She handed in her resignation on a Thursday. Her manager was blindsided. HR scheduled the exit interview. When asked why she was leaving, she said "a better opportunity came along."
That's what the exit interview captured. That's what went into the report. That's what leadership believed.
Here's what actually happened.
Sixty-seven days earlier, on a Tuesday afternoon, she sat in a meeting where her idea was dismissed for the third time in two months. Her manager didn't notice. That evening, she updated her LinkedIn profile. By Friday, she'd responded to a recruiter. Within three weeks, she was interviewing. By day fifty, she had an offer. She spent the final seventeen days going through the motions, waiting for the right moment to resign.
Her manager thought everything was fine. She seemed engaged. She hit her deadlines. She didn't complain.
The decision to leave happened on day one. The resignation happened on day sixty-seven. Everything in between was theater.
This is the 67-day window. And most companies have no idea it exists.
The 67-Day Window: When Employees Actually Decide to Leave
The 67-day window is the average period between when an employee mentally decides to leave and when they formally resign. During this window, the employee appears present but has already departed psychologically. They're interviewing, negotiating, and planning their exit while their manager assumes everything is normal.
The number comes from analyzing patterns across thousands of voluntary departures. Some employees move faster. Some take longer. But the median sits at approximately 67 days, and the pattern is remarkably consistent.
The 67-Day Timeline
The entire process happened in plain sight. Nobody was watching.
Why Exit Interviews Don't Tell You Why Employees Leave
Exit interviews are theater. Both parties know it.
The departing employee has nothing to gain from honesty and something to lose. They need a reference. They might return someday. They have colleagues still at the company. So they say something safe: "better opportunity," "career growth," "compensation," "relocation."
HR records the answer. Leadership sees aggregated data showing that people leave for "better opportunities." Everyone nods. Nothing changes.
The actual reasons employees leave are almost never what they say in exit interviews.
Research from the Work Institute consistently shows that the top drivers of voluntary turnover are career development, work-life balance, manager behavior, and job characteristics. Compensation rarely ranks in the top three. Yet exit interview data at most companies shows compensation and "better opportunity" as the leading causes.
The gap between exit interview data and retention research reveals the problem: employees tell you what's safe, not what's true.
If you're using exit interview data to diagnose why employees leave, you're solving the wrong problem. The data is a record of what people were willing to say on their way out the door. It has almost no relationship to why they actually started looking.
The Real Reasons Employees Leave (And Why Managers Miss Them)
The decision to leave is almost never about one thing. It's about an accumulation of moments that shift how an employee sees their relationship with the company.
They stopped feeling heard.
Employees don't expect every idea to be implemented. They expect ideas to be considered. The moment an employee realizes their input doesn't matter, something changes. They stop contributing in meetings. Managers interpret this as "buying in" or "being a team player." It's actually the first stage of disengagement.
A software engineer told us about the meeting that started his 67-day window:
"I raised a concern about the architecture. My manager said 'let's take that offline' and then never followed up. That had happened before. But this time I realized it was always going to happen. I started looking that week."
They stopped seeing a future.
Employees don't leave companies. They leave ceilings. When someone looks at their manager's job and thinks "I don't want that" or "I'll never get that," the countdown begins.
The problem is that managers rarely have explicit career conversations. They assume employees will ask if they want to talk about growth. Employees assume if there were opportunities, someone would mention them. Both sides wait. The employee eventually concludes there's no path forward and starts looking for one somewhere else.
They stopped trusting leadership.
Trust erodes through small inconsistencies, not dramatic betrayals. A commitment made and not kept. A value stated and not demonstrated. A message that doesn't match observed reality.
A marketing director described her trigger moment:
"The CEO said in an all-hands that our team was 'critical to the company's future.' Two weeks later, our budget got cut by 30%. I didn't expect unlimited resources. But I expected the words to mean something. Once I realized they didn't, I couldn't unhear it."
They stopped being challenged.
High performers don't leave because work is too hard. They leave because work is too easy. When someone masters their role and sees no new problems to solve, they start solving the problem of finding a new job.
