1 in 3 Job Posts Are Fake: The Candidate Harvesting Trap (with Katrina Kibben)

One in four US employers are now using AI in their hiring. And one in three job posts are not real jobs at all.

On this episode of The KeyHire Small Business Podcast, Corey Harlock sits down with keynote speaker, LinkedIn Top Voice, and bestselling author Kat Kibben to unpack why AI recruiting is really a math problem that most companies are getting wrong.

Kat has spent more than 15 years helping organizations like Zoom, Monster, and Olly write better job posts and navigate hiring with confidence, with work featured in the New York Times, NPR, and Forbes. Kat's core message is simple but powerful: AI cannot find what you cannot define. And most businesses, especially small businesses, have a definition problem. When you never clearly outline what you are looking for, every step that follows, from interview questions to scorecards to the first ninety days, is built on a shaky foundation.

Early in the conversation, Corey and Kat find instant common ground on the number one mistake business owners make when they hire. They do not define the role. Instead, they build what Corey calls a wish list, a job description packed with responsibilities no single person could ever handle. Kat explains that a strong job post is not about industry or seniority. It comes down to four categories of information: aptitude and the tools you use, teamwork and who you work with, who you present ideas to, and autonomy or whether you can do the work on your own. Both admit their own hiring record proves the point, with Kat noting that even as a recruiting expert running a recruiting business, only one in four early hires was the right one.

From there, the discussion moves into how AI is reshaping recruiting and hiring, for better and for worse. Kat ran a head to head test of the major large language models, feeding each one identical information and asking for a job post. The results hallucinated a location that did not exist, invented an inaccurate pay range, and exaggerated requirements back into old school formats, because these tools are trained on decades of flawed job postings. Kat warns that AI is a tone thief, and explains why prioritizing speed over results leads to biased screening, rejection loops, and lawsuits.

The episode also takes a fascinating and slightly alarming turn into candidate harvesting, fake job posts, and the rise of fraudulent candidates. Kat walks through activity now being investigated by the FBI and CIA, where foreign agencies mirror the job postings of well known American companies to collect resumes, then send AI backed candidates into interviews at large organizations to steal entire code bases and client lists. Corey shares his own firsthand experience interviewing three separate candidates who all listed the same three companies, every one of which turned out to be a fake website.

What makes this conversation so valuable is how practical it stays. Kat and Corey agree that great hiring is a discipline, not a gut decision, and that most owners get it right only one in three or one in four times. They dig into how to build interview scorecards from a well defined job post, why you have to define what a good answer sounds like before you ever ask the question, why communication and timeline transparency keep top candidates engaged, and how findable job titles help the right people discover your role.

If you are a small business owner, entrepreneur, or leader trying to use AI in your hiring process without falling into expensive traps, this episode is a must listen. Tune in for a smart, data backed conversation on job postings, AI recruiting, candidate screening, and why defining what you want is still the most important hiring decision you will ever make. Stop grinding, start growing.

KEY TAKEAWAYS

AI cannot find what you cannot define: the job post sits at the bottom of everything, and if you get it wrong, your interview questions, your screening decisions, and your six month performance benchmarks are all built on nothing.

You are writing a wish list, not a job post: a job post exists so that anyone reading it can say a clear yes or a clear no, and if unqualified people keep applying, the post was never clear enough to let them decide.

Four categories make a job post work: the tools you use and how you use them, who you work with, who you present ideas to, and whether you can do the work autonomously. Not years of experience, not degrees, not detail oriented.

AI is a tone thief trained on bad data: in a head to head test of the major models with identical inputs, one invented a location that did not exist, one added an inaccurate pay range, and all of them reformatted requirements back into the old templates they were trained on.

Speed is not a benchmark of success: time to fill measures things you do not control, and when you anchor people to time, they start rejecting anyone scored below an eight. That checkbox looks like efficiency and can get you sued.

Your job post is an attack surface: one in three postings are not real jobs, foreign agencies mirror well known company postings to harvest resumes, and fraudulent candidates have used deepfake video to get hired and walk out with code bases and client lists.

Define what a good answer sounds like before you ask the question: if you skip that step, you will default to whoever communicates most comfortably, which is bias, not assessment.

Findability is free and almost nobody does it: Google your job title alongside the word resume. If the resumes you would hire do not come up, nobody is going to find your role.

LINKS & RESOURCES

Kat Kibben's company, Three Ears Media: https://threeearsmedia.com/

Connect with Kat Kibben on LinkedIn: https://www.linkedin.com/in/katrinakibben/

Connect with Corey Harlock on LinkedIn: https://www.linkedin.com/in/coreyharlock/

Learn more about KeyHire Solutions: https://www.keyhire.solutions

Subscribe on Apple: https://podcasts.apple.com/us/podcast/the-keyhire-small-business-podcast/id1643962763

Subscribe on Spotify: https://open.spotify.com/show/1FT9oqXSek3jMfiKrZPLQs

EPISODE CHAPTERS

0:00 - Introduction: one in four employers use AI, one in three job posts are fake

