Why Organisations Have Never Invested More in Hiring-
Yet Continue to Miss the Best Talent
Dr Vassilia Orfanou, PhD, Post Doc, COIO, LUDCI.eu
Writes for the Headline Diplomat eMagazine, LUDCI.eu
From Governance Failure to Everyday Practice
The first article in this series argued that recruitment is not simply an HR activity. It is a form of human-capital governance. The systems used to identify and select people reflect decisions made across the organisation: executive priorities, financial constraints, legal controls, procurement choices, technology design and hiring-manager behaviour. When those elements are poorly aligned, recruitment can remain procedurally compliant while repeatedly producing weak outcomes.
Systemic recruitment malpractice rarely announces itself as a dramatic breach. More often, it appears as normal practice: a vacancy that was never genuinely open, a six-stage interview process without a clearer decision, an algorithm deployed without validation, or a candidate rejected because “fit” was interpreted as familiarity. Each incident may appear minor. Repeated at scale, however, these practices distort labour-market signals, waste candidate time and reduce the probability that organisations will recognise the talent they claim to need.
Ghost Jobs: When a Vacancy Becomes Fiction
A job advertisement normally communicates a clear market signal: an employer has work to be done and intends to hire someone to do it. Ghost jobs break that assumption. The term describes vacancies advertised without a firm or immediate intention to appoint. They may remain online to build a future talent pipeline, test salary expectations, demonstrate growth or reassure overworked teams that support is supposedly coming. A 2024 research preprint examining Glassdoor data estimated that as many as 21% of advertisements in its sample could display characteristics associated with ghost hiring. Because the study is a preprint, the precise figure should be treated cautiously; the underlying governance concern should not.
A vacancy published without a genuine hiring mandate transfers the cost of organisational uncertainty to candidates. Applicants tailor CVs, write cover letters, disclose personal data and often undertake assessments for an opportunity that may not exist in any meaningful sense. When the company simply aims to obtain market intelligence or create a pool of potential applicants; the candidate receives false hope. This is more than poor communication. It is an asymmetry of intent. It is lack of transparency.
The damage is intensified by a wider crisis of trust. Greenhouse’s 2024 State of Job-Hunting report found that candidates increasingly encounter ghost jobs, automated barriers and unexplained silence. When job advertisements cease to function as reliable statements of intent, the labour market becomes noisier and less efficient for everyone-including employers with genuine vacancies.
Interview Theatre: More Stages, Not More Evidence
At the other end of the process lies interview theatre: the assumption that additional stages necessarily produce better judgement. Candidates are asked to complete screening calls, panel interviews, psychometric tests, presentations, case studies, “culture” discussions and final executive meetings. The accumulation of stages creates the appearance of rigour. Yet rigour is not measured by how much candidate time a process consumes; it is measured by whether each stage adds valid evidence.
The classic Schmidt and Hunter meta-analysis demonstrated that selection methods differ substantially in predictive validity. Structured interviews and work samples can be useful because they connect assessment to defined job requirements. Repeated unstructured conversations, by contrast, may simply multiply intuition, similarity bias and inconsistent questioning. Google’s structured-interview guidance reflects the same logic: use consistent questions, clear rubrics and comparable evidence rather than relying on conversational chemistry.
Interview theatre becomes systemic malpractice when organisations cannot explain what each stage measures, why it is necessary or how it improves the final decision. The candidate experiences delay and uncertainty; the organisation incurs management time and risks losing strong applicants. A lengthy process may reduce anxiety for decision-makers, but that does not mean it reduces hiring error.
AI Without Governance: Automating Yesterday’s Bias
The most prominent warning came from Amazon. A Reuters investigation reported that the company abandoned an experimental recruitment engine after it learned patterns pertaining to disadvantaged women. Trained on historical applications from a male-dominated technology workforce, the system penalised indicators such as the word “women’s” in CVs. The algorithm did not invent the inequality; it converted historical preference into automated ranking.
The case matters because automation changes the scale and opacity of error. A biased manager can affect a limited number of decisions. A biased system can screen thousands of people before any human reviews the outcome. The apparent neutrality of mathematics may also discourage a challenge: a score looks objective even when its inputs, assumptions and validation are weak.
Regulators have begun responding. The U.S. Equal Employment Opportunity Commission warns that software and AI used in employment decisions can violate anti-discrimination protections, including where tools disadvantage people with disabilities. New York City’s Automated Employment Decision Tools rules require bias audits and candidate notices for covered systems. These measures express a basic governance principle: an employer cannot outsource accountability to a vendor or an algorithm.
“Culture Fit”: When Familiarity Masquerades as Merit
Few phrases sound more harmless than “culture fit”. Teams need shared standards, workable relationships and alignment with organisational values. The problem arises when fit is left undefined. Lauren Rivera’s influential study, Hiring as Cultural Matching, found that employers in elite professional-service firms assessed candidates not only for skills but also for cultural similarity in leisure interests, experiences and self-presentation. Familiarity became evidence of suitability.