This is counterintuitive for managers who assume that employees want less stress. Top performers want the right kind of stress: meaningful challenges with adequate support. Remove the challenge and you remove the engagement.
They started being managed differently.
Sometimes the trigger is a manager change. The new manager has a different style, different expectations, different communication patterns. The employee who thrived under the old system struggles under the new one.
Sometimes the trigger is the same manager behaving differently. A performance review that felt unfair. A project assignment that felt like a demotion. A one-on-one that felt like an interrogation instead of a conversation.
Employees are acutely sensitive to changes in how they're treated. They notice before managers realize anything has changed.
The Behavior Misread: Why Managers Are Always Surprised
The most dangerous phrase in retention is "they seemed fine."
Managers are blindsided by resignations because they're reading employee behavior backwards. The signals that look like engagement often indicate disengagement. The behaviors that suggest someone has "settled in" actually mean they've checked out.
The employee who stops pushing back.
Before: They challenged decisions, asked hard questions, advocated for their ideas.
After: They agree with everything. Meetings are smoother.
Manager interpretation: "They've finally bought in. They're being a team player."
Reality: They've stopped caring. Agreement is easier than advocacy when you're planning to leave.
The employee who stops complaining.
Before: They voiced frustrations, flagged problems, expressed concerns.
After: They're pleasant and positive. No more complaints.
Manager interpretation: "They've adjusted. Things must be better."
Reality: They've given up on things getting better. Complaining requires believing change is possible.
The employee who becomes more self-sufficient.
Before: They asked questions, sought feedback, requested input.
After: They handle everything independently. They don't need much from their manager.
Manager interpretation: "They've really grown. They're operating at a senior level."
Reality: They've stopped investing in the relationship. Independence is preparation for departure.
The employee whose work improves.
Before: Solid performance with occasional issues.
After: Consistently strong output. Deadlines met. Quality high.
Manager interpretation: "They've hit their stride. Everything is clicking."
Reality: They're auditioning for their next job. The performance improvement is for their resume, not for you.
The employee who takes more PTO.
Before: Rarely used vacation time.
After: Using PTO more frequently, often on short notice.
Manager interpretation: "Good for them. Work-life balance is important."
Reality: Those "vacation days" are interview days.
The pattern is consistent: disengagement behaviors often look like positive developments if you're not watching carefully. By the time the resignation arrives, the manager has months of evidence that everything was fine. The evidence was lying.
The Warning Signs That Actually Matter
If exit interviews don't reveal why people leave and observable behavior often misleads, how do you know if employees are in their 67-day window?
You watch for changes, not states.
An employee who has always been quiet isn't a concern. An employee who was vocal and became quiet is a signal. An employee who has always used all their PTO isn't a flag. An employee who never used PTO and suddenly starts is worth a conversation.
Engagement pattern changes:
Meeting participation drops. Not absence, but contribution. They're present but not adding.
Slack or email response times lengthen. They're still responsive to external stakeholders but slower with internal requests.
Voluntary project participation disappears. They used to raise their hand. Now they don't.
Office presence patterns shift. This applies to hybrid environments: someone who chose to come in three days now comes in one.
Relationship pattern changes:
One-on-ones become transactional. Updates are delivered, questions are answered, but the conversation doesn't go deeper.
Cross-team collaboration decreases. They're less interested in relationships that won't follow them to their next role.
Mentoring or training involvement drops. Investing in others requires believing in the organization's future.
Career pattern changes:
They stop asking about opportunities, promotions, or growth paths. The questions disappear because they've answered them by looking elsewhere.
LinkedIn activity increases. Profile updates, new connections, engagement with industry content.
Conference or training requests stop. Professional development investment shifts to efforts that benefit their next role, not their current one.
Workload pattern changes:
They stop taking on extra responsibilities. Discretionary effort contracts.
Documentation suddenly improves. They're preparing for their absence.
Knowledge sharing increases. They're transferring what they know, often unconsciously, before they leave.
None of these signals alone confirms someone is leaving. But clusters of changes, especially in employees who previously showed opposite patterns, warrant attention.
