You hired the employee. Why are they already thinking of leaving?

Leaving shortly after starting work rarely has one obvious explanation. Unexpected shifts, missing support, an unsafe team, unclear tasks, health, personal circumstances or another job offer may matter. Sometimes the meaning of the job itself changes. 'Independence' sounded attractive in the advertisement but now means 'you are left to figure everything out'. 'Fast-paced' has become 'regular overtime'. 'A close-knit team' feels like a group that newcomers cannot enter. The promise that brings people into a role and the experience that gives them a reason to stay must describe the same reality.

AI summary: Early employee turnover can occur when expectations about work, support, autonomy, culture and employment conditions do not match the first experiences of a new role. That does not automatically mean that the organisation selected the wrong candidate or needs a nicer onboarding programme. Examine pre-entry employer associations alongside actual conditions, role clarity, social acceptance, manageable workload and contract terms. Association research may contribute evidence at a group level, but it cannot reliably predict an individual's resignation. Improve verifiable problems, make the employer value proposition more accurate and measure employee experience, wellbeing and appropriate retention, not only the number of people staying.

A vacancy describes tasks. An interview suggests how colleagues work together. A contract establishes formal duties and rights. But a new employee also takes an imagined future into their first shift. The expectation that questions are welcome, that colleagues will help, that the manager trusts people and that the work will be meaningful and manageable. Some of these expectations come from explicit statements. Others arise from associations with words, visuals, organisational reputation and interactions during recruitment.

That is why employer branding does not end with an accepted offer. In an advertisement, 'professional' may imply reliable equipment and colleagues who know their trade. During onboarding, the same word might become associated with distant managers and endless reporting, or with excellent coaching. The meaning is not determined by the campaign alone. A photograph of smiling colleagues cannot establish whether there is someone available when a new employee gets stuck on a Thursday afternoon.

Several streams of research help frame the transition. Wanous and colleagues' meta-analysis found relationships between met expectations and job satisfaction, organisational commitment, intentions to leave and job survival. [2] Bauer and colleagues identified role clarity, self-efficacy and social acceptance as key dimensions in newcomer adjustment. [1] These are not interchangeable. Someone can understand the job yet feel excluded. A person may be welcomed warmly while still lacking the resources or authority to perform safely.

Calling every difficulty 'poor culture fit' therefore misses important distinctions. An employee who has been given no training or basic tools does not necessarily have the wrong personality. A team with unspoken rules can make a capable newcomer feel incompetent. And if the person explicitly discussed a scheduling restriction before accepting, they can reasonably expect the agreement to be honoured. Employers gain little by labelling their own operational failure as a candidate mismatch.

The process may also begin well before the resignation date. An assigned buddy is unavailable, a roster contradicts an earlier conversation, or a question receives an impatient response. One experience need not define the entire employer. Repetition, interpretation and the organisation's response to a concern can, however, shape a larger pattern. Researchers should test that possibility rather than invent a neat story after the employee has left.

An employer association is not a permanent label. It may come from the advertisement, a friend, an interview, a familiar brand, assumptions about the profession and everyday actions at work. Someone might picture an innovative company and encounter an unnecessarily complicated approval process. The interpretation 'innovation is just branding here' is plausible. It is not a conclusion researchers can draw without speaking to real employees and examining the circumstances.

Positive and negative experiences can coexist. Pay may match the agreement while the team atmosphere disappoints. Colleagues can be friendly while the supervisor is unavailable. A person can enjoy the technical work but feel unsafe admitting a mistake. A single general onboarding satisfaction score cannot separate these mechanisms, let alone explain individual choices.

A useful diagnostic distinction separates an expectation gap, a meaning gap and an actual failure to deliver. The first exists when expected work differs from what the employee encounters. The second occurs when a word such as 'flexibility' acquires a meaning opposite to what the person assumed. The third concerns something the employer is responsible for providing, such as protective equipment or agreed working hours. These issues may overlap, but their remedies differ.

