[Podcast] Do my teams hate me? | Navigating the AI Revolution in Recruiting with Aaron Elder, CEO at Crelate

Aaron Elder [00:00:00]:
The true power is AI’s ability to see connections between things that you couldn’t see, because I was able to understand all the data. And if it can surface those things to you, so then you can then take action. That’s probably the best use of AI, not just send out more emails.

Kortney Harmon [00:00:16]:
Hi, I’m Kortney Harmon, director of industry relations at Crelate. Over the past decade, I’ve trained thousands of frontline recruiters and have worked with hundreds of business owners and executives to help their firms and agencies grow. This is the full desk experience original podcast where we will be talking about growth blockers across your people, processes, and technologies. Welcome to another episode of the full desk experience.

Kortney Harmon [00:00:52]:
Do my things hate me for not implementing AI in my recruiting firm yet I’m the president of a recruiting firm, and AI and automation tools for hiring and staffing have really taken off in the last year. A lot of my competitors seem to be embracing these new technologies wholeheartedly. However, I’ll admit I’ve been a bit hesitant to jump on the AI bandwagon. There’s just so much out there. It all seems really overwhelming trying to evaluate and implement different AI tools. Part of me worries that by not automating more processes, we’re really going to fall behind our competitors who are able to operate more efficiently with AI. But I also have concerns about the over reliance on AI. You know, removing that human element that I believe is crucial for making great hires.

Kortney Harmon [00:01:37]:
I want my firm to stay cutting edge, but not at the expense of sacrificing our personalized approach. Now, at the same time, I know AI could potentially help us screen more candidates, streamline outreach, and make data driven hiring decisions. I’m just not sure if we have the resources and knowledge in house to properly evaluate and implement AI safely. So do my teams hate me for dragging my feet on AI up to this point? Have I missed the boat on adopting these technologies? Or is it reasonable to be a bit cautious with how we approach AI in our hiring process? I want to do what’s best for my team and my firm long term, but I don’t want to stubbornly cling to old methods. If AI is really the future, what do you think now? Erin, there’s no better person to answer this question than you. Who has is the technology space, has the mind, has the vision. So with this letter that we received, how do you interpret this? What are your thoughts on this? And have they missed the bandwagon?

Aaron Elder [00:02:35]:
Well, thank you for the question. It’s funny, my first thought is that I would be willing to bet that your teams are using AI whether you want them to or not. Already recruiters can just hop over to chat GBT and paste and stuff and summarize a document and generate a press release and generate a blog post and generate an email and copy and paste it back into your ets. Today, probably already happening. It might be interesting, actually. They might be moving faster with AI than you’re actually wanting them to, if you will. And you might be mindful of the quality of the work that’s coming out of those things. In some businesses where the team was using AI and just literally copy and pasting it directly and then not editing it, and they were sending out stuff that didn’t make any sense and then that was embarrassing for the company.

Aaron Elder [00:03:17]:
But from their head they’re moving fast and they’re getting AI right, but they weren’t thinking about it and audio editing it before it went out. I think it also depends on the size of the business and where you’re at. From one angle, AI actually presents a risk to the business. And if you don’t already have an AI policy, you might want to have one of those because recruiters might already be using it in ways that you’re not expecting. They might be copying your candidate database and uploading it into chat GPT today. I don’t know what the liability outcomes of that is, and I don’t know what of your proprietary data you want in these engines. And so you might want to have a policy around what is an appropriate use of AI in your business is one thing to consider, because I think the adoption is happening faster than businesses even realize. And so I’m actually building on the prior point here.

Aaron Elder [00:04:00]:
My first thought was, what are you trying to achieve? What are the actual problems of the business today, and how can AI play a role in that? I think if you’re not thinking about AI playing a role in that, then you’re definitely missing the mark. Is it a magic bullet as much solid problems? Definitely not. And do you need to be intentional and thoughtful about it? Yes. Right. Do you need to be so slow, but it doesn’t matter or you can’t get value out of it? The answer is no. I think what I’m seeing, though, and this is across industries, is that if you’re intending to use AI for faster, higher quality spamming, I just want to send more personalized sounding things to more people faster. You’re pretty late to that game that’s been going on for a long time, and you’re not going to get any better response than anybody else who’s doing that is, in fact, from people I’ve talked to and what we’re seeing. Response rates are dropping through the floor because everyone’s doing it.

