The clearest signs an employee is about to quit show up as changes from that person's own baseline, not as a single bad day. The risk climbs when weaker output meets quieter team participation. It climbs further when someone starts cancelling 1:1s, reshuffles their schedule or pulls away from anything that runs past next month.

Managers usually catch these signals too late, and the reason is simple: the evidence sits in different places. Performance numbers, calendars, engagement comments, peer feedback, and the small everyday behaviors no dashboard tracks. The practical move is to read the pattern across the last few weeks. Then open a stay conversation about workload, growth and manager support, before the employee feels accused of anything.

What signs show an employee may quit?

Read the clearest warning signs as a cluster of pre-quitting behaviors rather than one isolated red flag. The most reliable cues are visible drops in output and effort. After that come shifts in teamwork, attitude, schedule habits, customer energy, the manager relationship and the willingness to commit to future work.

People who are leaving tend to pull back from the parts of the job that ask for more than the contract demands. You might notice productivity slipping first: fewer resolved tickets, slower cycle times, even though the work technically still meets the brief. What disappears is the polish, the proactive fix, the extra quality check. Discretionary effort fades when someone stops volunteering and stops helping outside their strict role. And team behavior changes when peer assists thin out and the optional meetings vanish from their calendar.

This is not just anecdote. A 13-month follow-up study from Gardner, Van Iddekinge and Hom found that pre-quitting behaviors correlated with voluntary turnover at r = .31 and predicted who left beyond age, tenure and even managers' own hunches. A one-unit rise on their pre-quitting scale was linked to 6.48 times higher odds of someone actually leaving.

The 12 observable pre-quitting behaviors

The second tier of signals is more relational than productivity-based. The employee may sidestep roadmap ownership, decline to mentor, or dodge projects that stretch past the next few weeks. Watch for a newly cynical tone and for meetings where their focus runs thinner than usual. Repeated complaints about growth matter, and so do comments about fairness or recognition, especially when the same person also mentions their manager is hard to reach.

What changed compared with their baseline

Schedule changes only carry weight once leaving early starts to repeat rather than happen once. Short-notice absence or clustered PTO tells you more when it lands alongside other signals instead of standing alone. One final cue deserves real attention: fading energy for customers, stakeholders or the company mission, especially in someone who used to tie their daily work to that bigger purpose. The common thread here is deviation. You are comparing this person against their own earlier self, not against the team average.

Which resignation signs need caution?

Some popular resignation signs only become useful after you check for non-quit explanations. PTO or late arrivals can point to a job search, but they can just as easily point to health needs, caregiving pressure or plain burnout.

Start with the employee's usual pattern before you read anything into the change. Someone who has always guarded their focus time has not suddenly disengaged because they skipped one optional meeting. A person who starts leaving early after a conflict, cancels several 1:1s in a row and stops touching future-facing work is a completely different case, and deserves a different level of attention.

Absence works better than lateness as a withdrawal signal, though it still cannot carry the conclusion alone. A meta-analysis of lateness, absence and turnover found the absence-turnover correlation at .25 while lateness-turnover sat at just .01. That is exactly why clock-watching alone proves so little.

LinkedIn polishing deserves the same restraint. An updated profile might mean someone wants more recruiter visibility, or it might mean they finally cleaned up a neglected public page. The safest response is curiosity about working conditions, not a quiet investigation into whether they are interviewing.

What data strengthens quit-risk signs?

Give quit-risk signs more weight when separate data sources point in the same direction at once. An engagement-score drop, a pattern of cancelled 1:1s and thinner peer feedback say far more together than any one of them says alone.

Managers usually pick up tone, effort and meeting behavior first. HR can sharpen that picture: calendar data shows fewer meaningful conversations, engagement scores slide, feedback from colleagues turns sparse. This matters because many avoidable exits happen in silence. Gallup found that 42% of employee turnover was preventable, and that nearly half of voluntary leavers had no proactive conversation with a manager or leader about their satisfaction, performance or future in the three months before they walked.

Use the data to decide where to act first. An engagement dip should trigger a manager check-in. Missed 1:1s should trigger a repair of the meeting cadence. Weaker peer feedback should prompt a closer look at team connection or simmering conflict. When the same employee also shows stalled career movement or a sudden workload spike, your retention response needs to move faster, and tracking your retention rate over time shows you whether those interventions are actually holding people.

How should managers respond to quit-risk signs?

When quit-risk signs surface, resist the urge to confront the employee with a resignation accusation. The stronger play is to confirm the pattern, ask a stay-interview question, lift one concrete burden and lock in a fast follow-up.

Begin by setting the last few weeks against the employee's earlier norm. Then ask yourself whether workload, health pressure, fuzzy priorities or team conflict could explain what you are seeing. Your first conversation should lean on observable language, something like: "I noticed our 1:1s have been harder to schedule lately, how are things going?" Asking "Are you quitting?" only pushes the person into self-protection and shuts the door you were trying to open.