Research on psychological contracts considers employees' beliefs about mutual obligations beyond the legal contract. Morrison and Robinson proposed a model for how perceived psychological contract violation can emerge. [5] It is a framework for questions, not evidence that everyone who resigns quickly feels betrayed. Investigators must establish which promise the person believed was made, who communicated it and what happened in practice.

There is also good reason to be accurate about difficult aspects of a job. Explaining demanding periods, rotating shifts or genuine learning requirements does not automatically make a role unattractive. It helps people form more realistic expectations. Phillips' meta-analysis of realistic job previews examined their relationships with expectations, recruitment outcomes and employee turnover, finding that results depend on design and context. [4] The objective is not a pessimistic advertisement. It is an informed choice that still feels credible once the work begins.

The following categories are a diagnostic framework, not eight neurological personality types, a proven ranking or a model for individual employee classification.

An association can change without becoming negative. Some newcomers discover they can cope with challenging conditions and gain confidence. Others encounter an actual contractual or safety problem that requires immediate correction. Do not describe the latter as 'unhelpful beliefs'.

Separate the role, team, manager and employer. 'I don't fit here' may refer to a work pace, a particular supervisor, team dynamics or a non-negotiable family arrangement. Treating every such case as a company-wide employer-brand crisis can distract from a specific process. The reverse is equally damaging: a systematically misleading recruitment promise may be dismissed as one isolated manager's mistake.

The literature gives us stronger starting points than one resignation anecdote. Bauer and colleagues combined 70 independent samples of newcomers. Their model places role clarity, self-efficacy and social acceptance between socialisation experiences and later outcomes. [1] A checklist may establish what the job entails without making a person feel accepted. A friendly welcome cannot compensate for lacking the authority or equipment to do the job.

Saks, Uggerslev and Fassina's meta-analysis investigated relationships between socialisation tactics and adjustment. Social elements, including learning from experienced colleagues and affirming the newcomer's identity, featured prominently in their findings. [3] That makes buddy support, predictable learning and respectful treatment worthwhile hypotheses. It does not establish that the same buddy programme will work equally well in a hospital, factory, consultancy and field-service operation.

Kristof-Brown, Zimmerman and Johnson reviewed several forms of fit: person-job, person-organisation, person-group and person-supervisor fit. [6] The phrase 'the candidate was not a fit' is therefore not an explanation unless the relevant relationship or task is specified. Fit research is also not permission to exclude people based on demographic background or subjective comfort with similarity. An organisation may be able to change the environment rather than blame the newcomer.

Allen studied how socialisation tactics, job embeddedness and turnover related in a large financial-services organisation. [7] The findings support examining social integration and attachment to the organisation, but one study in one context cannot be transferred uncritically to, say, technical workers operating alone. Ashforth, Sluss and Saks also examined socialisation and newcomer learning in a longitudinal graduate sample. [9] Together, these studies offer mechanisms to investigate, not a fixed onboarding recipe.

Hom and colleagues' overview of a century of employee turnover research reminds us that leaving is complex, with different pathways, attitudes and reasons. [8] A disappointing event in week two is not a deterministic signal of resignation in week six. The cited literature contains meta-analyses, models and studies with different designs. It does not prove that a particular intervention will reduce turnover by a specified percentage at your company.

This entire case is fictional. The organisation, people, experiences and proposed alternatives are not Neurofactor client results or measured findings.

Imagine a regional technical service company recruiting field technicians. The campaign shows colleagues solving difficult faults together, well-equipped vans and the promise of 'independence, a team behind you and opportunities to develop'. Interviews are constructive. A new hire called Sam, a fictional character, has relevant experience and deliberately chooses this employer. Sam expects autonomous jobs but accessible support with unfamiliar equipment and difficult clients.

Day one involves paperwork, access cards and brief introductions. A planned shadowing shift is cancelled because a colleague is absent. On day two, Sam receives a route with several appointments. The plan is technically possible but leaves little time for discussion. A question posted in a group chat remains unanswered while everyone is travelling. No one intended to abandon Sam. Nonetheless, the association starts to change: independence means I cannot reach anyone.

