Every departing employee gets a 30-minute exit interview with HR: how was your experience, what could we have done better, would you recommend us to a friend. We went back through eighteen months of these notes. The same three complaints appear, independently, from people who never worked together and in several cases never met. We'd like to explain what actually happens to this feedback, because "we take this seriously" is the stated policy and considerably less true in practice.

The complaint that appears nine times

"My manager gave almost no feedback between reviews, so I never knew where I stood until it was too late to fix anything." Nine different departing employees, across four different teams, independently said some version of this sentence over eighteen months. Nobody cross-referenced these nine exit interviews against each other until we did it for this post. Each one was filed, individually, as one person's departure story, rather than recognized as the same structural problem recurring across the org.

Why the pattern never gets noticed in real time

Because exit interviews are conducted by whichever HR person happens to be assigned that month, filed into a shared folder, and almost never re-read as a set. Reading nine exit interviews side by side to find the pattern requires someone to decide, proactively, that this folder is worth re-opening — and nobody's job description includes "periodically audit the exit interview archive for repeating structural complaints," so the folder accumulates individual stories and never becomes the aggregate signal it could be.

The sentence that ends almost every interview

"Thanks so much for this — we really value the feedback." Said sincerely, by a genuinely caring HR person, about feedback that will be filed and then functionally never acted on, because the HR person taking the note usually has no authority to fix a manager's coaching habits, no mandate to escalate a pattern across nine separate departures, and no process that routes "I never got feedback between reviews" to whoever's actually responsible for manager training.

What actually gets acted on

The complaint that's specific, recent, and attached to someone still in a position to cause more departures if ignored — "my manager screamed at me in a 1:1 two weeks ago" gets escalated fast, because the risk of inaction is immediate and legible. "Nobody gave me feedback for eight months" describes a slower, structural problem with no single incident attached, and slow structural problems without a single incident compete poorly for attention against literally everything else happening that week.

Why the interview still gets conducted anyway

Partly because it genuinely helps the departing employee process their experience and leave on better terms, which has real value independent of whether the company acts on anything. Partly because conducting exit interviews is itself a defensible HR practice — "we always ask for feedback" is a sentence a company can say regardless of what happens to the feedback afterward, and the interview's existence does real reputational work even when its contents do none.

What we're not saying

We're not saying exit interviews are worthless — they occasionally do surface something specific enough and urgent enough to drive a real fix, and the closure they provide departing employees is a real, if secondary, benefit. We're saying the folder, taken as an aggregate, usually contains a clearer diagnosis of what's wrong with the company than any engagement survey, and almost nobody treats it as the diagnostic tool it actually is.

Our recommendation

Read the last twelve months of exit interviews together, once, specifically looking for the sentence that repeats. It will be there. It was there in ours — nine times, same root cause, never flagged until we went looking specifically for this post.

If your own exit interview folder hasn't been read as a set since it started, email us at contact@b2bsaas.guru. We'll help you find the pattern. We already know it's going to be an uncomfortable one.