A facilities management company loses its best contract manager, Joanna, after nine years. In her resignation meeting she says she has had “an opportunity too good to turn down”. Her director is puzzled. Her engagement survey scores were high. She had a pay rise in April. There had been no complaint.
What he does not know is the sequence. In June, a request for two days’ leave during her father’s surgery was approved only after three reminders and a remark about “busy periods”. She said nothing. That evening she replied to a recruiter’s message she had ignored for a year. The new job pays 4% more, which is not why she is taking it. Asked by a friend, she says: “Nothing was wrong, exactly. I just suddenly saw how it would be if something really went wrong.” The company’s retention analysis will record her reason as “career opportunity”. The actual cause was one conversation about leave that her director has forgotten.
The usual account of why people leave jobs is gradual. Dissatisfaction builds, the employee begins to look, finds something better and goes. On this account, satisfaction surveys should predict resignations and pay rises should prevent them.
They do, a little. But decades of turnover research show that the gradual account describes only some departures and leaves much unexplained. This guide sets out what the research says about how people actually decide to leave, and what that means for anyone trying to keep them. It is part of the HR Psychology hub.
Shocks start most departures.
A specific event prompts people to reconsider. Slow dissatisfaction alone is the less common route.
Embeddedness holds people.
Links, fit and what would be sacrificed keep people in place and cushion shocks.
Prevent, notice, repair.
Handle sensitive moments carefully, ask early and put mistakes right quickly.
How well does job dissatisfaction predict quitting?
Less well than is generally assumed.
A meta-analysis of turnover research covering hundreds of studies found that job satisfaction and organisational commitment were related to actual turnover, but only modestly. Pay level and pay satisfaction were weaker predictors still. The best single predictor was a stated intention to quit, and even that was far from decisive.
Many dissatisfied people stay for years, and many apparently contented people leave. Something beyond accumulated attitude is at work. Two ideas developed by Thomas Lee, Terence Mitchell and colleagues supply much of the answer: the unfolding model, which explains how people come to leave, and job embeddedness, which explains why people stay.
What is the unfolding model of turnover?
Lee and Mitchell proposed in 1994 that leaving does not follow a single process. Their central observation was that many departures begin with a shock: a specific, jarring event that prompts a person to think about their job in a way they had not been doing.
A shock can be negative, positive or neutral, and it need not come from work.
- Work shocks: being passed over, a new manager, a reorganisation, a colleague’s dismissal, a broken promise, a humiliating meeting.
- Personal shocks: a partner’s relocation, a birth, a bereavement, a significant birthday, a health scare.
- External shocks: an unsolicited job offer, a friend’s success elsewhere.
The model describes four main paths.
| Path | Trigger | What happens | Search for alternatives? |
|---|---|---|---|
| 1 | A shock that matches a pre-existing plan (“if my partner gets the transfer, I’ll go”) | The plan is enacted, with little deliberation | No |
| 2 | A shock that violates the person’s values or goals | They reassess whether they belong here and leave | No, they leave without another job |
| 3 | A shock, often an unexpected offer or a negative event | They compare the current job with specific alternatives | Yes |
| 4 | No shock | Dissatisfaction accumulates until they leave or search | Sometimes |
In a study of accountants who had left their firms, the great majority of departures could be classified into the model’s paths, and most involved a shock. Paths begun by a shock were considerably more common than leaving through slowly accumulated dissatisfaction alone.
The fourth path is the traditional account. It is real, and it is a minority. Joanna’s departure followed the third: a negative event made her receptive to an alternative she had been ignoring.
Why does a single event carry so much weight?
Because it changes what the person pays attention to. Most people, most of the time, are not evaluating their job. They are doing it. Inertia is powerful, and the status quo is the default.
A shock breaks the routine and prompts a question: “Do I still want to be here?” Once the question has been asked, the person begins to notice things they had stopped seeing and to interpret new events in its light. Recruiters’ messages get answered.
Shocks that carry a message about the relationship are especially powerful. Joanna did not leave because of two days’ leave. She left because of what the episode told her about how the company would treat her when it mattered. One event was taken as a sample of the whole relationship. This is the mechanism of the psychological contract, examined in our guide to the unwritten deal.
People judge an experience largely by its most intense moment and its most recent one, not by an average of the whole. Nine good years can be outweighed by one vivid incident near the end.
What is job embeddedness?
It is an answer to the opposite question: why do people stay? Mitchell and colleagues proposed that people are held in a job by a web of connections, some of which have nothing to do with how satisfied they are.
| Dimension | In the organisation | In the community |
|---|---|---|
| Links | Colleagues, teams, mentors, friends at work | Family nearby, children’s schools, friends, clubs |
| Fit | Skills match the job; values match the culture | Likes the area, the climate, the way of life |
| Sacrifice | What would be lost: pension, seniority, interesting projects, flexibility | What would be lost: home, a partner’s job, a short commute |
In studies of grocery store and hospital employees, job embeddedness predicted both intention to leave and actual turnover beyond what was predicted by job satisfaction, organisational commitment, perceived alternatives and job search.
Embeddedness explains the dissatisfied stayers. Someone with close friends at work, a specialised skill, children settled in a local school and a pension that vests in two years may be thoroughly fed up and still not leave. It explains contented leavers as well: a satisfied employee with few ties is easily moved by a shock.
Embeddedness also cushions shocks. A well-embedded person who is passed over for promotion has more reasons to absorb the disappointment. A weakly embedded one has fewer.
