Latest / Hiring & Firing with Fexingo: Recruitment, Performance, and Workforce Decisions / How One Job Description Kept Out Half the Applicant Pool
Transcript
- Lucas: Alright, let's start with a number. The average job description on LinkedIn or Indeed right now contains more than 80 individual requirements — skills, years of experience, education, certifications, personality traits. Luna: Eighty? That sounds like a wish list, not a job description. Lucas: Exactly. And a 2024 study from the Society for Human Resource Management found that the most effective job descriptions — the ones that actually get the right person in the door — have fewer than 10 requirements. Luna: So we're talking about a 90 percent reduction in what we ask for. That's a huge gap between what recruiters write and what actually works. Lucas: It is. And the problem isn't just inefficiency — it's exclusion. The 2025 Greenhouse Hiring Benchmark Report looked at over 3 million job applications and found that when companies reduced their listed requirements from 15 to 6, applicant volume went up 102 percent. And the quality of hire, measured by manager satisfaction at 90 days, actually stayed flat or improved slightly. Luna: So you double your pool without sacrificing quality. That seems like a no-brainer. Why aren't more companies doing it? Lucas: Because the people writing the descriptions — hiring managers and recruiters — are afraid of lowering the bar. They think every line item they drop is a concession on quality. But the data says the opposite. When you list 15 requirements, you're really screening for the candidate who's willing to apply to a job with 15 requirements, not the candidate who can actually do the work. Luna: There's also the legal angle. Over-specifying credentials can get you into trouble with the EEOC. I remember a case from last year — a retailer settled for about $200,000 because they listed 'Bachelor's degree preferred' for a warehouse supervisor role. The EEOC argued that requirement disproportionately screened out Black and Hispanic applicants and wasn't actually needed for the job. Lucas: Yeah, that's a 2025 settlement. And it's not an isolated case. The EEOC has been increasingly aggressive about job descriptions that create artificial barriers. And it's not just degree requirements — it's also years of experience, specific software tools, even physical requirements that aren't validated. Luna: I want to go back to that Greenhouse data. Double the applicants — but does the applicant pool get less relevant? Or are people with the right skills just more likely to apply when the list is shorter? Lucas: Great question. The Greenhouse report tracked that. They looked at 'relevant applicants' — people who had at least half the skills listed — and that number also went up, by about 70 percent. So the pool got bigger and more qualified. The key insight is that when you drop the fluff requirements, you attract people who actually have the core skills but might have been scared off by that fifth year of experience or that niche certification. Luna: Let's get specific. Can you give me a concrete before and after example? I think that's what our listeners need to hear. Lucas: Sure. I talked to a mid-size SaaS company — about 500 employees — that was hiring a senior software engineer. Their original job description had 14 requirements. Including '5+ years of React experience,' 'experience with AWS Lambda,' 'knowledge of Docker and Kubernetes,' 'Bachelor's degree in Computer Science or related field,' 'experience leading a team of at least 3 engineers,' and so on. Luna: That's a lot for one person. Lucas: Right. So they cut it down to six. They changed '5+ years of React' to 'strong proficiency in React,' dropped the AWS Lambda requirement entirely, kept Docker and Kubernetes as 'nice to have,' removed the degree requirement and replaced it with 'equivalent experience or willingness to learn,' and changed 'team lead experience' to 'ability to mentor junior engineers.' Luna: And what happened? Lucas: Applicant volume went from about 40 applicants in the first two weeks to over 120 in the same period. time to fill dropped from 45 days to 22. And the person they hired — a woman who had been working as a front-end developer at a small agency — had never used Kubernetes. But she learned it in her first two months, and her manager rated her as a top performer at the six-month mark. Luna: So the 'nice to have' became a 'learn on the job.' That's a perfect example of how over-specifying screens out the people who could actually grow into the role. Lucas: Exactly. And there's another layer here. The study also found that women and people of color are less likely to apply to a job unless they meet nearly every requirement listed. Men and white candidates are more likely to apply even if they only meet 60 percent. So by having 14 requirements, you're not just shrinking the pool — you're skewing it demographically. Luna: That's a really important point. The act of writing a long list doesn't just affect who applies — it affects which groups apply. And that's where the legal exposure comes back in. Lucas: Right. So what's the practical takeaway? I think there are three rules. One: for every requirement you add, ask yourself — is this truly essential to do the job on day one? Two: replace years of experience with proficiency levels. And three: use a 'must-have' column and a 'nice to have' column, and keep the must-haves to six or fewer. Luna: And I'd add a fourth: test your own job descriptions. If you can't explain why a certain requirement is there, it probably shouldn't be. Lucas: That's a good one. And look, I know there are recruiters listening who might be thinking, 'My hiring manager will never go for that.' And that's a real challenge. But you can start with a pilot. Pick one role, write a stripped-down description, and track the data. Show them the results. The Greenhouse report gives you a benchmark to point to. Luna: I think that's where the tension really lives — between what hiring managers think they need and what the data says actually works. And this episode is a good example of why we keep coming back to these conversations. A lot of what we think is common sense in hiring turns out to be common practice that doesn't hold up to scrutiny. Lucas: Yeah, and speaking of that — it's honest moments like this that make me glad we have a small group of listeners who chip in to keep these conversations going ad-free. If these episodes have moved your thinking forward in some small way, a handful of people already help out at buy me a coffee dot com slash fexingo. No pressure, but that's literally what lets us do deep dives like this without selling ad space. Luna: It's a pretty straightforward model. Listener supported, no sponsors, no filler. And it means we can keep drilling into the data instead of the fluff. Lucas: Exactly. So back to the practical side — I want to suggest one more thing. Even if you can't reduce requirements, you can change how you present them. Order matters. Put the most essential skills first, and group the nice to haves at the bottom. Studies show that candidates read the first three bullet points and then scan. If the first three are reasonable, they'll read on. If the first three are aggressive, they click away. Luna: That's a small change that costs nothing and could have a big impact on who applies. I'm going to try that with my own team. Lucas: Okay, let's wrap with one final number. The average time a candidate spends reading a job description is about 12 seconds. If your list is longer than a few bullet points, you've already lost most of them. So keep it tight, keep it honest, and let the data guide you. Luna: I think that's a solid place to leave it. Next episode we're going to talk about something that affects every single hire — reference checks. Do they actually work, or are they just theater?