1:06 - Housekeeping and today's guest

2:34 - Kat's background and how job posts became the whole problem

4:34 - The number one mistake: you never defined the role

5:10 - What a job post is actually for

6:12 - The four categories every job post needs

7:48 - Hiring is a discipline, and the experience gap owners never close

9:07 - Even a hiring expert gets it wrong three out of four times

9:54 - Why using AI to recruit is a math problem

10:33 - AI cannot find what you cannot define

12:49 - Hiring for current revenue instead of future revenue

14:01 - AI is a tone thief: the battle of the LLMs

15:55 - Speed over results, and the screening checkbox that gets you sued

18:05 - Losing good candidates to your own calendar

19:12 - Publish your timeline before you post the job

21:14 - Why owners freeze on the final decision

22:02 - Define it up front so you know it when you see it

23:11 - Turning the job post into your interview scorecard

24:10 - What happens when you wing the interview

25:09 - Define what a good answer sounds like

26:42 - 37% quit in the first 90 days

27:42 - Using AI on interview questions means going line by line

27:56 - LLMs are not smart, they are word predictors

30:05 - Talk back to the tool

30:34 - The good enough culture that sunk hiring

31:30 - AI optimized resumes and the easy apply problem

32:32 - When 90% of your applicants look great

33:41 - Fake jobs, fake candidates, and the FBI investigation

36:08 - Corey's story: three candidates, three fake companies

37:38 - The deepfake test: turn your head to the left

38:21 - Indeed's agents and the monolithic algorithm problem

40:10 - Rejection loops that follow candidates across companies

41:35 - Actionable advice: spend your time on the job post

42:53 - Findability, and why made up job titles hurt you

43:48 - Results over speed

46:36 - Where to find Kat

47:25 - Corey's closing thoughts on how long hiring actually takes

Transcript

SPEAKER_02 0:00

Did you know that one in four US employers are now integrating AI into their hiring? And shockingly, one in three job posts aren't real. They're not for real jobs. It's this thing called candidate harvesting. We're gonna get into that today.

SPEAKER_00 0:18

Welcome to the Key Hire Small Business Podcast, hosted by Corey Harlock, creator of Key Hire Solutions, where small business owners learn how to build stronger teams, simplify growth, and scale confidently. Got feedback or topic ideas? Send us a text. If you're listening on a handheld device, tap the link in the show notes to message the show. Subscribe or follow on your favorite platform. And follow Key Hire Solutions on social media for more insights and updates between episodes. To learn more about how KeyHire can help scale your business, visit KeyHire.solutions. Connect with Corey on LinkedIn through the link in the show notes. Mention you're a listener, and he'll accept your request. And now, on with the show. Here's your host, Corey Harlock.

SPEAKER_02 1:06

If you haven't had a chance, uh two weeks ago we had a great conversation with Steve Rogelberg about the power of one-on-one meetings. Great conversation. He's got a great book out there. So go back and listen to that conversation if you haven't had a chance. Steve was great, uh, learned a ton. He has such an amazing outlook on how to hold a one-on-one and the power of one-on-ones. I know small business owners, we hate talking to people sometimes, but this actually, his whole thing is you shouldn't even be talking to them. You should be listening to them. So maybe that's the angle you need so you can start scheduling those one-on-ones. Anyway, go back and listen to it. Today we have a super interesting uh conversation. We have Kat Kibben with us, and we are going to be talking about how people are using AI in their in their recruiting and hiring, how they shouldn't be using it, how they should be using it, and how it's affecting what's going on out there in the world. Maybe you need to update your process. So let's we'll get into this. Today, our guest is Kat Kibbon, uh keynote speaker, link uh LinkedIn top voice on hiring and best-selling author with over 15 years of experience helping organizations like Zoom, Monster.com, and Olivitamins write better job posts and navigate recruiting with confidence. Kat's insights have been featured in the New York Times, NPR, and Forbes. And today, Kat is bringing a sharp take on why AI recruiting is a math problem. Companies are getting wrong. Let's welcome Kat to the show.

SPEAKER_01 2:34

Thank you so much for having me.

SPEAKER_02 2:36

Hey, thank you. Did I get your uh I always like to ask, did I get the bio or mailed it.

SPEAKER_01 2:40

Although I always tell people if that's the most interesting part, I have failed you. So we're gonna try to give you a little something a little more interesting than the highlights.

SPEAKER_02 2:48

We're just I'm just getting to know you. I didn't want to get too spicy, you know. So you are, you said this is your your geek, geek out topic, uh, AI hiring. And I know in particular, your niche really is kind of like job posts and and creating job descriptions and job posts and things like that. So just give us a 30,000 overfoot uh overview of you know where your expertise lies and where it came from.

SPEAKER_01 3:16

Yeah, so honestly, if you looked at my resume, you might be a little confused, but it makes sense now because I've served as a managing editor of a blog about recruiting where all I did was talk to recruiters and create case studies about their work. I have set on the marketing.

SPEAKER_02 3:33

I'm sorry you had to do that. Pardon? I'm sorry you had to spend that much time with recruiters.

SPEAKER_01 3:39

No, honestly, as someone who is a former recruiter and also a marketer in this space, I gotta tell you that I think that was the most important education I ever had. Because what it did was expose me to the fact that there's always a million ways to solve a recruiting problem. Um, and really that perspective for the boundaries and controls of that. And I think when I started my own business after being a technical copywriter on an employer brand team where job postings were coming up all the time, it was really obvious to me that this was the beginning of the problem and why we struggle so much in recruiting. And it's because we never define what we're looking for. And now that we've layered AI on top, that problem is even more obvious because we're trying to use technology powered by this really unclear information. And surprise, we're not getting the results we want.