Do people really leave managers, not companies?
The saying is too simple and contains something true. The immediate manager controls most of what an employee experiences day to day: workload, recognition, flexibility, development, whether a problem is heard. Managers are also the source of many shocks, and the people best placed to prevent or repair them. Joanna’s shock was delivered by a manager.
But people also leave good managers, for pay, for careers, for family reasons, because the company is failing. And they stay with poor managers when embeddedness is high. A fairer statement is that the manager is the most important single factor the organisation can influence in the short term.
Is quitting contagious?
Yes.
Across samples in hospitality and banking, employees were more likely to quit when their co-workers were less embedded and were engaged in job search. Colleagues’ search behaviour predicted an individual’s leaving over and above that individual’s own satisfaction and embeddedness.
There are several reasons. A colleague’s departure is a shock in its own right. It removes a link, reducing embeddedness. It supplies information: someone you respect has found something better. And it makes leaving seem normal. Departures therefore cluster. The period after a respected person leaves is high-risk for those who were close to them, and it is the same dynamic that follows redundancies, described in our guide to layoff survivors.
What are the warning signs?
Research on pre-quitting behaviour by Timothy Gardner and Peter Hom suggests that the obvious signs, such as dressing smartly for an interview or taking unexplained leave, are not the reliable ones. More reliable are signs of disengagement.
- Less effort and less initiative than usual.
- Reluctance to commit to long-term projects.
- Less interest in pleasing the manager.
- Becoming quieter in meetings; fewer suggestions.
- Reduced interest in training or advancement.
- Leaving promptly and doing the minimum.
These overlap with what is now popularly called quiet quitting, covered in our plain-language explainer. The common feature is withdrawal of discretionary effort. By the time such behaviour is visible, the question has usually already been asked.
What can an employer actually do?
The research suggests four lines of action, each addressing a different part of the process.
Prevent avoidable shocks. Many are self-inflicted: a clumsy announcement, a promotion handled badly, a refused request, a broken commitment. The guides on being passed over, declining a raise request and fair process are, in effect, about shock prevention.
Detect them. Managers who hold regular, real one-to-ones tend to know when something has happened. The simplest tool is the stay interview: a periodic conversation in which the manager asks what keeps the person here and what might cause them to leave.
Repair. A shock need not end in resignation. Had Joanna’s director noticed and said, “I handled your leave request badly, and I’m sorry. It should have been approved the same day”, the story might well have ended there. Repair has to be quick, because the reassessment begins at once.
Build embeddedness. Strengthen links through teams, mentoring and joint projects. Improve fit through role design and development. Increase what would be sacrificed by leaving: interesting work, flexibility, benefits that grow with service, a clear route ahead.
Are counteroffers and retention bonuses worth it?
Sometimes, but they address a late stage of a process that began earlier. By the time someone resigns, they have experienced a shock, reassessed, searched, compared and decided, and they have begun to detach emotionally. A counteroffer speaks to pay. It seldom speaks to whatever started the process. The psychology is covered in salary negotiation psychology for managers.
Retention bonuses increase the sacrifice of leaving and can keep people through a defined period. They buy time and not commitment. An employee held by a bonus while the original cause goes unaddressed tends to leave when it vests.
Do not treat every departure as preventable or every leaver as a failure. Some turnover is healthy, and some shocks, such as a partner’s relocation, are beyond an employer’s influence. The aim is to avoid losing people you wanted to keep for reasons you could have dealt with.
What should you learn when someone goes?
The real reason, which is harder to obtain than it sounds. Leavers give safe, socially acceptable explanations, such as “a better opportunity”, that hide the trigger. Asking “what first made you think about leaving?” and “was there a particular moment?” comes closer. The limits of the method are discussed in our guide to exit interview psychology.
Look for patterns in shocks across leavers. If several departures trace back to the same kind of event, such as a reorganisation, a particular manager or the way leave is handled, that is where the leverage lies.
After any event likely to act as a shock for your team, whether a restructuring, a colleague’s exit, a rejected promotion or a difficult review, schedule one-to-ones within the week and ask directly how people are taking it. The conversation costs half an hour. It is the cheapest retention measure there is, and the one most often skipped.
Joanna’s director eventually learned the real story from a mutual acquaintance. He changed one thing. Leave requests for family medical reasons are now approved the same day by default and reviewed afterwards. It is a small rule, and it removes one of the ways a manager could tell someone, without meaning to, what they would be worth to the company on their worst day.
Frequently asked questions
What is the main reason employees quit?
There is no single reason. Research shows many resignations are triggered by a specific event, or shock, that leads the person to reassess their job, such as being passed over, a change of manager, a broken promise or an unexpected offer. Pay is usually one factor among several.
What is the unfolding model of turnover?
It is a theory developed by Thomas Lee and Terence Mitchell proposing that people leave jobs through several distinct paths, most of which begin with a shock that prompts them to reconsider, as opposed to a single process of growing dissatisfaction.
What is job embeddedness?
Job embeddedness is the set of forces that keep a person in a job: their links to people at work and in the community, how well the job and location fit their life, and what they would sacrifice by leaving. It predicts staying beyond job satisfaction.
How can managers reduce employee turnover?
By handling potentially shocking events with care, holding regular one-to-ones and stay interviews to detect problems early, repairing mistakes quickly and strengthening employees’ connections, fit and stake in the organisation.
Discover more from Kurums | Business Intelligence
Subscribe to get the latest posts sent to your email.