During the second week, a customer praises Sam's work. That reinforces 'I am capable'. At the same time, a late shift change is made without consultation. The supervisor calls it 'being flexible'. Sam understood flexibility as having some say over scheduling. The same word now produces an opposite meaning. Sam does not raise the concern clearly, worried about appearing difficult as a newcomer.

A few weeks later, Sam considers another role. In a hypothetical exit interview, the explanation is simply: 'It was not what I expected.' If the organisation records 'poor fit', it misses the important questions. Did the schedule contradict an agreement? Was shadowing promised? Could starters really reach someone? Were these conditions common across teams? Did managers have the authority and time to correct them?

The organisation develops three hypothetical improvement routes. A - Realistic expectations: prospective employees see an authentic workday, including busy periods, call support and actual scheduling rules. B - A workable start: early shifts include a reachable buddy, checks on competence and an unambiguous escalation contact. C - Feedback and correction: conversations after one, two and four weeks compare expectations and reality, with visible ownership of unresolved issues. There is no tested winning option here and no invented retention increase.

A sound follow-up study does not merely ask which concept sounds nicest. It asks whether candidates understand the job more accurately, whether newcomers can work safely, whether help is actually available and whether the pattern of early departures changes. Researchers must also be ready to discover that compensation, physical demands or scheduling are fundamentally unattractive. That would call for a change in working conditions, not a more persuasive campaign.

An exit interview can be useful, but it is an imperfect instrument for reconstructing an employee's first weeks. A departing person may simplify the story, avoid conflict, protect personal relationships or assign new meaning to events afterwards. Managers may hear only what fits their existing explanation. 'The culture was not right' becomes a convenient label that leaves the operational trigger untouched.

Collect different evidence at different moments instead. What did the candidate believe about the role when accepting the offer? Which promises were explicitly confirmed? What was agreed in writing and in scheduling discussions? What happened during the first shifts? When did concerns arise, and how did the organisation respond? Timing matters because a single setback is different from repeated failure to address an agreed condition.

Do not speak only to people who leave. Include newcomers who remain, established colleagues and managers. Someone who resigns may notice a problem that long-serving employees have come to normalise. Equally, someone who initially struggled but later settled in may explain which support genuinely helped. Use open questions rather than selecting answers to confirm a preferred account of why employees resign.

Power matters. A newcomer may fear that criticism will influence a probationary review or their relationship with the manager. A survey administered by the very person whose behaviour is under discussion may therefore provide incomplete or flattering answers. Participation should be voluntary, respondents should understand confidentiality limits, and sensitive interviews may need an independent researcher. The employer should never imply that honest criticism puts a contract at risk.

Exit interviews belong in a broader evidence set alongside actual work agreements, onboarding records, workload observations and carefully conducted conversations. Report group patterns only when sample sizes and privacy safeguards are adequate. In a small team, a seemingly anonymous quote can still identify the speaker.

These are possible work-related signals, not permission for covert surveillance. Monitoring private social accounts or constructing a personal resignation risk score is not a substitute for respectful conversations and genuine workplace improvements.

To make association research useful, translate each attractive phrase into observable behaviour and verifiable evidence. Otherwise the analysis gets trapped in generic concepts such as 'people-centred' or 'professional'. An association-to-reality matrix clarifies what was promised, what the newcomer understood and whether the workplace can actually deliver it.

The matrix matters only when someone can act on the findings. Recruitment may craft the message 'room to grow', but operational leaders determine whether training time exists. HR can design a buddy system, but scheduling decides whether the buddy is ever present. Closing a promise-to-experience gap therefore requires line management, operations, HR and sometimes senior leadership, not a new slogan from the employer-brand team.

People can also interpret a credible promise differently. An experienced specialist may hear autonomy as freedom from micromanagement. A novice may interpret it as the chance to practise with support nearby. These are hypotheses to explore in actual work contexts. They should never become fixed psychological labels attached to whole age or demographic categories.

