I found some interesting data last week while doing some research for my new presentation. It’s from BambooHR and they’re saying time to fill has actually gone up over the last 3 years. Yes, even with all that AI investment and all those keynotes about the future of work, time to fill is heading in the wrong direction from 19 to 38 days. What’s the deal? Every survey headline boasts about how some random percentage of teams are using AI recruiting tools to speed up one task or another. Why aren’t we faster yet? 

My hypothesis is that the fundamental issue comes down to content quality and clarity. Job postings and job architecture are the foundational data layer for every single AI aspiration in your organization. You cannot use AI effectively for job post writing, interviewing, setting compensation, or making job offers if you can’t explicitly define what you are looking for in the first place. AI can’t do that part for you. 

Now, it’s complicating you being found by the right candidates, too. About 30% of ChatGPT users are using the tool for job search. That means millions of people are using AI search engines and candidate-matching agents to find jobs. If your job post doesn’t have the information they’re looking for, you’re never going to be found. 

Skills-Based Hiring And Other Tools That Won’t Work For AI Search Engines

Unfortunately, even with all the intel in the world at their fingertips, many hiring managers still rely on made-up requirements like requiring five years of experience or arbitrary college degrees. When they’re pressed on why five years is required instead of four, the logic falls apart quickly because those numbers are often pulled out of thin air. When you feed vague expectations into an AI system and ask it to find a "highly collaborative team player," you end up with a useless pile of candidates. AI cannot guess what you need when they’re matching candidates to your job. You have to define the specific, real-world experiences required to do the work. 

The most commonly repeated answer to this challenge has been skills-based hiring, but it’s not the right answer. It’s an easy answer to a really complex problem. The inherent flaw is that it assumes words have universal meanings across every job. That’s not true. Collaborative means one thing to a dock team and another to a remote team. Context of the experience is what defines if someone is qualified for a job, not just saying you have a skill. No software tool or algorithm can provide the internal context of your specific organization (yet).

To circumvent this lack of intel, some companies are experimenting with tracking technologies that measure daily employee activities. Early reports from employees can be summarized with one word: creepy. Generally, I get why organizations want to do this, but it frequently triggers deep employee mistrust. Workers have been told for years that AI is coming to replace them and surveillance tools are usually interpreted as a step toward automated layoffs.

How To Optimize Job Posts For AI Search Engines and Candidate-Matching Agents: A Practical Action Plan

The right answer isn’t a technology plug in or complicated prompt. Writing job posts for an AI future means evolving into a new language of work that has to be understood by a human and a machine enough to make a decision. As an industry, we’re nowhere near that part, so I’ve spent the last 3 weeks researching and writing a 1 hour presentation for HRCI. I hope to share it with more of you in 2027 at your monthly meeting or state conference (book a meeting here). 

Here’s what I think would actually work to make your jobs more findable when people are using AI search engines and candidate-matching agents. 

• Optimize your job titles. If you listen to nothing else I recommend, just do a quick search. Your job title and the words “resume example.” If you would not hire any of the people that pop up, you need to change your job title. More on that here.
• Rewrite mandatory requirements from vague skills to experiences. Stop asking for "5+ years of experience" or generic traits. Instead, define the exact experiences that qualify someone. For example, "used JavaScript to build a website for a construction company with 100+ employees.” Specificity gives both humans and AI search engines clear criteria to evaluate.
• Include a clear location, salary, and schedule. These are the 3 most important criteria for job seekers. This will be part of their prompts when they use candidate-matching agents. If you don’t include it, your job is not going to be found by these agents. 
• Write like a person. I love when people ask me how I learned to write with a voice. I used my voice. Voice to text is your BFF because you will naturally use keywords you might not consider using - keywords that job seekers use in their prompts, creating a more specific match when they’re looking to discover a role. 

As candidates adopt more tools and job boards get more expensive, it’s important for every team to optimize their job posts for AI search.