Exit Interviews Reveal What Went Wrong
02/28/07
Exit interviews, like autopsies, can reveal what went wrong and why. But how you manage them, what information you gather and how you use it make the difference between a valuable trouble-shooting tool and just another paperwork exercise. These suggestions will help you get your money’s worth from this posthumous morale survey.
The Devil Is in the Details
How you conduct the conversation has a major impact on the quality and amount of information you’ll get from it.
For obvious reasons, never allow immediate supervisors to interview departing employees. The ideal interviewer is an HR staffer with proven interviewing skills who is adept at putting employees at ease and drawing them out. In addition, HR people tend to have high credibility with the rank and file and are perceived as less likely to betray confidences than someone in the employee’s chain of command.
Ensure privacy. Ringing telephones, buzzing pagers, and drop-in visitors inhibit employees’ responses and devalue the interview’s importance. Prevent these interruptions; give the meeting the respect it deserves.
Consider doing the survey by mail. While it’s easier to hold an exit interview on the employee’s last work day, some HR authorities recommend sending a postage-paid questionnaire to the employee’s home several days after separation. This gives people time to collect their thoughts, reflect on their entire work experience, and perhaps gain a better perspective than they had on their last day. The downside: You’ll get fewer responses.
Whether the interview is done in person or by mail, ask workers to be specific about their criticisms. Vague remarks such as “lack of challenge” or “inadequate training” give you nothing tangible to go on. Ask them to spell out why they feel the way they do, along with explicit examples whenever possible. Urge them to propose how any toxic conditions that caused them to leave could be improved or eliminated.
What to Ask
It’s important to gather profiling data on employees such as age group, length of time employed, department, division, and job classification or title. This information helps you identify the typical high-risk turnover candidate.
Understandably, most employees won’t want to level with you about their reasons for leaving. Some are merely ill at ease; others may fear reprisals from ex-supervisors. Nevertheless, you’ve got to encourage them to supply enough details to let you pinpoint the main reasons why most people are quitting. It’s the only way to get to the root of the problem. Your discussion should focus on these topics:
Salary
Benefits
Opportunity for advancement
Training
Relationship with supervisor
Relationships with coworkers
Reasons for leaving that aren’t related to the job or company
About salary: While it’s important to know if your pay scale has slipped below that of competitors, don’t capitulate on pay complaints unless an employee is virtually indispensable. If you negotiate too willingly, other workers may try to cut themselves a better deal just by threatening to quit.
Always leave space for spontaneous comments. Sometimes they’re more useful than answers to your canned questions. Moreover, no matter how many questions you ask, none of them may address why this particular employee resigned.
Revise your interview questions as often as necessary to reflect changes in technology, training methods, job classifications, and higher management decisions that realigned reporting relationships or shifted communication channels.
Follow Up
Summarize and report interview results promptly to each department manager—even if a departing employee comments negatively on him or her. Concerned supervisors should want to know about such criticism and adjust their behavior accordingly. Include comparative data that show how each one’s turnover rate compares with his or her peers.
Share this information with senior managers, of course, because they have the authority to take corrective action in high-turnover jobs or locations and slow down that expensive revolving door.