Step 1 - Define the departure problem without assuming a cause. Are you studying voluntary exits in the first 30, 90 or 180 days, resignation during training, early end of a temporary contract or dissatisfaction among people who stay? Separate voluntary and employer-initiated exits and specify which employees enter the denominator. A turnover rate without a clear cohort and observation window is difficult to interpret.

Step 2 - Audit work conditions, safety and contractual basics. Compare promised and actual pay, hours, location, work responsibilities, equipment, protective measures and schedules. Examine reports of discrimination, harassment, bullying, unsafe work and unreasonable workloads. Verified breaches require organisational action and appropriate protection. They should not be reframed as an association or reputation problem.

Step 3 - Map the promise made before entry. Review vacancies, advertisements, careers pages, interview scripts, offer letters and examples recruiters and managers use. Identify which statements are specific, verifiable and applicable to the actual team. Ask candidates what they believed they heard. A video suggesting unlimited support can create an expectation even when the copy does not explicitly promise it.

Step 4 - Record expectations before the start. Ideally before day one, ask new employees what they expect from the role and what conditions are most important to them. Begin with open questions before offering predefined categories. Ask how confident they are and where each expectation came from. Do not collect unnecessary personal information or intimate details about family or health.

Step 5 - Study first experiences in context. Examine role clarity, practical guidance, real tasks, availability of tools, schedules, social welcome and manager contact. Gather evidence at relevant moments, such as the first shift and subsequent weeks, rather than relying solely on a general happiness score. Allow factual incidents to be reported without forcing employees through psychological profiling.

Step 6 - Explore shifts in association. Ask what the employee now associates with the job, team, manager and employer. Compare with expectations, but do not treat every change as caused by onboarding. Well-designed explicit association questions are often enough. A validated implicit task might add group-level evidence in a suitable research design; it must not become an individual hiring, monitoring or dismissal measure.

Step 7 - Consider competing explanations. For each gap, compare plausible causes: unrealistic recruitment messaging, insufficient training, demanding working conditions, alternative opportunities or external personal circumstances. Determine whether the issue occurs across roles, within one team or in a distinct individual situation. Studying only people who leave creates selection bias and overlooks contrasting experiences.

Step 8 - Test a relevant intervention responsibly. Inaccurate schedule information calls for clarity and compliance. An unrealistic vacancy claim calls for correction. A lack of help calls for someone genuinely available. Specify the proposed mechanism, the experience to improve, the expected outcome and potential harms before changing the process. Where possible, use sound cohort or team comparisons without withholding essential support from a control group.

Step 9 - Evaluate outcomes beyond the resignation count. Measure suitable recruitment, expectation alignment, role clarity, available support, voluntary early departures and longer-term work outcomes such as safe performance and retention. Account for differences in job, location, manager, contract type and labour market. Report counts, time windows, missing responses and uncertainty. A fall in departures is not a success if people's working conditions deteriorate.

Not every step requires a neurological measurement or a large formal programme. When the immediate problem is an inaccessible work system, missing protective gear or a schedule that violates an agreement, correcting that condition matters more than investing in a sophisticated association study.

1. Describe real work, not just corporate values. Show what employees do frequently, which periods are demanding, what they do independently and what assistance is available. Avoid relying exclusively on words that could describe any employer. Mention difficult conditions that legitimately affect the applicant's decision.

2. Offer realistic exposure to the role. A conversation with a future colleague, a relevant task example or a properly arranged visit can give candidates a better basis for judgment. Make conditions transparent, accessible and fair. A realistic preview must not become an undeclared selection test or unpaid productive labour.

3. Explain what attractive terms actually mean. 'Flexible hours' needs details about availability, shift swaps, notice periods and choice. 'High responsibility' needs clear decision authority and support. Otherwise candidates may fill the gaps with interpretations that the employer never intends or can deliver.

