Employee Retention Research: Stay Interviews, Exit Interviews, and What Actually Works

Voluntary turnover costs U.S. businesses an estimated $1 trillion per year (Gallup). Replacing a single employee costs between 50% and 200% of their annual salary — and for senior or specialized roles, that figure can reach 400%. Yet 42% of employees who voluntarily leave say their manager or organization could have done something to prevent it (Gallup, 2024).

The problem is not that retention is impossible. The problem is that most organizations do not systematically listen to the people who are still there — and by the time they conduct an exit interview, it is too late to change anything.

This guide covers how to use qualitative research to understand what keeps employees and what drives them away, before they hand in their notice.

The Listening Gap

Gallup's 2024 research found that 45% of employees who left reported that no manager or leader proactively discussed their job satisfaction, performance, or future with the organization in the three months before their departure. Not one conversation.

Meanwhile, global employee engagement fell from 23% to 21% in 2024 — only the second decline in 12 years — costing the world economy an estimated $438 billion in lost productivity. 70% of team engagement is attributable to the manager, and low-engagement teams experience turnover rates 18% to 43% higher than highly engaged teams.

As Dick Finnegan, the world's leading authority on stay interviews, put it:

"Everything I had been taught about turnover in HR was wrong. Most employees don't leave because of the actions of the CEO — they leave because of the actions of their direct supervisor."

Retention is not an HR problem. It is a management problem. And it requires a management-level research practice to fix.

Stay Interviews: The Most Underused Retention Tool

A stay interview is a structured, one-on-one conversation with a current employee designed to understand what keeps them at the organization and what might cause them to leave. Only 17.2% of employers conduct stay interviews — yet companies that regularly seek employee feedback see a 14.9% reduction in turnover rates.

When to Conduct Stay Interviews

• Every 6–12 months for established employees, separate from performance review cycles
• At 90 days and 180 days for new hires
• Quarterly for employees in transition roles or after organizational changes
• Never combine stay interviews with performance reviews — the power dynamic corrupts the data

Who to Prioritize

Start with the people whose departure would hurt most:

• High performers and high-potential employees — highest cost of loss
• New hires in their first year — highest flight risk
• Employees in critical or hard-to-fill roles

Then expand organization-wide.

The Core Questions

Dick Finnegan's SHRM-validated 5-question framework:

1. "What do you look forward to when you come to work each day?"
2. "What are you learning here?"
3. "Why do you stay here?"
4. "When was the last time you thought about leaving, and what prompted it?"
5. "What can I do as your manager to make your experience at work better?"

Question 4 is the most diagnostically powerful. It surfaces real risk factors that employees rarely volunteer unprompted. The key is strong follow-up probing — "What prompted it?" and "What happened next?" — to reach specific, actionable detail.

Getting Honest Answers

The central challenge: employees fear that honesty will damage their standing. Research shows employees systematically underreport dissatisfaction with management and culture, instead citing safe answers like "better opportunity" or "pay."

How to get past the filter:

• Normalize the conversation. Frame it as something every employee participates in, not a signal that something is wrong.
• Separate from performance reviews entirely. If employees associate the conversation with evaluation, they will self-censor.
• Send questions in advance. Give employees a week to reflect before the conversation.
• Start with positive questions. "What do you look forward to?" builds rapport before the harder questions.
• Use a neutral interviewer for sensitive topics. When exploring management quality or organizational culture, third-party or AI interviewers get dramatically more honest responses.

Exit Interviews: Getting Truth from Departing Employees

Exit interviews capture why employees left. But standard paper-based exit surveys achieve participation rates as low as 15–30%, and even when employees participate, they systematically underreport the real reasons for leaving.

Timing

Conduct exit interviews during the notice period but not the actual last day. Too early and the employee is still in "professional mode." Too late and they are mentally checked out.

Consider a follow-up survey 3–6 months post-departure. With distance and perspective, former employees are often more candid about the real reasons they left.

The Honesty Problem

Departing employees protect relationships. They say "I got a better offer" when they mean "My manager never once asked about my career goals." They say "It was time for a change" when they mean "I watched three colleagues get promoted past me."

How to get real answers:

• The interviewer must not be the employee's direct manager. Ever.
• Third-party interviewers yield the most honest feedback but require budget.
• AI-powered interviews remove the social pressure entirely. Employees can be candid without worrying about burning bridges, giving a bad reference, or making the conversation awkward.
• Guarantee anonymity and demonstrate it through actions — aggregate findings, never attribute specific feedback to individuals in reports.