The 67 Days You Can't Get Back
Here's the cost math that most companies never calculate.
An employee decides to leave on day one. They resign on day sixty-seven. During those sixty-seven days, the company had opportunities to intervene. Every single one was missed.
The moments that could have changed the trajectory:
What actually happened:
Day 3: Nothing.
Day 14: The one-on-one covered project updates.
Day 30: No skip-level meetings were scheduled.
Day 45: Nobody knew anything was wrong.
Day 67: The resignation arrived. The manager was shocked. HR scheduled the exit interview. The employee said "better opportunity." Leadership wondered why turnover was increasing.
The cost of losing that employee: somewhere between $75,000 and $225,000 depending on role and seniority, using SHRM replacement cost methodology. Plus the institutional knowledge. Plus the client relationships. Plus the impact on the team that watched another colleague leave.
The cost of the interventions that could have prevented it: a few hours of intentional conversation, spread across sixty-seven days.
The ROI calculation isn't complicated. Companies just don't do it until after the resignation, when it's too late to matter.
The CLOVER Approach to the 67-Day Window
The CLOVER Framework provides a structure for the daily manager behaviors that prevent employees from entering the 67-day window, and that can pull them back if they've already entered it.
The framework was developed by Clover ERA based on patterns observed across successful retention interventions. It translates retention research into specific, repeatable actions.
Communication
The 67-day window opens in silence. Employees don't announce that something has shifted. They wait to see if anyone notices. When nobody notices, they conclude nobody cares.
Communication closes the window by making employees feel seen before they feel invisible.
"We're paying attention. You're not invisible here."
Learning
Employees who feel stagnant start looking for growth elsewhere. Learning signals that the company is investing in their future, not just extracting their present.
"Your growth matters to us. We want you to be better, not just productive."
Opportunity
The 67-day window opens when employees stop seeing a future. Opportunity closes it by making the path forward explicit.
"There's a future for you here. Let us show you what it looks like."
Vulnerability
Employees are more likely to share concerns with leaders who share their own. Vulnerability creates psychological safety for honesty.
"You can tell us the truth. We can handle it."
Enablement
The 67-day window opens faster when employees feel blocked. Enablement removes friction before frustration accumulates.
"We want you to succeed, and we'll remove what's in your way."
Reflection
You can't spot the 67-day window if you're not looking. Reflection builds in the regular check-ins that surface problems before they become departures.
"We're checking in because you matter, not because we're checking up."
For Managers: The Three Questions That Spot the Window
You can't have a career conversation with every employee every week. But you can ask three questions regularly that surface whether someone has entered the 67-day window.
These questions take two minutes. They surface sixty-seven days of problems before those problems walk out the door.
For Executives: The Dashboard That Predicts Turnover
Individual managers can watch for signals in their direct reports. Executives need aggregated indicators that surface risk across the organization.
Leading indicators (predict turnover):
Employee engagement pulse scores: trend lines matter more than absolute numbers. A team whose score dropped from 78 to 72 is at higher risk than a team that's held steady at 68.
Internal application rates: employees who believe in the company's future apply for internal roles. Declining internal mobility is a canary in the coal mine.
One-on-one completion rates: managers who skip check-ins create information vacuums where the 67-day window opens unseen.
Time-to-fill for recent departures: if departing employees' roles are taking longer to fill, word is spreading that something is wrong.
Lagging indicators (confirm the problem):
Voluntary turnover rate by manager: variation matters. If one manager's team has 25% turnover while peers are at 10%, the problem isn't the market.
Turnover by tenure band: high turnover in the 1-2 year band suggests broken career paths. High turnover in the 3-5 year band suggests ceiling effects.
Exit interview response rates: declining participation means employees don't believe their feedback matters. They're leaving without bothering to explain.
Glassdoor review frequency and sentiment: external signals often appear before internal metrics move.
Build a dashboard that tracks these indicators monthly. Review it with HR and department heads. The goal isn't perfect prediction. The goal is spotting problems in aggregate before they become a crisis.
The Cost of Missing the Window
Every voluntary departure that could have been prevented represents a failure to notice.
