Does the way you write a job description change who applies?
I run a business founded on the belief that people can do lot more than we credit them for, and we should hire like it.
So when I set out to analyze a large batch of resumes that came to me through two different job descriptions for the same role – the first a traditional job description heavy on industry jargon, and the second a competency-focused version – I thought the answer would be an unequivocal yes. Writing a job description that clearly describes what the work actually involves should help qualified people realize that they’re qualified for the job.
Of course, it was more complicated than that.
My study
I qualitatively coded 116 resumes for core competencies (eg, data entry accuracy), positive indicators (promotions, awards), and negative indicators (resume presentation issues, short tenure). Then I ran a model-based clustering algorithm on the coded data to see if natural “types” of candidates emerged in the pool, and whether either job description drew more of any type.
Three profiles emerged from the data:
Low signal, low risk: clean resumes, modest skill markers, few red flags.
Strong executive function, high positive indicators: candidates who lead with awards, promotions, and self-direction. Generalists who seem to learn on the job.
Strong core skills, high risk indicators: candidates with the strongest skill signals, but also the most concerning markers around tenure and resume presentation.
Then I looked at the second part of the question: did either job description draw more of any one profile?
The answer: no. Both job descriptions drew candidates from all three profiles, in roughly similar proportions. The jargon-heavy description and the competency-focused version drew remarkably similar mixes of people.
What I took away
Writing a competency-oriented job description didn’t draw a different kind of candidate. At least not in this analysis, at this sample size (I am definitely going to try this again on a larger pool).
But here’s what it did do: it gave me a much clearer view of who was actually in the pool. A traditional review would have flattened these candidates into ‘qualified’ or ‘not qualified’. Clustering on competencies showed me that the pool had real structure underneath the typical binary.
What this means if you’re hiring
So, no, the job description didn’t sort candidates, but the way we read resumes does. The job description is only part of the problem. The question is less “how do I attract the right kind of candidate?” and more “how do I see the candidates I’m already getting more clearly?”. A clearer job description is still worth writing because it gives both you and the candidates a sharper basis for comparison. The bigger lever, though, is in how you read what comes in.
I presented this work at the Modern Modeling Methods Conference at Fordham last week. It started when I finally had real-world data to test ideas about skills-based hiring that I’ve been working with for years.
There’s a lot more to learn here, so watch this space!
