Definition · Workforce planning and analytics
Flight Risk
Definition
A flight risk employee is someone judged likely to leave the organization voluntarily in the near future, often the next 3 to 12 months. Flight risk is assessed by managers, through stay interviews, or with attrition risk models that score employees on factors such as tenure, pay position, manager changes and recent promotions.
Also called: Retention risk, Attrition risk, Turnover risk
Signals that predict leaving
Some signals show up in HR data, others only in conversation.
- Pay below market or below peers, especially a low compa-ratio after a strong review.
- Time since last promotion or pay change well above the norm for the role.
- A new manager, or a manager who has lost several people recently.
- Tenure milestones: risk tends to rise around the first 90 days, at one year, and after vesting dates or bonus payouts.
- Schedule or commute changes, which matter a great deal for hourly staff.
- Behavior changes: fewer voluntary shifts, more single-day absences, dropping out of projects, sudden interest in using up PTO.
None of these proves anything about one person. A rise in Friday and Monday absences can mean job interviews or a sick child.
How attrition risk scores are built
Many HR analytics teams build a predictive analytics model trained on past leavers. The model looks at what was true of people in the months before they resigned and scores current employees on the same variables. Outputs are usually bands (high, medium, low) rather than exact probabilities, because the probabilities are not precise enough to act on for individuals.
Keep the inputs defensible. Do not use protected characteristics such as age, sex, pregnancy or disability, or obvious proxies for them. Be careful with data from monitoring tools and personal browsing; in several states and under GDPR in Europe, using that data can create privacy and notice obligations. Rules vary by jurisdiction, so check before you build.
Acting on flight risk
The score is only useful if it leads to a conversation. Focus first on high-risk people whose departure would be regrettable: strong performers, scarce skills, people with knowledge nobody else has. Our page on regrettable attrition covers how to define that group.
The best follow-up is a stay interview: ask what keeps them, what might make them leave, and what one change would help. Then fix what you can, whether pay, schedule, a development opportunity or a manager problem, and do it before a counteroffer is the only option. Counteroffers made after a resignation have a poor record of keeping people for long.
Flight risk in high-volume hourly workforces
In contact centers, warehouses and retail, individual scoring matters less than spotting risky cohorts. A hiring class whose 30-day attendance is already slipping, a shift whose supervisor changed last month, or a site where a competitor just opened down the road paying $1.50 more an hour: each is a group-level flight risk. Track early warning signals by cohort and respond with group fixes such as schedule stability, pay adjustments or supervisor coaching.
Mistakes to avoid
- Labeling someone a flight risk and then quietly passing them over for opportunities. That makes the prediction come true and can look like retaliation.
- Sharing individual risk scores widely; keep them to HR and the direct manager.
- Treating the model as a verdict instead of a prompt to talk.
- Ignoring the managers whose teams consistently show high risk.
Frequently asked questions
What does flight risk mean in HR?
It means an employee is considered likely to leave voluntarily in the near future, based on signals such as pay position, career stagnation, manager changes or changes in behavior.
What is the difference between retention risk and attrition risk?
They mean the same thing in practice. Retention risk is the risk of failing to keep someone; attrition risk is the risk of them leaving. Both describe the likelihood of voluntary turnover.
How accurate are flight risk models?
They are useful for ranking groups and directing attention, but individual predictions are often wrong. Treat scores as a reason to have a conversation, not as a fact about the person.
Should employees be told they are a flight risk?
Usually no. Have a stay conversation about what would keep them instead. Telling someone they have been scored as likely to leave can damage trust and change their behavior.