For decades, recruitment has been built on a familiar ritual. A vacancy opens. Candidates are invited to apply. CVs pour in. The first task is always the same: filter those CVs, read them, rank them, reject them, shortlist from them, and decide who is “worthy” of progressing.
This process is so deeply ingrained that most people never question the logic. Many recruiters would still say, almost by reflex, “Well, of course we start with the CV. What else would we use?”
The awkward truth is that the CV is one of the weakest predictors of future job performance that we have. The scientific literature has been clear about this for decades. What has changed more recently is that we now have something much better: robust personality profiling at the level of sub-factors (the narrow traits that sit underneath the broad Big Five). These sub-factors are more predictive, more efficient, and more closely linked to retention and long-term success than anything printed on a CV.
This article examines that evidence, and argues that the industry’s attachment to starting with the CV is no longer defensible if we want to be taken seriously as an evidence-based profession.
The uncomfortable truth: CVs are among the weakest predictors of performance
Despite their central place in recruitment, CVs are not supported by a strong research base. In fact, the opposite is true.
One of the most cited reviews in this area is Robertson and Smith’s 2001 paper in the Journal of Occupational and Organisational Psychology. They make an extraordinary observation: given how widely CVs and application forms are used, it is “astonishing” how little serious research has been done on them. Where they have been studied, the results are not flattering.
The studies that do exist show that CV evaluations are heavily influenced by factors which have nothing to do with real job performance, such as:
- Layout and formatting
- Use of “competency” buzzwords
- Writing style and phrasing
- Inclusion of a photo and physical attractiveness
- Irrelevant but positive-sounding qualifications and courses
In other words, CVs are excellent at capturing impression management. They are far less impressive at capturing genuine capability.
When researchers look for hard predictive validity, the numbers are consistently small. Meta-analytic work over several decades shows that:
- Education typically correlates with job performance at around r = .10 to .20
- Years of experience sit in the r = .15 to .20 range
- Training and qualifications rarely exceed r = .10 to .15 on their own
To translate that: these variables typically explain less than 5 per cent of the variation in job performance. They are not useless, but they are nowhere near strong enough to justify being the primary gateway into the recruitment funnel.
Yet in most organisations, they still are.
The false comfort of “experience” and “qualifications”
If we strip the CV down to its supposed essentials, three components usually dominate the early sift:
- Where someone has worked
- The job titles they have held
- The qualifications and training they have collected
These feel intuitively meaningful. They also fit very neatly with how hiring managers like to think: “We want someone who has already done this job, in this sector, for this long.”
Unfortunately, the evidence does not back up that comfortable story.
Experience does not equal performance
Van Iddekinge and colleagues conducted a large meta-analysis on pre-hire work experience, covering dozens of independent studies. The headline result was simple: amount of prior experience shows only a weak relationship with future job performance, and in some cases no meaningful relationship at all. It did not reliably predict turnover either.
People can accumulate years in a role without ever becoming especially competent, and experience in one context often fails to transfer cleanly to a different organisation, culture or system.
Training does not equal behaviour
The training section of a CV tells you that somebody attended a course, passed an exam, or collected a certificate. It does not tell you whether they:
- Understood the content
- Retained it
- Can apply it
- Will choose to apply it consistently under pressure
A long list of courses can impress on paper but carry very little predictive power once other factors are taken into account.
Qualifications do not equal capability
Educational attainment does relate to labour-market outcomes in broad terms, but its relationship with specific job performance is modest. Degrees are often more about signalling than about the day-to-day behaviours that drive performance in a particular role.
Most importantly, none of these CV components captures how someone is likely to behave once they are in post. They do not tell you:
- Whether they are reliable
- How they respond to stress
- Whether they cut corners
- How they handle conflict
- Whether they will fit the culture
- How they deal with ambiguity
- Whether they will stay
The CV is, fundamentally, a historical document. It tells you where someone has been. It does not directly measure how they will perform when they get here.
The science has moved on – hiring practice largely has not
While recruitment practice has remained anchored to the CV, occupational psychology has spent the last 60 years refining what actually predicts performance.
Three broad conclusions now enjoy strong empirical support:
- Job performance is strongly linked to stable behavioural traits
- Personality makes a meaningful contribution to prediction, especially traits related to conscientiousness and integrity
- More granular, “narrow” traits are often more predictive than broad personality factors
In the early decades, personality research was hampered by poor measures and inconsistent models. Results were mixed, and many practitioners concluded that personality “doesn’t work”. That view is now badly out of date.
From the 1990s onwards, with the emergence of the Big Five model and proper meta-analytic techniques, a very different picture emerged. Studies by Barrick and Mount, Tett and colleagues, Salgado and many others have shown that:
- Conscientiousness is a reliable predictor of performance across a wide range of jobs
- Emotional stability (low neuroticism) also contributes meaningfully
- Other traits show role-specific value (for example, extraversion in sales and customer-facing work)
The typical validities for broad traits sit in the r = .20 to .30 range. That is already significantly better than education and years of experience, and it reflects a consistent, generalisable relationship across occupations and cultures.
Crucially, personality measures do not produce the same level of adverse impact that cognitive ability tests can, which makes them attractive from a diversity and fairness perspective as well as a predictive one.
But the real improvement comes when we go below the broad Big Five factors and look at sub-factors.
Sub-factors: where prediction gets sharper
Broad traits are blunt instruments. “Conscientiousness” is a useful concept, but it lumps together many specific tendencies:
- Dependability
- Orderliness
- Achievement-striving
- Self-discipline
- Cautiousness
Similarly, “agreeableness” hides distinctions between diplomacy, empathy, trust and cooperativeness. “Emotional stability” covers resilience, emotional control, and sensitivity to stress.