Aaron Elder [00:04:53]:
Candidates, prospects, they are just inundated with an onslaught of AI content. It’s a funny thing, Kortney, I’ve told you already, but, like, I get 60 emails every morning of personalized AI things telling me that. And my response was to go to my head of it and say, please go buy a better AI to filter out more of these things. Right. It’s just too much. It’s too much. And so if you’re looking to use AI to solve that, you’re late to the game. And what you should be thinking about is how are you going to differentiate in that circle of automated mediocrity? Because that’s how you’re gonna get in front of someone.

Aaron Elder [00:05:28]:
I think the same things still apply, like getting to be top of mind with the candidate at the right time, where they are. I don’t know. It goes back to the old thing.

Kortney Harmon [00:05:37]:
No, I think that’s great. And so are you saying it’s getting to the human touch faster? It’s like in aid of having that human touch, do you think it’s in conjunction with.

Aaron Elder [00:05:46]:
I think what I’m getting at is that is, I think this is one of those big solid and it has to be a smart. And, like, do your teams hate you for not doing that? Maybe, but it’s the same way that you might hate your personal trainer for telling you that you need to eat less and work out more. Right? Like, you’re just telling them news that they don’t want to hear. You’re telling them, like, listen, AI is a valuable tool. We can use it. Here’s how we’re going to use it. It’s going to help us in these areas. It’s not appropriate in these areas.

Aaron Elder [00:06:14]:
But at the end of the day, you’ve got to have a way of differentiating yourself amongst all the other noise out there to get in front of your candidates and be top of mind at the right time, right place, et cetera. In fact, it’s funny, I actually feel, and I don’t know this is true, but, like, I think the next big thing is going to be, I think you might probably heard this, is that vinyl records outsold CDs for the first time, I think, last year. Right. It’s funny. It’s because vinyl record sales have been increasing while CD sales have been dropping off a cliff. What I say by that is that I think there’s going to be a world of offline or analog type things in response to the deluge of stuff. And it’s funny, I’m starting to see some inklings of this, whereas I get the 60 emails a day, and that’s not an exaggeration, it’s 60 inbound random things trying to sell me and my company on something. But then every now and then I get a package in the mail and it’s a mug from some company I haven’t heard of.

Aaron Elder [00:07:07]:
And so now, well now I got a mug and I got an email from the guy. And when I see the phone number pop up, I’ll probably answer, I’ll talk to, I’ll see what’s going on. Right. But it’s this creative, different idea to get me to get their name top of mind for me. And I think the same thing applies there. And I don’t think AI is going to send a mug for you. I mean it can be part of an automated process, sure, at scale, but how are you getting different? I guess is my thought. So that’s one big thought.

Kortney Harmon [00:07:30]:
Oh, that’s great.

Aaron Elder [00:07:31]:
It’s funny too. You had a lot of things in that email. One of the things was AI can help me screen candidates. I think businesses need to be very intentional about what that means because you might be one day having to defend those decisions and I don’t know if you could defend them. Like for example, let’s assume that the AI gets proven to be racially biased, for example, and it comes out next year. Who’s liable for all the candidates that you screened out over the last year based on this AI engine that you use? Bias? I think the answer is probably you is what will all we come down to it? So just be very careful about that aspect of it. I prefer to look at AI as surfacing ideas, surfacing next steps, summarizing things, aggregating things, surfacing the right information at the right time, saving menial tasks. And actually the true power is AI’s ability to see connections between things that you couldn’t see because I was able to understand all the data.

Aaron Elder [00:08:27]:
And if it can surface those things to you, so then you can then take action. That’s probably the best use of AI, not just send out more emails. And so the question is what tasks today can be easily replaced by AI? And then more importantly, what tasks tomorrow can be replaced? Because the speed at which these things are improving is pretty drastic. Let me give you an example. You might be paying either an onshore and offshore resource to tag your database for you, go through and clean it up and tag it. That might be a great example of where an AI could apply some rules for you and go tag these things and produce the outcome that you wanted. You would have to make an investment upfront in finding the right AI engine and to tune it appropriately or the tool to do it and then to set it up and they get it to run and make sure it didn’t screw things up. Right.

Aaron Elder [00:09:13]:
But it’s no different than what you have to go teach a person to do. Go do. Here’s the tags that we use. Here’s what they mean. Go find these things. Go apply the tags. That seems a pretty savings that would come from that. I don’t know if you’ve got other ones Kortney, you can share them.

Aaron Elder [00:09:27]:
I think response generation, summarizing resumes, these are all sort of marginal time savers which can then increase throughput. But I don’t think it’s going to allow you to replace or get rid of that recruiter or that sorcerer. At least not today.

