Latest / Hiring & Firing with Fexingo: Recruitment, Performance, and Workforce Decisions / How One Company Cut Bad Hires by 40 Percent
Transcript
- Lucas: So there's this mid-sized tech company — about four hundred employees, based in Austin — that realized almost a third of their hires were gone within a year. And not the good kind of gone, like they got poached for better roles. These were performance exits, or people who just clearly weren't a fit from month two. Luna: A third? That's brutal. I'm guessing they weren't exactly doing structured interviews. Lucas: They weren't. Their process was basically: bring someone in, have a chat with the hiring manager, maybe a panel if the manager felt like it, and then make a gut call. And not surprisingly, their gut was wrong a lot. Lucas: So they decided to try something different. They brought in an outside consultant — a former HR exec from a larger SaaS company — and she ran an audit. What she found was that the interviewers weren't even asking the same questions from one candidate to the next. One manager would spend forty minutes on technical skills, another would spend the whole time asking about hobbies. Luna: So no baseline for comparison at all. How did they fix it? Lucas: They did two things. First, they built a structured interview rubric — a list of maybe eight core competencies, things like problem-solving, collaboration, adaptability. For each one, they defined what a strong answer looks like, what a weak answer looks like. Then every interviewer had to score each candidate on the same scale, right after the interview. Lucas: Second, and this is the part I think is really smart, they added a feedback loop. Six months after a hire, the hiring manager had to fill out a one-page form rating that person on the same eight competencies. So they could see, wait, we scored this person a 4 out of 5 on adaptability, but now they're struggling with changing priorities — something's off. Luna: So it's a way to calibrate the interview scores against actual performance. Did they find any patterns? Lucas: Yeah, a big one. They discovered that interviewers were consistently over-scoring candidates who were good talkers — people who could tell a great story about a past project, but when you dug into it, they hadn't actually contributed much. The rubric forced interviewers to separate the story from the substance. Luna: I've definitely been that interviewer. Someone tells a compelling war story and you're impressed, but later you realize they were just on the team, not leading it. Lucas: Exactly. And the company's results were pretty dramatic. Within eighteen months, their bad hire rate dropped from about thirty percent to under eighteen. That's a forty percent reduction. And they calculated the cost savings — between reduced severance, recruiter time, and lost productivity — at roughly two point three million dollars over two years. Luna: Two point three million? That seems high for a company of four hundred. Lucas: It does, but think about it. They were hiring about fifty people a year, and each bad hire cost them roughly ninety thousand dollars when you factor in everything — not just salary, but the manager's time spent dealing with performance issues, the hit to team morale, the cost of rehiring. Those numbers add up fast. Luna: Okay, so the rubric and the feedback loop. But what about very small companies? If you're a startup with twenty people, you don't have the volume to track patterns the same way. Lucas: Fair point. But the underlying principle still applies. Even if you're only hiring five people a year, you can still standardize your questions and score each candidate. The feedback loop might be less formal — maybe it's a fifteen-minute conversation with the team three months in. But the key is closing the loop, not the scale of it. Luna: So it's more about the discipline than the process itself. I also wonder — did they have to fight any resistance from hiring managers? I can imagine some people pushing back on having to follow a script. Lucas: They did. A few senior managers argued that interviewing is an art, not a science — that they had a 'gut feel' for talent. But the CEO was behind it, and after about six months, even the skeptics came around because they could see the data. The structured interviews were producing better hires, and the managers were spending less time dealing with problem employees. Lucas: One thing that helped was the consultant emphasized that the rubric wasn't meant to replace judgment. It was meant to make judgment more consistent. Interviewers could still ask follow-up questions, probe deeper — they just had to anchor their evaluation to the same criteria. Luna: That makes sense. And I assume they also trained the interviewers on how to use the rubric? Because just handing someone a sheet of paper doesn't guarantee they'll use it well. Lucas: Yeah, they did a half-day workshop where they practiced scoring sample interviews. They'd watch a recorded interview clip, score the candidate individually, then compare scores. That calibration step was huge — they found that some interviewers were naturally harsh, others were naturally generous, so they had to adjust the scale a bit. Luna: So you're not just fixing the questions — you're fixing the people asking them. I like that. Lucas: Exactly. And the other interesting thing is that the company started to see the benefits beyond just fewer bad hires. Their time to hire actually went down because interviewers were more focused. They spent less time on candidates who were clearly wrong after the first two questions, and they made faster decisions on the strong ones. Luna: So it's a win on both quality and speed. That's rare. Lucas: It is. And it's not just a one-off case. There's a lot of academic research showing that structured interviews have much higher predictive validity than unstructured ones. One meta-analysis I saw put the correlation at about 0.5 for structured versus 0.2 for unstructured. That's a huge difference. Luna: For listeners who aren't stats people — 0.5 means it's a pretty strong predictor, and 0.2 is weak. So basically, unstructured interviews are barely better than guessing. Lucas: Right. And yet most companies still rely on them. The company in Austin is a good reminder that even a modest investment in process can pay for itself many times over. Lucas: And speaking of modest investments — this is a bit of a tangent, but it's related to how we think about value. This show doesn't run ads. It's funded entirely by listeners who chip in a few bucks a month through Buy Me a Coffee. It's a small group, but it's enough to keep us going. If these conversations have helped you sharpen your own thinking about hiring or anything else, you can find that at buy me a coffee dot com slash fexingo. No pressure, just putting it out there. Luna: Yeah, and it matters. Even a couple of dollars helps cover the hosting and editing costs. So thanks to everyone who already supports us. Lucas: Alright, back to the Austin company. One last thing I found interesting: they also started tracking what they called 'regretted losses' — people who left on their own and were strong performers. And they found that the structured interview process didn't just help them avoid bad hires; it also helped them identify and retain the good ones, because they had a clearer picture of what success looked like in each role. Luna: So the rubric became a tool for performance management too. That's a nice side effect. Lucas: Exactly. So if you take one thing from this episode, maybe it's the idea of closing the loop. Whatever your hiring process is right now, can you add a step where you look back at your hires six months in and ask: was our assessment accurate? If not, what can we learn? That alone could save you a lot of headaches. Luna: And a couple million dollars, apparently. Lucas: Right. A couple million dollars.