A good stay conversation digs into what the employee still looks forward to at work. SHRM's stay-interview framework adds questions about what they want to learn, why they stay and when they last seriously thought about leaving. Close every one of these conversations with a single visible action.

The follow-through is where most retention attempts quietly fail. People share what's wrong, nothing visibly changes, and the conversation itself becomes one more reason to leave. If you want a fuller menu of moves, our roundup of data-backed retention tactics pairs well with the stay conversation.

How does AI surface attrition risk earlier?

AI earns its place when it connects resignation signals that managers normally see scattered across separate systems. Sprad's Talent Management Workspace and Atlas AI can surface attrition-risk clusters from engagement, 1:1, performance, career and business data, so HR acts before a resignation email ever lands.

Predictive analytics works best when it hands managers earlier context and leaves the actual conversation to a human. The useful output is an explainable prompt, not a verdict: this person's engagement fell, their 1:1 cadence changed, their development path stalled. That gives HR a concrete reason to back the manager with a stay conversation or a career review. Recent people-analytics research tests attendance, office automation and enterprise social-media data as added predictors of voluntary turnover, which is exactly why clusters beat any single flag. For Sprad, Atlas AI fits the retention workflow by joining signals that usually stay apart, letting engagement data, performance notes, meeting history and business-system context live in one place. That is the same logic behind our work on spotting resignation signals early.

How should HR monitor employees responsibly?

Design quit-risk monitoring as a support system employees can actually understand, not a surveillance net. People need transparency about how their data is used, managers need to stay in the loop as human reviewers, and access should be limited to the signals genuinely needed for retention action.

Continuous monitoring crosses a trust line the moment employees feel watched instead of supported. HR should spell out which signals the company uses, who can see them and what action they may trigger. In GDPR-heavy environments that usually means data minimization, role-based access, clear retention rules and documentation that legal or a works council can review. Eurofound's 2024 analysis warns that systematic, continuous tracking can infringe privacy and data-protection rights, which is reason enough to keep the guardrails tight.

The strongest guardrail is human accountability. Attrition-risk outputs should start a supportive conversation. They should never trigger discipline, quiet exclusion from projects or an automated employment decision.

Resignation risk needs timely conversations

The genuinely hard part is that most resignation signs look like performance problems at first glance. Treat them only as discipline issues and you will miss the workload strain, the career frustration or the broken relationship that produced the withdrawal in the first place. The very same signal can turn into an exit or a retention win, and which one it becomes depends largely on how quickly the manager moves.

The fastest retention wins almost always come from one visible fix to workload, growth or manager access, applied while the employee is still open to staying. Analytics can make the signals far easier to see, but the person on the other end still needs a trusted human across the table, not a risk score.

This week, review your team for baseline changes over the past two to six weeks and pick one person who needs a non-accusatory stay conversation. Pair that human check with a simple signal review across engagement, 1:1 cadence and performance data, and let Sprad's Talent Management Workspace do the connecting so you can spend your time on the conversation itself.

Frequently Asked Questions (FAQ)

Does increased PTO mean an employee is about to resign?

No, increased PTO on its own does not mean an employee is about to resign. Absence has a stronger link to turnover than lateness does, but PTO can equally reflect health needs, caregiving or burnout recovery. Treat it as meaningful only once it shows up alongside other withdrawal signs.

Are cancelled one-on-ones a resignation signal?

Cancelled one-on-ones become a useful warning sign when the pattern shifts from the employee's normal behavior. Direct proof on cancellations alone is thin, yet meaningful weekly manager conversations tie strongly to engagement. If 1:1s start disappearing while effort also drops, that combination is your cue to act.

How long should managers watch quit-risk patterns before acting?

Managers should usually compare the behavior against the employee's baseline over roughly two to six weeks. A single missed deadline or one quiet meeting is far too weak to interpret. A repeated pattern across output, attitude and schedule, however, is enough to start a supportive conversation.

What should a stay interview ask when someone seems ready to quit?

A stay interview should ask what keeps the employee at the company and what might push them to leave. Cover what they want to learn, what they still look forward to, and what you as their manager could do better. Every one of these conversations needs a visible follow-up to count.

Can a disengaged employee stay without resigning?

Yes, a disengaged employee can absolutely stay without resigning. Engagement predicts turnover risk, but job embeddedness, the labor market and personal constraints can keep someone in place. That still matters, because the person may remain present while their effort, morale and influence on the team quietly decline.

Can AI predict resignation risk accurately enough to use?

Yes, AI can help identify resignation risk when it works from clusters of signals rather than a single flag. It should combine engagement, 1:1, performance, workload and career data with human review at the end. Use the result to prompt support, never to make an automated employment decision.