This creates a self-replicating organisation. Decision-makers recruit people who resemble those already inside, then interpret the resulting homogeneity as proof that the selection model works. Candidates who would challenge assumptions or broaden the organisation’s capabilities can appear risky precisely because they are different. A stronger standard is culture contribution: not “Does this person resemble us?” but “What capability, perspective or constructive tension will this person add?”
Assessment Without Validation
Psychometric and behavioural assessments can be valuable, but only when they are reliable, relevant to the role and validated for the decisions being made. The SIOP Principles for the Validation and Use of Personnel Selection Procedures make clear that selection instruments require evidence of validity, fairness and appropriate use. A polished dashboard or proprietary score does not remove that obligation.
The risk is particularly acute when AI claims to infer personality from CVs, language or online behaviour. An external audit of two algorithmic personality-prediction systems found substantial instability in their outputs, raising doubts about whether they could function as valid testing instruments. SIOP’s recommendations for AI-based assessments accordingly emphasise that new technology should face the same scientific scrutiny as established tests.
Assessment becomes malpractice when it is used because it appears sophisticated rather than because the organisation can demonstrate what it measures, how accurately it measures it and why the result is relevant to performance. The danger is not merely an inaccurate score. It is the institutional confidence that an inaccurate score can create.
Candidate Ghosting: The Governance Failure Everyone Can See
Some of the most damaging practices require no advanced technology. In Greenhouse’s 2024 survey, 61% of jobseekers reported being ghosted after an interview. Silence after an application is frustrating; silence after multiple interviews is a statement about how the organisation values other people’s time.
Candidate ghosting is often explained by workload, changing budgets or delayed decisions. Those pressures are real, but they do not justify indefinite silence. Modern recruitment platforms make basic status communication inexpensive. When an organisation cannot close the loop, the failure is rarely technological. It is a failure of ownership: nobody is clearly accountable for the candidate once the internal process stalls.
The reputational consequences extend beyond the vacancy. Candidates are also consumers, suppliers, investors and potential future employees. A company may spend heavily on employer branding while allowing its recruitment process to communicate indifference. The contradiction is difficult to hide.
Conclusion: Isolated Frustrations, Systemic Pattern
Ghost jobs, interview theatre, algorithmic bias, cultural matching, unvalidated testing and candidate ghosting are usually discussed as separate defects. They share a common architecture. In each case, the organisation transfers uncertainty or risk away from itself and towards the candidate. It gains more applications, more assessments, more data or more time, while providing less clarity, accountability or evidence in return.
That is why systemic recruitment malpractice cannot be solved by asking individual recruiters to work harder. The problem lies in mandates, incentives, validation standards and ownership. Organisations must decide whether their recruitment system exists to manage applicants or to identify future value.
The final article in this series will turn from diagnosis to governance. It will introduce the Recruitment Audit, examine recruitment debt and set out practical responsibilities for boards, executives, HR leaders, hiring managers and technology vendors. The central question will be simple: if recruitment is an investment in future organisational capability, what would it look like to govern it as seriously as one?
References
Dastin, J. (2018). Amazon scraps secret AI recruiting tool that showed bias against women. Reuters, 10 October. (Accessed 21 July 2026).
Greenhouse (2024). Ghosting, ghost jobs and bots: Candidates reveal their top concerns in the 2024 State of Job-Hunting report. 10 December. (Accessed 21 July 2026).
Google re:Work (n.d.). Use structured interviewing. Google re:Work Guides. (Accessed 21 July 2026).
New York City Department of Consumer and Worker Protection (n.d.). Automated Employment Decision Tools. (Accessed 21 July 2026).
Ng, H. (2024). Why is it so hard to find a job now? Enter Ghost Jobs. arXiv preprint arXiv:2410.21771. (Accessed 21 July 2026).
Rhea, A.K., Markey, K., D’Arinzo, L., Schellmann, H., Sloane, M., Squires, P., Kahn, F.A. and Stoyanovich, J. (2022). An External Stability Audit Framework to Test the Validity of Personality Prediction in AI Hiring. arXiv preprint arXiv:2201.09151. (Accessed 21 July 2026).
Rivera, L.A. (2012). Hiring as Cultural Matching: The Case of Elite Professional Service Firms. American Sociological Review, 77(6), pp. 999 – 1022. (Accessed 21 July 2026).
Schmidt, F.L. and Hunter, J.E. (1998). The validity and utility of selection methods in personnel psychology: Practical and theoretical implications of 85 years of research findings. Psychological Bulletin, 124(2), pp. 262–274. (Accessed 21 July 2026).
Society for Industrial and Organizational Psychology (2018). Principles for the Validation and Use of Personnel Selection Procedures. Fifth edition. Bowling Green, OH: SIOP. (Accessed 21 July 2026).
Society for Industrial and Organizational Psychology (2023). Considerations and Recommendations for the Validation and Use of AI-Based Assessments for Employee Selection. Bowling Green, OH: SIOP.
U.S. Equal Employment Opportunity Commission (2022). Artificial Intelligence and the ADA. (Accessed 21 July 2026).