Latest / Hiring & Firing with Fexingo: Recruitment, Performance, and Workforce Decisions / Why Exit Interviews Are a Hiring Tool
Transcript
- Lucas: So there's this mid-sized tech company, about four hundred employees, and they had a turnover problem that just wouldn't quit. First-year attrition was hovering near thirty percent, which is ugly for any firm, but for them it was personal — every bad hire cost them roughly one point five times that person's salary in recruiting, onboarding, and lost productivity. Luna: Right, and that math is brutal. But what did they actually do about it? I'm guessing they didn't just throw more money at recruiters. Lucas: No, they did something smarter. They went back and read every exit interview from the previous two years — forty-seven in total — and coded the reasons people gave for leaving. And what they found was that nearly two-thirds of those departures traced back to a mismatch between the job as it was advertised and the job as it actually was. Luna: So the classic bait and switch, but not necessarily malicious — just a fuzzy job description and a vague interview process. Lucas: Exactly. People were hired for one thing and then spent their days doing something else. One engineer said the role was pitched as greenfield development, but he spent eight months maintaining legacy code. A marketing hire was told she'd own strategy, but she ended up doing vendor coordination. Those aren't rare stories, but this company actually listened. Lucas: So they did the obvious thing first — they rewrote every job description for the roles with the highest turnover. And they didn't just tweak the bullets. They went to the teams and asked: what does this person actually do in the first ninety days? What does a typical Tuesday look like? What's the least glamorous thirty percent of this job? Luna: That's a great question, because nobody sells the boring parts. Every job description says 'collaborate with cross-functional teams,' but nobody mentions the weekly status report that takes three hours. Lucas: Right, and they made that explicit. One role now literally says 'you will spend about twenty-five percent of your time on data entry.' And you know what? Some candidates self-selected out, which is fine. The ones who stayed knew what they were signing up for. Lucas: But the bigger shift was in the interview process. They built a structured rubric around the specific tasks that showed up in the exit data. For their customer success roles, they added a role-play where the candidate handles an angry client — because that was a top complaint. For their data team, they added a debugging exercise, because people said the job was less about building models and more about fixing broken pipelines. Luna: So they turned exit interview complaints into interview questions. That's a pretty direct line from 'this job sucked because X' to 'do you actually want to do X?' Lucas: Exactly. And the results were striking. Within twelve months, first-year turnover dropped from thirty percent to twenty-one. That's a nine-point improvement, which for a company of that size means about a dozen people who didn't quit, plus the savings from not having to rehire and retrain. They estimated it saved them north of two million dollars in hard costs. Luna: That's a huge win. But I have to ask — exit interviews are notoriously unreliable, right? People are polite, they don't want to burn bridges, they say 'I'm leaving for a great opportunity' when really they hated their manager. Lucas: That's the biggest criticism, and it's valid. So the key is how you collect and code the data. This company didn't just take the first answer at face value. They did a few things to get closer to the truth. Lucas: First, they normalized the process. Every exit interview was conducted by someone from HR, not the direct manager, and they used the same structured questions every time. No leading, no 'was it something we did?' They asked things like 'what part of the job took up more time than you expected?' and 'when did you first consider leaving?' Luna: That 'when did you first consider leaving' question is sneaky good. It pushes people to give a timeline instead of a generic reason. Lucas: Yeah, and it surfaces early warning signs. One woman said she started thinking about leaving in week three, when she realized the 'mentorship program' was just a shared drive with PDFs. That's the kind of detail you'd never get from 'what's your reason for leaving?' Lucas: Second, they did a follow-up survey six months later. Not with everyone — they focused on people who left for 'another opportunity' and asked them to rate their new job against their old one. Were the things they complained about actually better elsewhere? That gave them a sanity check on whether the complaints were real or just venting. Luna: That's a smart move, because if someone leaves saying 'I want more growth' and then ends up in a job with even less autonomy, maybe growth wasn't the real issue. Lucas: Exactly. And that's the thing — exit interviews are just one data point. The companies that use them well treat them like a signal, not a verdict. They aggregate, they look for patterns, and they pair them with other data like engagement surveys and stay interviews. Luna: Stay interviews — that's a good bridge. Because if you only talk to people when they're leaving, you're always reacting. But if you check in with your best people before they get to that point, you might catch the mismatch earlier. Lucas: Right, and that's actually where this company went next. They started doing a thirty-day check-in with every new hire — not a performance review, just a conversation about whether the job matches what was promised. And they found that issues were showing up way before the six-month mark. Lucas: One new engineer told them, pretty bluntly, 'the job description said I'd be working on the new platform, but my onboarding has been all about the legacy system.' That's exactly the kind of mismatch they'd been catching in exit interviews — and now they were catching it in week two. Luna: So they could fix it right then, either by adjusting the work or by having a conversation about expectations. That's way cheaper than losing someone at month eight. Lucas: Exactly. And that's the real lesson here — exit interviews aren't just a way to say goodbye. They're a source of intelligence that should feed back into how you hire, how you onboard, and how you manage. If you treat them as a formality, you're throwing away data that could save you millions. Luna: And honestly, this is why I love this podcast. We get to dig into these real-world examples and pull out the actual mechanics. It's the kind of thing you can apply the next day. Lucas: That's the whole point. And it's also why we keep making these episodes — because listeners tell us they're useful, and that keeps us going. We're a small independent show, no ads, no corporate backing. Just a handful of listeners who chip in a few bucks a month through buy me a coffee dot com slash fexingo. That's literally what funds this operation. Luna: Yeah, and it's not a lot of people, but it's enough to keep us ad-free and focused on the content. If you've gotten something out of these conversations, that's the way to keep them coming. Lucas: No pressure, just putting it out there. So — back to the data. One of the most interesting patterns they found was that the mismatch wasn't always about the job duties themselves. Sometimes it was about the level of autonomy, or the pace, or even the team culture. Luna: Can you give an example? Lucas: Sure. One senior product manager left because she said the role had 'strategic' in the title but she spent most of her time on project management. That's a classic scope mismatch. But when they coded for that, they realized the job description did say 'strategic' — but the interview process never asked about project management tolerance. So they added a question to every PM interview: 'How do you feel about managing Jira tickets?' Luna: That's such a simple fix. And it probably cut down on a lot of pain for both sides. Lucas: It did. And the beauty is, they didn't need to hire a consultant or buy new software. They just listened to what people were saying on their way out and let that inform the hiring process. That's the kind of thing any company can do, regardless of size. Lucas: So if you're sitting on a pile of exit interviews, don't just file them away. Spend an afternoon with a spreadsheet and start coding. You might be surprised what you find. Luna: And if you do, I'd love to hear about it. Maybe we'll do a follow-up episode with your numbers. Lucas: That would be great. For now, the takeaway is simple: exit interviews are a hiring tool, not a goodbye ritual. Use them.