The direct costs are measurable: recruiting expenses, onboarding time, productivity ramp-up, potential signing bonus or salary premium to attract a replacement.
SHRM estimates these direct costs at 50-200% of annual salary depending on role seniority. For a mid-level professional making $100,000, that's $50,000-$200,000 per departure. For a company losing ten preventable employees per year, that's $500,000-$2,000,000 in direct costs alone.
The indirect costs are harder to measure but often larger: institutional knowledge that walks out the door, client relationships that weaken, team morale that dips as colleagues watch people leave, and the management time consumed by constant recruiting cycles instead of strategic work.
Our turnover cost calculator models both direct and indirect costs based on your specific salary bands, time-to-fill averages, and productivity assumptions. Most executives are surprised by the total, not because the math is complicated but because nobody had added it up before.
The 67-day window is where these costs are created. It's also where they can be prevented.
The Clover ERA Platform captures these signals before it's too late.
Frequently Asked Questions
The 67-day window is the average period between when an employee mentally decides to leave and when they formally resign. During this window, the employee appears present but has already disengaged psychologically. They're interviewing, negotiating, and planning their exit while their manager often assumes everything is fine.
The figure comes from pattern analysis across thousands of voluntary departures, identifying the typical timeline from trigger event to resignation. Individual cases vary, some faster, some longer, but the median clusters around 67 days. The number represents an average, not a rule.
Employees often don't surface dissatisfaction because they've learned it doesn't lead to change, they fear career consequences, they don't want confrontation, or they've already concluded the situation is unfixable. By the time an employee enters the 67-day window, they've often tried indirect signals that went unnoticed.
Sometimes. The earlier you intervene, the better the odds. Employees in the first few weeks of the window are still evaluating. Employees at day fifty with an offer in hand are much harder to retain. The key is catching the signals early and having an honest conversation about what would need to change.
Research consistently shows that employees who accept counteroffers leave within twelve months at significantly higher rates than employees who never received one. By the time a counteroffer happens, trust is usually too damaged to repair. The better investment is preventing employees from reaching the point where they're entertaining outside offers.
Watch for changes in behavior patterns: increased PTO usage, decreased meeting participation, reduced discretionary effort, increased LinkedIn activity, and shifts in communication style. No single signal is definitive, but clusters of changes warrant a direct conversation.
Departing employees have little incentive to be honest and potential downside from candor. They need references, may have colleagues who remain, and might return someday. Safe answers like "better opportunity" or "career growth" protect them. These answers are true in a narrow sense but don't explain why they started looking in the first place.
Manager behavior is the most common trigger, not in dramatic forms like harassment but in accumulated moments: feeling unheard, seeing limited growth, experiencing unfair treatment, or losing trust in leadership. Compensation is rarely the trigger, though it's often the excuse given at exit.
Restructures often trigger the 67-day window for surviving employees. The announcement creates exactly the kind of moment that causes employees to reevaluate their relationship with the company. This is why turnover after restructure spikes in the months following layoffs: the restructure triggers hundreds of 67-day windows simultaneously.
The Question Worth Asking
Here's the question that reframes retention:
How many of my team are somewhere in their 67-day window right now?
At any given time, in any given company, some percentage of employees have decided to leave but haven't told anyone yet. They're in meetings, hitting deadlines, appearing present. They're also interviewing, networking, and waiting for the right moment to resign.
You can't know the exact number. But you can know whether you're doing anything to find out.
When did you last have a genuine career conversation with each of your direct reports? Not a performance review. Not a project check-in. A conversation about their future.
When did you last ask "how are you doing?" and wait long enough for a real answer?
When did you last notice a change in someone's behavior and follow up to understand why?
The 67-day window is always open for someone. The question is whether you're watching.
See Who Might Be in Their Window
We help companies identify turnover risk before resignations arrive. In a 15-minute Turnover Analysis call, we'll show you:
The typical 67-day window patterns for your industry and company size
Which leading indicators are most predictive for your workforce
What a manager-level early warning system looks like in practice
No pitch. No pressure. Just a conversation about what you're not seeing.
Schedule Your Free Turnover Analysis