I spent many years writing a blog about recruiting at Recruiting Daily. It was my job to hire speakers and plan our content calendar. I had to figure out which topics were the most popular and make sure we were ahead of the trends to get those people to join our community. One topic always won: sourcing.
No matter what, if I scheduled a session about sourcing candidates and finding their contact information, it would double whatever attendance we had at other sessions. Back then, we all knew why. At the time, finding candidate email addresses and getting them to respond was the hardest part.
Then, AI came along. Contact information became the easy part. I mean, ask the last guy who tried to online date one of my friends. We knew *everything* about him by the time she left for that date. Joking, kinda. The new problem? Can you write a job post that has enough information that both a human and a machine can match candidates to a job?
Why Job Posts Dictate AI Recruiting Success
We’ve all read enough bad job postings to know that most recruiters do not write clear, concise job posts. Most of them find the entire task of writing a job post to be an administrative burden and put very little effort into it. For years, that was ok. Recruiters could post the job and no matter how bad it was, we hired someone. People would “understand what we really mean” somewhere in those requirements.
But now, the first step of most AI recruiting tools is “enter a job post here” or “generate a job post.” That sounds good until you realize it’s built off one very flawed assumption: the job post you use will be concise, specific, clear, and use the keywords candidates need to understand if they are qualified. And those job posts aren’t good enough.
I believe that’s why AI recruiting projects are failing right now. You tell me. How are you going to use AI for interviews if you never define the skills you’re interviewing for? How are you going to implement AI salary analysis if you don’t know what skills you’re paying for? Short answer? You don’t accomplish anything without a clear definition of the job.
How To Write Job Posts That Can Power AI Recruiting
The first step? Stop using AI to generate first drafts of job posts based on a short prompt and a job title. Use your brain first. Voice to text your first draft, then use AI to proofread or ask if there’s anything you missed that would make the job post more clear to a reader.
Then, train recruiters on how to write a job post candidates can find. They need to know how to recognize bias, edit generic skills in mandatory requirements, and when to incorporate AI.
Finally, always use a job post audit to evaluate the quality of each job post. The audit is there for consistency. If a person doesn’t know what makes a job post bad, you won’t get anything that’s consistently good. You can buy my job post audit and tailor it for your team here.
In my presentation, The Billion Dollar Problem: Why You Need to Rewrite Your Job Posts In An AI World, I share my whole how-to for generating high quality job data and why it matters for AI recruiting. If you want to get HRCI or SHRM credits, buy this on-demand version. I’d love to train your association or team in 2027. Book a call here any time.