the stated brief.

For applicants, the results point to the importance of reading job descriptions closely and mirroring terms that accurately reflect their experience. The most commonly absent phrases in the sample were not specialist jargon, but standard language around operations, governance, and delivery.

JobSpace AI said its platform has analysed more than 5,000 real UK job applications since launch, and the latest sample adds to that picture by quantifying how many candidates may be missing interview thresholds because of wording rather than lack of experience.