SPEAKER_02 4:34

Shocking, right? And I've talked about this a ton. The number one biggest mistake companies make, especially small business owners, when they start to hire, is they don't clearly define the role. Exactly. It kind of turns into this person A just quit. Okay, what did they used to do? Smack that on a piece of paper, but let's brainstorm and then let's make a wish list. And if they could do this, and what if they could do that? Oh, and what if they could do this? And you end up with what I that's why I hate job descriptions because I call them wish lists. People just, I'm like, this isn't even one person doesn't even do all this stuff. Like it's it's impossible, right?

SPEAKER_01 5:10

Yes. Well, and they never even accomplish the task, right? So I think we write job posts because we're supposed to have them. But if I ask people why do job posts exist, I find them believing it's an administrative task and not the actual purpose, which is number one, to define what we are looking for. And number two, that it's so clear that anyone can read this and say, yes, I'm qualified. Yes, I want to. No, I'm not qualified. No, I don't want to do this.

SPEAKER_02 5:43

Right. You want to get a hell yeah or a hell no? You don't want a bunch of people that go, I I I fit in this box. Let me go kick the ticket.

SPEAKER_01 5:54

Why are all these unqualified people applying to this? And I'm thinking, if you wrote something clear, they could make a clear decision.

SPEAKER_02 6:03

Okay, Kat, we are starting to mind meld here. This is great. I love it. Uh, what else have you learned from talking to all these recruiters and in your past experience?

SPEAKER_01 6:12

I think they have a misperception that job postings should be unique because of their industry, because of the level, that an executive job posting would look wildly different than an entry-level one. And what I've actually found through my research is what makes a good job posting is actually just providing enough information to accurately characterize a job. And ultimately that comes down to kind of four categories of information: the aptitude, like what tools do you use and how do you use them? Teamwork, who do you work with? Because that kind of puts you with at a level within the business, right? Who you present ideas to, right? Do I actually present ideas to executives or am I presenting externally, just to my peers, to no one? Right. And then who you are, my brain just went bink, right? So, and then the last one is autonomy. Can you do it by yourself or do you need someone to watch over you? And if we can really look at all four of those benchmarks, we can produce something great. Unfortunately, the current template that everyone uses, right? That really long-winded about us at the beginning, four paragraphs about our mission and values. And then all of a sudden at the bottom, we bury a little bit of highly collaborative team player joining our group of rock stars to accelerate the future of work and not one specificity to make a decision, right? And so we just have, I think we have a lot of people who have good intentions but don't actually know what they're supposed to do.

SPEAKER_02 7:47

I talk about this a lot on this podcast, how hiring is a discipline. And, you know, if you look at the small business owner, and I've said this a million times, people listening are probably like, here he goes again. But when you start a small business, you hire your neighbor, you hire your neighbor's cousin, you hire your buddy, and we all pitch in and we all do what needs to be done when it needs to be done, whether it's my job title or not, I'm just gonna jump in and help. And eventually, if they do a good job, the business takes off, but that experience level stays here, right? And and they get that gap in the experience and the um capacity of the people that are building the business. And no one starts a business because they say, Oh, I love to recruit, I love to hire, and I love to you know put people on pips. No one says that. They have a passion to sell or they have a passion for the product, but because they have to hire, and this is kind of what we talked about before we came live, before they start to hire, because they have to hire, they think they get good at it. Yes, and they don't measure their success, right? Uh, my stats tell me that a small business owner is they get it right between one and three and one in four times. That person is actually like the person who can drive the business, right?

SPEAKER_01 9:06

And I will say, as a small business owner who is specialized in recruiting and hiring, okay? Everyone, listen very carefully. I my business probably has the same level of success that about 21 in four of the people that I've hired, especially early in my business, were the right people. The other three absolutely were not. And that's really, really hard to come to terms with because I think, like you said, we didn't start a business to be a recruiter. We start a business because we love an idea and we are willing to make sacrifices to have that idea be real. And then we have to figure everything out. And even if you think hiring is your strong suit, it probably isn't. And I say this as literally someone who has casually called themselves an expert.

SPEAKER_02 9:53

Right. Yeah, yeah. That's funny. So, what what intrigued me about your background was you have this concept or theory that using AI to recruit is really a math problem. It is. And I'm I'm really curious to understand what you mean by that. Uh, and the other part of your your um of that is recruiting is a using AI to recruit is a math problem. And most business owners don't know the equation or don't know the formula to to make it work. So I'm interested to know what your theory is behind that and what what that's all about.

SPEAKER_01 10:32

Well, so I have this hypothesis, and I'm actually doing some research right now to validate this hypothesis. But after eight years of helping people write job postings, it was very obvious to me that White Gew said we cannot define what we are looking for. And I want to be really explicit when I say like this is not skills, that like magical thing of detail-oriented and collaboration, because that's never what we're actually looking for.

SPEAKER_02 10:59

And those aren't skills, those are personality traits.

SPEAKER_01 11:03

What we are looking for is experiences you have had that would prepare you for a parallel experience that is very likely to happen inside of my business. And we don't think of it that way as business owners because we were never actually taught that, right? So if you've read job postings, your first instinct is like, that means five years of experience, a college degree, a, and 12 years of this. And I'm like, why is it five, not four? Why is it 12, not two? Why is it a college and not a bachelor's, not a master's? And most people can't answer that question. And here's my point on that is that AI cannot find what you cannot define. And we have a definition problem. And job postings are at the core. And if I were to draw you kind of a picture and the business owner a picture of this, it would show the job posting kind of at the very bottom. And what it would show next is how everything in the experience should build off of that exact moment. Because if you could get that right, every other part of hiring is a lot easier. Tell me how you're gonna define the right interview questions if you don't define what you're looking for.