Modern assessment moves deeper, measuring personality at the level of sub-factors. These are narrower, behaviourally specific traits such as:
- Resilience
- Work ethic
- Detail orientation
- Diplomacy
- Cooperation
- Empathy
- Emotional control
- Reliability
- Persistence
- Flexibility
These are much closer to the language managers use when they describe high and low performers. They are also the traits that show up again and again in performance reviews, disciplinary procedures, exit interviews and engagement surveys.
Research over the last two decades has consistently shown that:
- Sub-factors often show stronger correlations with job performance than the broad traits they roll up into
- They provide incremental predictive power, explaining variance that broad scores miss
- They support job-specific prediction, allowing tailoring to the behavioural demands of particular roles
In several meta-analytic summaries, sub-factors related to conscientiousness and integrity (for example diligence, honesty, reliability, impulse control) achieve validities in the r = .30 to .45 range, with some integrity-oriented measures approaching r = .50 in certain contexts. That is a long way ahead of anything typically found on a CV.
From prediction to outcomes: performance and turnover
It is one thing to say that personality sub-factors predict performance; it is another to connect them to concrete outcomes such as turnover, culture fit and profitability. In practice, the link is straightforward.
Low scores on certain sub-factors show up as:
- Absenteeism and lateness (low reliability and work ethic)
- Frequent errors and rework (low detail orientation, low conscientiousness)
- Customer complaints (low empathy, low diplomacy, low emotional control)
- Team conflict (low cooperation, low agreeableness)
- Burnout and stress-related exits (low resilience, high sensitivity to pressure)
- Misconduct, rule-breaking and ethical breaches (low integrity and impulse control)
These are precisely the behaviours that cost organisations money and drive unnecessary turnover.
Conversely, high scores on the relevant sub-factors underpin:
- Consistent attendance and punctuality
- High quality work with fewer mistakes
- Strong customer relationships
- Cohesive teams
- Better coping under pressure
- Lower disciplinary incidence
- Longer tenure and stronger retention
When organisations build selection processes that actively measure and prioritise these sub-factors, they typically see:
- Reduced early attrition
- Fewer behavioural issues
- Better cultural alignment
- Higher productivity per head
- More stable teams over time
The CV, by contrast, gives at best a hazy, indirect hint at these outcomes, and often none at all.
Efficiency: the overlooked advantage of sub-factor profiling
Discussions about selection methods often focus on validity, fairness and candidate experience. They should. But there is another dimension where abandoning CV-first hiring is a clear win: operational efficiency.
CV-based sifting is:
- Labour-intensive
- Time-consuming
- Cognitively draining for reviewers
- Highly inconsistent between reviewers
- Difficult to scale without loss of quality
- Almost impossible to automate without building yet another biased filter on top of a biased input
Sub-factor personality profiling, in contrast, is:
- Fast and standardised
- Scalable to high-volume campaigns
- Easy to compare across large pools
- Consistent across assessors and time
- Straightforward to automate while retaining rigour
Instead of spending hours manually reading CVs to work out who “sounds good on paper”, recruiters can:
- Define the key behavioural sub-factors that matter in the role
- Assess all applicants quickly and objectively
- Focus their time on a smaller pool of candidates who are statistically more likely to succeed and stay
This does not remove the need for interviews, work samples or reference checks. It simply means those resource-intensive activities are used on better-matched candidates.
The result is a recruitment process that is not just more accurate, but faster, cheaper and less frustrating for everyone involved.
Why hasn’t the industry already shifted?
Given the strength of the evidence, it is reasonable to ask why CV-first hiring remains the norm.
There are several overlapping reasons:
- Inertia and habit: the CV has always been there, so it feels natural to keep using it
- Perceived simplicity: managers think they can interpret a CV quickly
- Lack of awareness: many decision-makers simply do not know the data on validity
- Symbolic comfort: having a CV in the file feels like a minimum requirement, even if nobody can defend its predictive value
- Fear of change: adopting robust, data-driven methods makes some people uncomfortable because it exposes just how weak their traditional filters are
None of these reasons is evidence-based. They are psychological and cultural. That does not make them trivial, but it does mean they should not dictate professional practice indefinitely.
What an evidence-based recruitment process looks like
An evidence-based selection process does not abolish the CV completely, but it does stop using it as the first and primary decision gate.
A more defensible process looks something like this:
1. Use a simple application form to capture essentials (right to work, basic qualifications where truly required, key experience markers).
2. Run all suitable applicants through a robust personality assessment that measures relevant sub-factors aligned to the role.
3. Combine those results with a small number of critical “knock-out” criteria (for example, licences, legal requirements).
4. Invite the best-matched candidates to structured interviews and, where appropriate, work sample tests.
5. Use the CV as a reference document, not a decision engine: something to provide narrative context, not a basis for snap judgement.
This approach:
- Uses high-validity predictors early
- Reduces the risk of superficial CV impressions skewing decisions
- Focuses recruiter and manager time where it adds most value
- Improves quality of hire
- Reduces turnover by selecting for stable behavioural strengths rather than historical signalling
It is not theoretical. Organisations that have implemented this pattern consistently report better outcomes.
Conclusion: why start with the weakest tool?
If recruitment were being designed from scratch today, with full access to the last 60 years of research, it is very unlikely that the CV would be chosen as the primary screening tool.
It is subjective, unstandardised, easy to manipulate, and weakly predictive. It reflects where someone has been, not how they will behave.
Personality profiling at the sub-factor level, by contrast, is objective, behaviourally relevant, statistically robust and operationally efficient. It predicts performance and retention in a way the CV simply cannot.
The profession now faces a straightforward question:
If we know that CV-first hiring is both inefficient and statistically flawed, and we know that personality sub-factor profiling is more accurate, more efficient and better correlated with reduced turnover,