Core Exit Interview Questions

1. "What prompted you to start looking for a new role?"
2. "What could we have done differently to keep you?"
3. "How would you describe your relationship with your direct manager?"
4. "Did you feel you had opportunities for growth and development here?"
5. "What would you tell our leadership team if you knew it would be taken seriously?"
6. "On a scale of 1 to 10, how likely are you to recommend this organization as a place to work?"

Engagement Pulse Research

Stay and exit interviews capture individual depth. Engagement pulse research captures organizational patterns.

Research identifies the strongest predictive indicators of employee attrition as: job satisfaction (single most influential factor), years with current manager, years in current role, and years since last promotion. Machine learning models analyzing engagement data can achieve 77.5% accuracy in predicting attrition.

The practical implication: if you collect structured engagement data regularly, you can identify retention risks before they become resignations.

What to measure:

• Manager relationship quality
• Growth and development opportunities
• Sense of purpose and meaning in work
• Workload sustainability
• Recognition frequency
• Trust in leadership

Common Mistakes in Retention Research

1. Only doing exit interviews. By then, the person has already left. Stay interviews are preventive; exit interviews are diagnostic.
2. Having the direct manager conduct stay interviews without training. Untrained managers turn stay interviews into performance check-ins or defensive conversations.
3. Treating it as an HR initiative instead of a management practice. Retention research works when managers own the conversations and act on them.
4. Asking but never acting. Nothing destroys trust faster than soliciting feedback and visibly ignoring it.
5. Relying on annual engagement surveys alone. Annual surveys measure sentiment at one point in time. Ongoing qualitative research captures the "why" behind the numbers.

McKinsey's Great Attrition research framed it clearly:

"People keep quitting at record levels, yet companies are still trying to attract and retain them the same old ways. Organizations can treat this as the 'Great Attrition' or reframe it as the 'Great Attraction' by genuinely listening and adapting."

Retention Research with Koji

Traditional stay and exit interviews are bottlenecked by the same constraint: someone has to schedule, conduct, and analyze every conversation. For a 500-person organization running annual stay interviews, that is 500 conversations — an impossible workload for most HR teams.

Koji makes this scalable.

How it works:

1. Start from a template. Koji offers HR-specific templates for Stay Interviews, Exit Interviews, New Hire Check-ins, Manager Effectiveness research, and Employee Engagement studies — each built on validated frameworks like Finnegan's 5-question model.
2. Customize with the AI Consultant. Tell it about your organization, your retention challenges, and what you want to learn. It builds a tailored interview plan with the right question types — open-ended for exploration, scale questions for measuring satisfaction, and yes/no questions for benchmarking.
3. Distribute the link. Send it to employees via email, Slack, or your HRIS. No scheduling required. Employees complete the interview on their own time, voice or text.
4. AI conducts the interview. Koji's AI interviewer follows the structured plan while adapting follow-ups in real time. When an employee mentions frustration with their manager, the AI probes deeper. When someone rates their growth opportunities as a 3 out of 10, it asks what would make it a 7. No human interviewer can maintain this consistency across 500 conversations.
5. Get actionable patterns. Koji synthesizes themes across all interviews — surfacing the top retention risks, the strongest engagement drivers, and the specific actions that would make the biggest difference. No manual coding, no spreadsheet analysis.

The result: the depth of 1:1 stay interviews at the scale of an annual survey, with the honesty of a neutral third-party interviewer.

Predictive Retention: From Reactive to Proactive

The most advanced retention practices use qualitative data to predict attrition before it happens. Research shows that engagement data collected 6–12 months prior reliably surfaces turnover patterns across job roles, tenure levels, and business units.

By running ongoing stay interview programs through Koji, organizations build a living dataset of retention signals. Over time, patterns emerge: which teams have the highest flight risk, which manager behaviors correlate with retention, which career development gaps surface repeatedly.

This transforms retention from a reactive scramble ("Why did our best engineer quit?") into a proactive practice ("Our engineering team's growth satisfaction scores dropped 20% this quarter — let's investigate before we lose anyone").

Key Takeaways

• 42% of turnover is preventable — but only if someone asks before the employee decides to leave.
• Stay interviews are the highest-ROI retention practice. Only 17% of employers use them.
• Exit interviews alone are insufficient. Departing employees underreport real reasons. Follow up 3–6 months later for honest feedback.
• The manager is the #1 factor. 70% of engagement variance is attributable to the direct manager.
• Neutral interviewers get more honest answers. AI-powered platforms like Koji remove the social pressure that causes employees to self-censor.
• Make it ongoing, not annual. Continuous qualitative research catches retention risks before they become resignations.