SPEAKER_02 12:18

How no, we don't, we'll just we're gonna make it up. We're gonna feel our way through the conversation.

SPEAKER_01 12:23

How do I pick who to interview that person if, again, I don't know what I'm interviewing for? And then let's go down the employee experience. Um, how do I do performance grading the first six months to say this was a successful hire if I didn't define the right benchmarks in the first place? It all trickles back. And ultimately, if you want to use AI for hiring and you miss that first step, good luck.

SPEAKER_02 12:49

Yeah, and I agree with you on that. And I've talked a lot about you know, when you define your role, you can't, you can't, you so the the two biggest mistakes small business owners make when they hire are don't define the role and they hire for current revenues versus future revenue. So they bring in a leader who on day one is already maxed out in terms of their capacity and ability to push that business forward. And AI, you can't just say, I need a sales manager, AI, create me a job description. You need to create it, and you need to sit down either with your team or by yourself and do exactly what you said, bullet that out. You can then give it to AI and say, okay, here's my industry, here's what we need, here's my company, here's my website. Now kind of help me round this out. And then you can also say, and help me understand, we're currently at 10 million, I want to grow to 30. What other skills should I be looking for in this individual? But like you said, you have to have it clearly outlined and defined what you're looking for before you start including AI. That's my personal thought.

SPEAKER_01 14:00

And the other element that you need to consider here is that AI is a tone thief. So if you upload something to a job post, and I've tested this across the major platforms Claude, ChatGPT, Gemini, a lot of the ones that I hear mentioned most of the time. I actually did a battle of the LLMs, and I gave every single one the exact same information and I said, write me a job post. And what most business owners don't know is that it hallucinates details. It will add things to your job post without your permission or you even giving them that information. In my case, I put in remote on all three. It added a location to one. It literally added a place that did not exist. It added a pay range that was not accurate. And so it also exaggerated the requirements and it reformatted requirements into the old school ways because all these tools are trained on all the job postings that have already existed. And I think you can I can you and I absolutely agree on the fact that that is not a good data set to be building anything on.

SPEAKER_02 15:11

Correct. Yeah, yeah. And that's really interesting that you actually did the the use case and plugged it all in and saw because I'm I'm a big fan of truth mode when I use AI. So first thing I say, uh truth mode and red tea, uh red team. Like uh give me an idea and then dissect the idea you just gave me to make sure it's you know you're not just giving me stuff here. Um it that's so interesting. How else are you seeing uh business owners using or misusing AI in terms of their job search or their um recruiting talent acquisition, even within the hiring process?

SPEAKER_01 15:54

I think the the overarching mistake is that they are prioritizing speed over results. So it's very obvious to me that we have been anchored into a timeline as a benchmark for success in recruiting because we still talk about time to fill. And I think time to fill is the most bunk metric for your success that has ever existed because you don't control it. We don't control how long it takes the hiring manager to get back to us, we don't control how long it takes the candidate to reply to that email. And all of those variables somehow fall into the recruiter's bucket as a benchmark. And so when AI came on and they started using it, their brains went, how do I do it faster? The consequence is well, now we're seeing it in courts where workday is being sued because of their use of AI and sorting. Because what ends up happening is when you anchor people to time as their benchmark, they do not have behaviors that are inclusive. Let me give a great example. So we've all heard about candidate scoring. That's a real thing where they you enter what you think you want, and then the machine says, this is a match. In some cases, that is simply a keyword count, and in other cases it's a little more intelligent. But let's just say there's a score. Most tools will give you a quick checkbox where it says, Do you only want to see eights, nines, and tens? And any recruiter who's overwhelmed by applicants, that's gonna be their first move, right? Remove anybody, reject anyone with a seven or lower. That can get you sued. But on the surface, it is a time-saving strategy. Right. And and ultimately, AI should only be used on systems that are ex that can be executed well by any person that are very easy to understand. And like screening is just not one of those things. It is our responsibility. It's a recruiter superpower to be able to read between the lines and understand. Yes, I think they can do it. Yes, I think they can do it here.

SPEAKER_02 18:05

That's that's a good point. And the the the one of the biggest battles business owners face is time. You know, it's their time, the training, being uh we talked about recruiting being a discipline. And the third thing is being able to um prioritize in your world, how often do you say, hey, I have this candidate, but we we need to talk to them. They're interviewing with two other companies. Yeah, I don't have time this week. Can we push it to next week? We can, but good people go quickly. And another concept I've been playing with a lot lately is if you solve for time, not meaning just going fast, but if you can solve for time and prioritization, you'll solve for quality of candidate as well, because the good people go quick. And if you keep kicking the can and pushing people out before you can interview them and get to that second interview or get them on site, they're going to get job offers and be off the market. And then you're left with kind of a uh what I it's not the right thing to say, but you're kind of left with the best of the worst people that are still out there looking.

SPEAKER_01 19:12

You know, I would go so far to say that I actually think it's a communication problem. It's not that the best people are always on the clock, it's that we never communicate the timeline. So, in that example, I think the bridge is when you post the job, you literally say, we will close this for applications on this date. We will complete the first round of screening by this date. We will complete interviews by this date and have an offer in your hand by this date. And I I've seen enough research on candidate expectations, why they drop out of a process to believe that that would have a very significant impact on your ability to keep a great candidate in a pipeline, no matter how competitive the other offers are.