4. Reserve real time for guidance. A buddy assigned in a spreadsheet is not useful if they are always on another shift. Protect support time, define what the buddy can answer and make escalation straightforward. Research supports investigating socialisation and guidance, but the best arrangement depends on the actual job. [3]

5. Give newcomers a realistic first success. This is not about 'hacking motivation'. It is about making tasks, competence boundaries, feedback and success criteria understandable. Increase complexity in step with experience and safety. A person who knows what good work looks like can judge their own progress more accurately.

6. Ensure managers understand recruitment promises. The employer-brand campaign should not tell a story that frontline managers have never seen. Discuss which conditions are explicitly or implicitly being offered. Amend claims that departments cannot fulfil. Let managers report operational limitations without treating the report as disloyalty.

7. Ask for specific experiences, not only feelings. 'Do you feel welcome?' can begin a conversation, but follow with 'Who could you ask for help yesterday?', 'When was reality different from what you expected?' and 'What happened after you raised it?' Specific examples make it easier to distinguish a general impression from an actionable cause.

8. Show that speaking up leads to action. When employees report missing equipment, unsafe work or an impossible schedule, someone must own the correction. Follow up to confirm that it happened. Asking questions without acting can reinforce the belief that concerns are unwelcome or futile.

9. Reject artificial loyalty pressure. People are not required to report positive feelings to be valued employees. Do not portray resignation or constructive criticism as a moral failure. Respect autonomy, privacy and the right to decide that a role does not suit one's needs. An honest departure can be preferable to prolonged mutual frustration.

10. Update the EVP only as reality improves. A credible employer value proposition describes what employees can expect now, not a future aspiration presented as a current fact. When actual improvements have been demonstrated, the messaging can become more precise. If a promise is true in one department but not another, avoid representing it as universal.

These decisions are not a guaranteed retention checklist. The conditions that matter for seasonal workers, clinical shift teams, technical field staff and consultants can be quite different. The right response is driven by the diagnosed cause, not the popularity of an onboarding trend.

The starting point for Neurofactor is the behavioural question: which expectations arrive with a newcomer, how are those meanings reinforced or changed, and which differences warrant action? Its research methods may combine association research, profiling, behavioural observations and business data. A method is not an outcome, however, and a sophisticated sensor does not explain why someone resigns.

Explicit association research can explore the words and meanings people spontaneously connect with an employer, role or team: 'space', 'pressure', 'safety', 'pride', 'learning', 'alone'. Be precise about the target of each association. 'I trust the organisation' and 'I trust my immediate supervisor' are different claims. Ask which experiences made those meanings credible. A list of positive adjectives, disconnected from conditions, will rarely support a useful decision.

Implicit association tasks can, under suitable conditions, examine relative links between concepts. Their relevance depends on task design, validity, reliability and enough participants for the intended analysis. Reaction times do not reveal an individual's hidden wish to resign, dissatisfaction or mental condition. Never use scores to rank job applicants, judge an employee's loyalty or exploit individual vulnerabilities.

Profiling may describe patterns across roles, teams and contexts when the sample and evidence allow it. Resist stereotypes: younger people do not automatically want more freedom, and experienced staff do not necessarily prefer predictability. Apparent differences can reflect actual tasks, working hours, commuting demands and access to support. Examine the environment before assigning a trait to a group.

Eye tracking may help determine whether candidates notice important information on a vacancy page or onboarding interface. It does not measure belonging. EEG, in suitable controlled studies, can describe time-related brain responses, but it is no detector of 'this employee will leave' or 'this employee believes the EVP'. For questions about daily team conduct, task observation, conversations, scheduling records and subsequent outcomes are usually more direct sources of evidence.

A useful research chain is: employer promise -> interpretation -> lived conditions -> possible association shift -> testable change -> outcome. This is a working model, not a universal law. If an employee likes the job but cannot combine the schedule with personal responsibilities, the report should state that plainly.

Sometimes the job genuinely does not fit a person's circumstances. Night work proves physically unmanageable. A changed shift pattern makes commuting impractical. Another employer offers more attractive terms. Unexpected caring responsibilities arise. The role requires qualifications that were never mentioned. None of these situations can necessarily be fixed by better communication.

