Exit interview report template for leadership

An exit interview report is the document that turns a stack of individual conversations into something leadership can act on: how many people left, the themes that repeated, the numbers behind each theme, and what is being done about the ones that clear the bar for action. It is built from the forms described in exit interview form template, using the method in how to analyze exit interview data. This page is the report itself: the structure, a filled example, and what to leave out.

A report that lists every comment from every exit interview is not more thorough, it is less useful — leadership has to do the theme-finding themselves, which is the one job this document exists to do for them.

What this report is for, and what it replaces

Without this report, exit interview forms sit in individual employee files and nobody sees the pattern across them. This report exists to answer three questions leadership actually has: how many people left and from where, what did they say in common, and what are we doing about it. It is not a transcript archive, not a performance record on any single manager, and not a substitute for escalating a serious disclosure the same day it happens — that goes through your standard HR process immediately, not into next quarter's report.

The report template

EXIT INTERVIEW REPORT — [QUARTER/PERIOD]
Prepared by: [name/role]   Distributed to: [leadership/department heads]
Period covered: [dates]   Report date: [date]

1. HEADLINE NUMBERS
   Total departures this period: [n]
   Voluntary: [n]   Involuntary: [n]   Contract end: [n]
   Departures with an exit interview completed: [n] of [n] ([%])
   Compare to prior period: [n] departures, [+/-X] change

2. TOP THEMES THIS PERIOD
   (List only themes that met your action threshold — see the
   analysis method for how these were coded and counted.)

   Theme: [short name]
   Mentioned in: [n] of [n] exits ([%])
   Departments/teams: [list]
   Illustrative quote (anonymized): "[quote]"
   Recommended action: [specific action]
   Owner: [name/role]   Target date: [date]

   [Repeat for each theme that cleared the threshold]

3. THEMES TO WATCH (below the action threshold, noted for next period)
   [Theme] — mentioned in [n] of [n] exits. No action yet; will
   revisit if it recurs next period.

4. ESCALATIONS THIS PERIOD
   [n] disclosures were routed to HR/legal outside this report's
   normal review cycle, per standard process. Do not name individuals
   or provide detail here; state only the count and that they were
   handled through the standard process, separate from theme analysis.

5. PROGRESS ON PRIOR PERIOD'S ACTIONS
   Action from last report: [action]   Status: [done/in progress/
   not started, with reason]

6. WHAT WE ARE NOT ACTING ON, AND WHY
   [Theme or comment that did not clear the threshold, or where the
   organization has decided not to act, stated plainly rather than
   silently dropped.]

7. METHOD NOTE
   Exit interviews completed by: [role(s)]. Themes coded using:
   [method, e.g. see how-to-analyse-exit-interview-data]. Quotes
   anonymized by: [method — no names, no identifying detail beyond
   department where department size protects anonymity].

A filled example

An invented quarter at a 340-person company, showing the report at a size leadership can actually read in five minutes.

Notice what this report does not contain: no employee names, no direct manager named next to a criticism (the coaching note goes to the director, not published with the theme), and no theme presented as fact beyond what the numbers support — 31 percent is stated as 4 of 13, not rounded into false precision.

Anonymizing without losing usefulness

Setting the action threshold before you need one

Decide the threshold before the first report, not while writing it, or it will bend to fit whatever story you already believe. A reasonable starting rule for a mid-size organization: three or more exits naming the same specific, actionable issue within the period becomes a numbered theme with an action and an owner; one or two mentions become a "theme to watch" carried into the next report; a single mention that is serious on its own (safety, harassment, discrimination) bypasses this threshold entirely and goes to escalation the same day, regardless of count.

What not to include

• Every comment from every exit. That is the raw data, not the report; keep it in the individual forms.
• A named manager next to unverified criticism. Route it to their boss privately with the count and the substance; publishing it in a circulated report is a different, harsher action than the evidence usually supports.
• Compensation figures from a single exit. One person's stated expectation is not a compensation benchmark; that is a separate exercise with its own data.
• Speculation dressed as a theme. If two people mentioned something vaguely related but for different reasons, say so rather than merging them into a theme that did not actually repeat.

Distribution and who sees what

The full report, including department-level themes, goes to whoever owns retention company-wide — typically an HR or People leader. Each department head sees their own section and the company-wide headline numbers, not other departments' quotes. This keeps the report useful without turning it into a public ranking of which team is unhappiest, which tends to produce defensiveness rather than action.

Tracking a theme across quarters

A single quarter's report answers "what happened." A short trend table answers the more useful question: "is this getting better." Add one to the report once you have three or more periods of data, tracking only themes that have appeared more than once.

The figures above are invented to show the format. The point of the trend column is to stop a theme from being reported as new every quarter when it is really the same unresolved issue moving slowly, or to catch a spike tied to a specific team or event rather than treating every period as disconnected from the last.

Common mistakes

• Publishing raw quotes without checking whether they identify someone. A quote that names a project only three people worked on is not anonymous just because the name is removed.
• Reporting a percentage from a tiny sample without the denominator. "50 percent said X" from two exits out of four is technically true and practically misleading; state the fraction, not just the percentage.
• No section 5. Skipping progress on prior actions is how a report becomes a ritual nobody expects to change anything.
• Writing the report the same day as the last exit, from memory. Build it from the forms and the coded themes, not recollection of the more memorable conversations, which is where recency bias creeps in.