SPEAKER_02 19:59

Yeah. And that's you just touched on the biggest candidate complaint is communication. They never know what's going on, they never hear back, they never know why. They could be silenced for two weeks and then they get a call. Oh, we want to bring you in. You're like, oh, I thought it was eliminated. And that's something we're really big on is always giving that agenda. Hey, here's what's happening next, here's what the timeline looks like, here's how we're gonna be moving forward with this. And people are always like, oh, wow, thanks. I appreciate I appreciate you letting me know. Uh but um, as I alluded to in the intro, you know, Steve Rogelberg's uh talk we had two weeks ago, he was talking about you know trying to get owners to communicate with the people they have, let alone the people they're trying to hire and interview, and then that becomes that priority issue. I just I gotta take this lunch, or I gotta make sale. Oh, we had an opportunity to go on a sales meeting. I gotta, I gotta push this out, whatever it is. It's it's just a challenge. And it's not a right or a wrong thing. They're one person trying to run their company and um they're trying to build that professional sales team, and they might still be kind of in that kind of experience capacity gap where they're just kind of at war every day trying to trying to make the business run. It's it's a real struggle they face.

SPEAKER_01 21:13

And where I see that like secondary mistake too, and I'd be curious what you see as well, is like, I think as much as there's a little bit of the I want the perfect apple in the corporate level, there's a whole different dimension of feeling like you need to get it right when it's your business. And it leads to a lot of delays. I know the reason why a lot of my clients who hire me who run small businesses and they hire me to maybe help them write that job post and really distill what they're looking for. What ends up slowing them down is at the end, they have all these people. And like you said, it's like they struggle to make that final decision because they're like, is this the best? Did I do the best? Did I define this correctly? This is a big risk, and we don't know how to navigate. That kind of uncertainty.

SPEAKER_02 22:02

You're right. It's um they want to get it, you're, I think you nailed it. They want to get it right. And it's always that, well, we we need this. Can we see someone else? We just want to contrast and compare. And uh what Key Hire does is we're always like, no, we're gonna define exactly what we're looking for up front so we will know it when we see it. Because if we get someone who's really good and we got to put them on, you know, hold put them on hold for a week so we can get you to interview someone else, and then they go, and now we're left with someone who's wasn't as strong because we wanted to compare and contrast, that's that efficiency piece, kind of that um the speed solves for the quality piece. Like if we can define, and I'm sure that's what you're doing with your job postings. Like, this is the two-dimensional snapshot of the exact right person. And then when we get people who can, you know, we call them 80%ers, if they check 80% of the box, they're worth talking to. If you can find someone who has 80% of what you're looking for, we need to carve out some time to have a chat with those folks, that that person, because they're they're a real strong fit.

SPEAKER_01 23:10

And that's this kind of element of that recruiting formula again, is like if we spend high quality time defining the requirements we are looking for out of the front, then I'm even telling people, okay, take that exact job. You're turning that into your interview scorecard, you're turning that into your interview questions. And then we know we are not screening for something we don't need because I think that's why people come back and do the whole, well, I'd just love to see a couple more, is when they get something in their head that says, Well, what about this? And I always tell them, and I'll push back to say, if you can't tell me how this will make them better in six months, like in six months, I want you to come back to me and say, This was the best hire of my life. This person changed everything for me. And if what you are looking for does not directly impact their ability to do that six-month benchmark thing that's gonna make you feel so good, you need to drop it.

SPEAKER_02 24:09

Yeah. Yeah, I agree. And going back to that interview piece, which is really important that you highlighted, if you wing the interview, you might have two really closely matched candidates you're talking to, but you give them each a completely different line of questioning. And then how are we supposed to come back? There's no baseline been set. We're not comparing apples to apples anymore. We're comparing apples and oranges because we didn't ask the same questions. We didn't run them through the same process, and those questions weren't based on the requirements of the of the job description or the job ad. So, and this is, I think you you nailed it when you said this is where they're like, oh, I'm not sure. And that's why it's important to do the work up front and create that process and make sure you you are asking the right questions and the same questions, not exactly verbatim, but you're you're hitting the same points throughout. Um, so you can compare and contrast and make sure you're getting the best person.

SPEAKER_01 25:09

When I train teams, I even have them go the far one step farther than that. So they have the question they're going to ask, and they can compare the answer that they got from each candidate on that question. But the first thing they have to do before they ever ask is tell me what a good answer sounds like. What are the benchmarks you're listening for for a good answer? Because I think that's where we miss out is we start to, and and this is bias that's buried deep inside of our brains, it's nobody's fault. But we start to lean into the person who has better communication style. We lean into the person who has a higher level of intellect or uses keywords and vocabulary that sounds comfortable to us. It is very natural that if you do not define what good means, you are going to lean into your bias. And the more that we are very distinct about the benchmarks, the more that we have the opportunity to actually hire the right person for the job. And I keep going back to it, I don't mean to harp on it, but that's what I mean by the formula. It's like everything rolls back to that very first definition. And if you get it wrong, I don't give a shit if you're using AI or not, pardon my language. It doesn't matter because you will not get the right answer at the end. It's like a calculator. If you enter the wrong data, you can't get it to know what you meant.

SPEAKER_02 26:31

And that's where the one out of three, one out of four comes because if you're just hiring on gut, it's your odds drop. Your odds drop dramatically. Go ahead. What do you think?

SPEAKER_01 26:42

Well, and then alternatively, look at the other side of the coin. 37% of people quit in the first 90 days because the job does not align with their expectations.

