Exit interviews that produce something usable
Most exit interviews are conducted by the wrong person, at the wrong time, and produce a diplomatic summary that changes nothing.
The exit interview is standard practice and largely ineffective. A departing employee, in their final week, is asked by their manager or the founder why they are leaving, and gives an answer calibrated to preserve a reference and avoid an awkward final Friday. For a related product-level perspective on employee monitoring transparency, see more details. For broader independent guidance on this point, consult CIPD retention resources.
The information exists — leavers know more about what is wrong with a company than anyone still in it — but the standard format is designed almost perfectly to not collect it.
Fix the three structural problems
Who asks. Not the person's manager, and often not the founder. In a small company the least bad option is frequently someone from a different part of the business, or an external adviser for senior departures. The question is who the leaver can be candid with, and the honest answer is rarely the person they are leaving.
When. The final week is the worst timing: the person is finishing handovers, is emotionally disengaged, and has an immediate interest in a smooth exit. A conversation a month after they have gone, once the new job is under way and the reference is secured, produces markedly more.
How. A conversation, not a form. Forms collect ratings; the useful material is in the specifics that only emerge when someone follows up on an answer.
'What would have had to be different for you to have stayed?' It is concrete, it is forward-looking, and it produces an actionable answer far more often than asking why someone is leaving.
Ask about the decision, not the destination
The new job explains where they are going. It does not explain why they were looking, which is the part you can act on.
Useful follow-ups: when did you first start thinking about leaving, what happened around that time, was there a point where something could have changed your mind, did you raise any of this and what happened. The timeline is often more informative than the reasons — it identifies the specific month in which the company lost someone, and usually the specific event.
Separate what you can change
Some reasons are structural and permanent: no promotion path, no possibility of a certain kind of work, geography, pay ceilings the business cannot move. Recording these is worth doing, but acting on them is not always possible.
Others are ordinary and fixable: a manager who never gave feedback, a promise not kept, equipment never provided, a workload never rebalanced. Sorting the answers into these two buckets is what turns a set of interviews into a decision.
One departure is an anecdote
The temptation after a difficult resignation is to act immediately on that person's account, which risks restructuring around a single perspective that may be unrepresentative.
Keep a simple record across departures — role, tenure, stated reason, underlying reason, whether anything was raised beforehand. After five or six leavers, patterns appear that no individual conversation showed, and those patterns are what justify a change.
Close the loop visibly
If exit feedback leads to a change, say so internally. Not attributed to the individual, but stated: this came up when people left, and here is what we did.
The team knows who left and roughly why. A visible response tells the people still there that raising something has an effect — which is the mechanism by which they raise it before resigning rather than afterwards.