Other issues must not be psychologised. Harassment, discrimination, unsafe working practices and failure to honour employment agreements are not negative associations to be erased with a cheerful induction campaign. The underlying conditions require correction, protection and appropriate procedures. HR, independent advisers, employee representatives, occupational-health professionals or other competent parties may need to act within their proper roles and with confidentiality safeguards.

Conversely, an employer can communicate honestly, guide newcomers well and still have some people leave quickly. Personal preferences and circumstances change. A decision to move on does not automatically mean that either side has failed. The research goal should not be zero voluntary turnover, but fewer preventable harmful mismatches and better-informed, sustainable employment experiences.

Be careful with causal inference. If newcomers with buddies stay longer, that does not by itself prove the buddy caused retention. Perhaps the most organised teams also happened to assign buddies, or more difficult roles received less structured support. Before-and-after comparisons may be affected by seasonality, labour-market conditions, changing pay or changes in the people hired. Responsible reporting acknowledges these alternative explanations.

An attractive dashboard is not always an informative one. Recruitment can deliver more applications while new hires leave faster. Onboarding can receive higher satisfaction ratings without improvements in safe work or performance. Recruitment, work experience and retention measures must be considered together rather than collapsed into one appealing but ambiguous score.

Choose a primary outcome that matches the actual question. If it concerns avoidable voluntary turnover within 90 days, define the cohort, observation period and departure categories. Separate those leaving during probation, training and independent work. Do not compare cohorts whose members have not yet had a complete 90-day opportunity to leave without using appropriate time-to-event methods.

Add relevant intermediate outcomes. Can employees describe their tasks and priorities? Do they know when and how to ask for assistance? Is their point of contact reachable? Are scheduling, equipment and safety arrangements functioning? Do their expectations of autonomy and development match practice? These measures can help test a mechanism, but they do not replace a valid causal comparison.

Monitor unintended consequences. Low turnover is not a success if employees remain mainly because alternative jobs are unavailable or complaints feel dangerous. Consider undesirable absence, workload, grievances, incidents and sustainable employability. Good retention means people can work safely, understand their situation and make meaningful choices, not that leaving has been made harder.

Where justified, compare jobs, worksites, recruitment cohorts and onboarding approaches. Protect employees in small teams from re-identification. Report actual counts, uncertainty, non-response and the limitations of satisfaction and association measures. For very small groups, careful action on individual operational problems is often more valuable than statistically impressive-looking percentages.

Continue evaluating after the first weeks. Someone may feel relieved after surviving a confusing induction yet discover months later that development has stalled. Another person may have a difficult start and gradually develop trust and competence. Employee learning and the credibility of the employer promise unfold over time, not on a single survey date.

Why do employees leave shortly after joining? Sometimes the actual role, environment or conditions differ from what they expected. Sometimes role clarity, guidance and social acceptance are insufficient. And sometimes working conditions, safety, personal circumstances or an alternative offer explain the decision. The resignation alone cannot identify which account is correct.

The next step is therefore not automatically another vacancy rewrite or a new onboarding app. Investigate what the person associated with the employer before joining, which lived experiences confirmed or changed those meanings, and what the organisation can genuinely improve. Both employer and employee should be able to make an informed decision.

A strong EVP does not merely convince someone to start. It remains credible when that person sees how the work is actually done on an ordinary Tuesday morning. That is the difference between attractive recruitment communications and an employer people have sound reasons to stay with.

Limits of the evidence: The studies offer theoretical models, meta-analyses and empirical associations involving expectations, socialisation, fit and turnover. They are not Neurofactor client results, do not prove that a specific campaign causes resignations and do not promise any fixed retention lift from onboarding interventions.

• Neurofactor research methods - choosing association, profiling or behavioural research around the actual question.
• Neurofactor applications - recruitment, employer branding, workplace and behavioural strategy.
• Associations in recruitment: what makes talent choose an employer
• Why does a well-known employer brand still attract so few applications?