SPEAKER_02 26:52

Yeah. Right. Yeah. And and beyond that, not just the job, but what they were told in the interview. It could be, and I I've talked about this before, where a business owner might say, Oh, we got a great culture, you're going to love it here. And then on the first day they get pulled aside and they're like, Oh, watch it for so-and-so and so-and-so. There's snakes, and they're like, Oh, hold on. That's not what I was told in the interview. And what the business owner meant to say was, we want to build a great culture and we see you being a part of that, but that feels scarier. So they just say, Oh, you're going to love our culture. We oversell and underdeliver. Yeah, that's that's right. And and that's that's a great point. What are your thoughts on? Okay, we've we've defined our needs, and you talked about, hey, now you have these needs clearly defined. You've done the work using AI to create questions and scorecards based on that job description.

SPEAKER_01 27:42

I love that. And I'm gonna tell people again, you gotta go line by line.

SPEAKER_02 27:47

Have you taken improv courses? Pardon? Did you have you taken improv courses? Yes, and right?

SPEAKER_01 27:56

So I think the issue right now is that we think AI should be able to understand all of the intricacies and we kind of give it too big of a data set. And what ends up happening is LLMs are not smart, right? I know you know this. I'm just when you say LLM, what do you refer to? Large language learning model, right? So the Chat GPT, Gemini, Claude, we think they're smart. People talk about these like they are beings that have intelligence levels, they are word predictors. They predict that if you use this word, the next word will be this, and the next question will be this, and the next piece will be that. That's literally how those formulas are written, and they move faster than our brains might predict the next thing. And what ends up happening is again, it just has a crap data set. The data set is the whole world, not what you need. And so if you're in that process of going from requirement to interview question, you have to go line by line. This is not upload my scorecard, add questions to fields. Yeah, and just let people save and we're good to go.

SPEAKER_02 29:00

Yeah.

SPEAKER_01 29:01

You really got to talk back to whatever tool you're using, especially in that environment, right? So I give it, and you know what it's gonna do first. I can almost bet, I haven't tested it yet, but now I'm gonna have to, is you're gonna give it that and it's gonna give you all situational interview questions. Well, situational interview questions aren't always the most helpful questions. So you're gonna need to go back, and I might say things like, I don't like that. These are too complicated, too detailed, too specific. Let's give me something more open-ended so that people can't just say yes or no or give repeat the same experience over and over again and nurture the conversation with it. Again, I think too many people are treating AI like search where they could enter a query and use the first result. Like I'm seeing too much of that behavior and not enough curiosity and contrast to say that's actually not good.

SPEAKER_02 30:04

Yeah, and one of my secret weapons, well, I think it is when I use any AI anymore, is I will give it instructions and I will say, before you take any action, please ask me any questions you have about this. And I will take I will do that about four or five times before I get them to print or create or give me any data back, just so we're aligned, at least as aligned as we can be. And then it comes back and then you got to play with it. Okay, what about this? And use your brain.

SPEAKER_01 30:34

If it's if you don't like it, don't use it. If it sounds good but might be misinterpreted, don't use it. Sometimes all of it is junk. Yesterday, I wanted to test a new platform, so I uploaded a hiring manager intake conversation and I used it and I tried to write the job post by uploading this transcript, asking some questions, giving it a really, really detailed prompt and running it. After 20 minutes of banging my hand against the wall, I deleted that chat and I went and did it myself because it was bad. Right. We have a good enough culture in recruiting that has existed since as long as I've been in it, 17 years, where people say this is just administrative. It doesn't matter. It's good enough. And I really feel like I hope AI is what breaks that pattern because that good enough culture has sunk us in the world of hiring.

SPEAKER_02 31:29

Well, and especially now, and this is something I wanted to ask you because off the off the lead, I I threw in the a stat from from some of your LinkedIn posts that one in three job job posts aren't real, the whole candidate harvesting thing. And then I've talked a lot about how people there there are people online telling you how to use AI to make an irresistible resume. So you take your job posting and you give your resume and say, find the keywords and make my my resume, you know, match this job posting. And especially for a small business owner, before they would post a job, I mean, you're gonna get lots of applicants, especially now, because people are looking for work. And I've always said the good thing about job boards is it makes it really easy for people to apply. The bad thing about job boards is it makes it really easy for people to apply.

SPEAKER_01 32:25

I I may have told the founder of ZipRecruiter that that easy apply was cruel before. I'm with you. I'm with you. Come on, please.

SPEAKER_02 32:32

Yeah. So so before you might get, say, we'll use easy numbers, 100 applicants, but you could sort them very quickly by zip code and you know, who had the right experience. But now with AI, you know, you you could sort them and get to that that top 10 of people like, okay, they look like they're in the ballpark and they're worth, we're they're worth at least, you know, either calling for a pre-screen or giving them a really thorough read and kind of force ranking them. But now you could be in a situation where 90% of the people who applied look like they're really good, and 10% you can eliminate easily. And and that's that's not a bonus. Like that's just more. We talked about how business owners don't have the time to run like to do the background work and run the hiring process on time, and now they're gonna have more stuff to sort through. It's gonna be harder for them to sort through because people are using AI to create a resume that is keyword rich and their ATS goes, oh yeah, this person looks great. This person looks great.

SPEAKER_01 33:41

Yeah. Because what they call intelligence is basic matching, right? And I think what it's really showing from my perspective is that the resume, again, was never really prepared for this problem. It's the same kind of reflective point because the resume is kind of the reflection of the job post. The job post is a reflection of the ideal resume, if we were doing it right. But again, I just feel like the format of the resume restricts us to a certain point, especially when you're getting into high, highly qualified roles, meaning you have to have some specific experience in order to do this. Because the longer the content, the easier it is for our brains to just do keyword matching and believe that there's something positive on the end. And now you have to layer on the fact that there are spam jobs, right? So people are posting fake jobs. There are fake candidates who are using that to hack into other people's business. And I people, when I first said this, people are like, you're a liar. So hold on, walk us through. So people are posting jobs. So what's happening right now, and this is actually being investigated by the FBI and the CIA, is that specifically where they're tracking this to is foreign agencies. So North Korean cyber, Russian cyber attackers. They create job postings that mirror the job postings of very well-known American companies. They are also creating resumes. And they are applying to these highly technical roles at well-known organizations. And in the attraction example, they are collecting resumes because they can use that to train their own tools and to start to identify trends in the American workforce, which isn't that bad. Like, I'm sure there's something worse they could do with all that. They could probably steal a lot of people's identities, but that hasn't been well tracked. This is where things get wild. So on the candidate side, they are sending people to interviews. These people are getting hired and going into these very large organizations and stealing entire code bases. They're stealing the entire client list. They are taking all of that back and behind firewalls that exist in come in countries like North Korea and recreating that exact organization there as a for-profit organization with another name.

SPEAKER_02 36:08

It's crazy. So I had nothing like that, but um, I've told the story before, but I was working with the um a SaaS company, and they were looking for an AW, someone to kind of move everything uh from their on-prem to AWS. And I was having a heck of a time finding that person. And I found two or three, and I was like, oh wow, this looks great. So I interviewed the first one, and they showed up and their camera was off. And I was like, hey, you can turn your camera on. Oh, I turned on, and they were kind of in this dodgy looking place. And um uh they said they were located in DC. Okay. Uh it was a remote role, and they did not have what I was looking for. But I'm like, okay, so that's that's a pass. Then the next one, same thing happened. Third one, same thing happened. And so I looked at all three of them, and they all had the same three companies on their LinkedIn profile. And when I went to those companies, they were all fake websites. They still had like the the the pictures were all stock photos, and some of the the sections still had like the gobbledygook that comes when you just like add a section, right? Uh and so I actually wrote a note to LinkedIn. I'm like, you guys got to be careful. There are there are people, these these people were not even in the US. They were sitting somewhere in Africa and they were doing exactly that. They were trying to get hired saying we live in DC and they were trying to get into these companies and steal information.

SPEAKER_01 37:38

The craziest. So when the first time I heard about it, the way that they caught them was they asked them to look to the left. Hey, can you just look to your left? And the the little like bot or kind of it's like a fake background, but for your face.

SPEAKER_02 37:54

Like what I have right now, yeah.

SPEAKER_01 37:57

So this it creates this illusion of a person on the front, yeah, and they the it wasn't coded to be able to turn its head. And one of the recruiters caught it, just like, you know, you're staring too long, and you're like, and out of curiosity, you said, can you turn your head to the left? And it ended up shutting down.

SPEAKER_02 38:21

Crazy. It's it's and then um do you have a few more minutes? Because now I want to touch on kind of what's going on with Indeed and what the Stanford article that came out. Like Indeed, they have their AI agents. So post your job, and our agents will send you the best candidates, but those agents are being trained on Indeed's info and not your company info, and it's not, you're not getting the best people, and it's not working the way it is either. What's your research that you found on those type of agents?

SPEAKER_01 38:52

So when everyone thought they were building AI, they built these monolithic algorithms, which is related to the other research you and I talked about from Stanford prior to uh starting the recording. So monolithic means there is one algorithm and this company uses it for everyone. And in the past, we had white labeling, which meant that everybody kind of had their own box, right? This is my deployment. Right.

SPEAKER_02 39:20

And it sat and it sat on your ATS and it was kind of siloed on your ATS.

SPEAKER_01 39:25

Yours. If you made changes, it did not impact anyone else inside of their system, nothing else. That's not how algorithms work. And they didn't build an algorithm that was cloned for each person that would evolve with that box of information. They built an algorithm that is influenced by all of the boxes of information. And ultimately, because it is influenced by everyone, that sounds right, right? Like on the surface, you're like, I get something better because I'm getting trained by thousands of companies, not just my users. But it's really, really wrong because you are trained by all the bad behaviors and now you can't control any of them. And the consequence is bias and ultimately bias that may get you sued.

SPEAKER_02 40:10

Right. And that Stanford report said, I forget, I know HigherView is one of them. I forget the other assessment that sat on top, but they they weren't associated with individual ATS. They sat on top and went across all of the companies that were using them, which were hundreds of major companies. And what could have happened and was happening, they said about 10% of the people got stuck in this constant loop of being, you know, rejection, these rejection loops, because they would apply to company A for a job and it would they would get rejected based on the results, and then they would go to company B, C, D, and regardless of the requirements of the job, because they got rejected one time, they just rejected them from all.

SPEAKER_01 40:53

And I want to clarify that it's not them, it's the again, it's the data, right? So my AI brain looked at this and said no. And so when my AI brain sees this same pattern of information again, it's gonna say no again because that's how it's trained. Remember when we talked about LLMs and they're just picking the next word? That's what they're doing with these resumes. This is a no, no, no, no, and it's replicating across all of these, and that's the danger of buying into these big systems and not asking a lot of questions about how other data influences yours.

SPEAKER_02 41:34

Well, Kat, this has been great. So we've we've gone on a good rant, but we always like to leave people with some actionable advice. So if I'm a small business owner listening to this and I'm thinking, oh my god, I don't even want to go near AI. This is scaring the crap out of me. Give them kind of some, or they're using AI and they're thinking, I'm I did what she said not to do. Um, what would you say?

SPEAKER_01 42:01

What would advise would you give? So my first piece of advice is to spend the most time on the job post itself, defining the requirements and really avoiding the traditional way of describing work. So I don't want to see years of experience, I don't want to see degrees. I want to see what you learned. You had your four categories. What were your four categories you said they should have great the tools they use, autonomy. Did they do it by themselves or with someone else? Who do they work with? Who do they present ideas to? Distill it down to that. I love that. I like that a lot. I think that's thank you. That's one of my hypotheses of my research, is that I actually think if we formatted it that way, we would get more predictive results, but I'll tell you about that more some other time.

SPEAKER_02 42:48

And yeah, we're gonna we're gonna have to, I'm thinking we might have we got I got so much I want to tell you.

SPEAKER_01 42:52

I love it. I told you we were gonna nerd out. Okay. So you need to spend more time on the job post itself. The second thing, and we did not touch on this today, but I think this is really important for small business owners, is to think about the findability of a job. Job titles are made up. Okay. So anytime you pick a job title, I want you to Google the job title and the word resume and make sure that resumes that you would hire pop up. Job postings similar to your work pop up. Because the easiest way to not get a good applicant is for them to not find it.

SPEAKER_02 43:27

Right. And people, especially the small business owners, they they want to come up with these unique and different titles because they have this role that they think is very unique and different. So we got to give it this funky name. But that's that's a great point. I never considered that. That it people aren't searching for it.

SPEAKER_01 43:47

Findability matters. Yeah, that's great. Okay. I really think this last piece is something that we touched on, but I'm just gonna say it again results over speed. Speed. Is not a benchmark of success, no matter how much pressure you feel under. Define the result and work backwards from there. Do not buy, use, sell, trade, any AI that tries to sell you on speed because if that's their benchmark of success, they've gotten something really wrong and they will rush you into the wrong result.

SPEAKER_02 44:21

And when you say speed using AI, meaning they say you'll hire faster.

SPEAKER_01 44:26

Save four days automating your interview process.

SPEAKER_02 44:30

Yeah, that's that's complete bullshit. Um, it takes what it takes, right? And the wrong person will appear, or the right person will appear when they appear. Uh that AI can't make the right person have a bad day at work and go look at a job board tonight. Right. You know, that that's gotta happen, right? And and you have to, that's how you get the warm body syndrome. That's how you hire people and you're like, oh, they're they're not, I got pain, they're close enough, let's hire them. Now I'm disappointed. I just did a post uh a little while back, but it's like you gotta sometimes you just gotta deal with the pain, and and when you get it right, it's short-term pain, and then you'll be so much happier with that result. If you go for the short-term relief, you're gonna have long-term pain, and it's gonna really it's going, it's going to repeat because if you're one in three, one and four effective and you're hiring, you're gonna do it, you might do it again, you might miss it a third time before you get the right person. And if you're not that person could be there three, six, nine, twelve months before you go, this isn't working out. And so that cycle could be months and years long before you finally get it right. And how much effect is that having on your business and your ability to grow and scale and keep good people and reduce turnover and all those buzzy things we like to talk about. And that's why I I agree with that so much. Like sometimes you just have to feel the pain. It but if you've defined what you need and you're confident in what you're looking for, it will be worth it.

SPEAKER_01 46:08

One in three works for a home run derby, all right? It does not work for hiring. And if you really want to save time, I think spend time where it needs to be spent. You're gonna spend time defining what you're looking for in the job post. You are going to be clear about how we benchmark and what a good answer is, and you're gonna prioritize interviews. That will save you a lot more time than any of these tools have saved a lot of people.

SPEAKER_02 46:36

Yeah, all great advice. Kat, if people are listening to this and they're like, hey, I want to know more, I want to know more about what Kat's up to, how can they get it?

SPEAKER_01 46:44

Yeah, so I'm the only Katrina Kibbon in the world. So if you spell my name right, you will find me. If you spell it wrong, you'll find a priest from the U.S. House of Representatives, and you will know that you are in the wrong place. I also have a company, it is named Three, like the number, ears, like the ones on your head, media after two dogs with four ears. But I will tell you that story if you reach out to me directly.

SPEAKER_02 47:06

Very cool. And uh, Matt, do we have a book? There it is, the bounce back factor available on Amazon. So cool. Thanks so much. This is this has been great. And uh, I'm gonna reach out to you because I want to know what your thoughts are on assessments.

SPEAKER_01 47:21

Up next. Thank you for having me.

SPEAKER_02 47:24

There you go. Kat Kibben, that was that was great. Some really, really good insight backed by data, which is incredible. Um, really enjoyed the conversation, and I think she really uh Kat really dialed in on that speed piece. I know I get asked all the time, how long do you think this is going to take to fill? And I always say, I don't know. The best case scenario is six weeks. It's gonna take a week to put everything together, a week to get into the market. Then if I find someone, it takes a two-week interview process. Then they need to give two weeks notice. So if we hit a home run and come out of the gate hot, we might have someone in seat in six weeks. That's the fastest this is gonna happen. But it's usually more than that because it it takes some dialing in and some work to find that right person. And you gotta screen a lot of resumes and talk to a lot of people and uh before you're gonna get to that right one. So uh sage words for sure. All right, that's enough for me. My name's Corey Harlock, I've been your host, and until next time, stop grinding, start growing.