Kortney Harmon [00:09:40]:
I agree with you and I think it’s just a matter of time of just helping us do more. We look at our activities today and we judge so many people with that KPI hamster wheel of you have to have 22 calls in this call block. You have to have 117 emails, you have to have so many text messages. I think that’s going to change. And the idea of the KPI hamster wheel and the do more mentality is going to help us get to conversations quicker. But I don’t want it to get us to a point where it’s just going to help those numbers escalate. And those numbers mean nothing because we’re treading water.

Aaron Elder [00:10:09]:
You need to get more into outcome. Right. Outcomes is what matters more. I love that the KPI hamster wheel to totally freaking get it. It’s a balance though, right? I mean that’s the analogy of you just got to do a certain number of reps in the gym if you want to build muscle mass, right? Like you know, as an example. Right. So there is some aspect of it. You just got to make a certain number of calls.

Aaron Elder [00:10:25]:
But also it’s the outcomes that those calls are leading to. And I think what you’ll find is that oftentimes as the frequency of automation increases on the outbound, the quality and the outcomes decreases. Unless youre a first mover with an approach or a technique, other cost savings, I think there is a value in consolidation of tech people are looking to have more on one platform. That is a cost savings. The problem is that AI right now, at least on the top end or even in the middle end, is very expensive to run. I think what youre seeing today is the affordable implementations of that for your business businesses that are truly going out and buying AI hundreds of thousands of dollars to go tune these things for their data and their scenario, to create their secret understandings of what’s going on. Right. And to create a differentiation.

Aaron Elder [00:11:09]:
If you’re just implementing someone else’s tool for an outbound campaign, you’re pretty far down on the food chain, I guess is my thought on that. What are the other areas? Coordinating ideating. So, for example, certain bank CEO’s say like we’re going to get rid of 25,000 jobs over the next two years because of AI. How are they doing that calculus? They’re doing that calculus by saying certain positions can straight up be eliminated within the organization through 100% automation, or that the remaining staff, if you’ve got 100,000 people and you want to lay off 25,000, then the remaining 75,000 have to get 30% more efficient in order to do, you know, et cetera. Let’s say it’s the former. What roles in a recruiting firm can straight up be eliminated by AI?

Kortney Harmon [00:11:56]:
It’s interesting because Maurice Fuller and I just had this discussion and I think where he’s like in his thought process, he was literally even saying like, manager level positions are going to be replaced because this data is going to be able to say, oh, well, I know this person’s output is x. I know where I need to get them. Here’s the recommendation of like the personal trainer at the gym. Don’t get me wrong, I feel like that could be the case. I don’t necessarily think that’s in the next six months to a year, because there’s so much fine tuning, there’s so much that actually has to be further developed and we might miss the mark. Go ahead.

Aaron Elder [00:12:29]:
Well, no, it’s funny on that because I personally think that middle managers are always, it’s always that weird role when you’re not a doer and you’re not a decision maker per se, you’re just kind of that middle manager. And those roles always get squeezed, particularly in economic downturns and in industrial revolutions, if you will. Right? Like if you don’t need the middle manager to aggregate the reports, then what are they doing right now? If there are player coaches who are driving people and inspiring them to achieve greatness. That’s a different skill set. Right. But if you’re just taking the orders from here and panning them there, that’s a minimal utility. Actually, that I think it’s a really good opportunity because those people also tend to be middle to late in career and so then cost more. And so there’s more meat on the, on the bone there in terms of cost savings versus removing an offshore part time sourcer who costs you $800 a month.

Aaron Elder [00:13:27]:
Do you know what I’m saying? It’s marginal. I think my point is that that’s actually kind of interesting is would AI replace that? And then the question is, at what cost? Because here’s an interesting math you got to do. If you got an offshore resource, who’s cranking through your data cleanup projects, doing it the way you want, they cost $800 or $1,200 a month, and you’re like, okay, that’s fifteen k a year. It’s getting done. Things are fine. Are you really thinking you want to go spend $50,000 and several months to go implement a tech to go replace that only to have it maybe not do it? It’s a steep roI curve for you to sort of think through. Right. And that’s kind of where we’re at.

Aaron Elder [00:14:04]:
Yeah, go ahead.

Kortney Harmon [00:14:05]:
No. Well, and I think that kind of plays into the last question with this to wrap up. I’m not sure if we have the resources and the knowledge in house to properly evaluate and implement. And I think that’s where everyone is. It’s like, there are some companies that are like, well, we have someone in charge of AI or, oh, we have lunch and learns once a week that we’re evaluating IAI to see if it’s best fitted for our organization. But who’s responsible to know if you’re getting the return out of it? Who’s responsible to evaluate if there’s overlapping tech of what you already have? Are you spending additional monies that you don’t need to. There are so many levers to pull in this process and things to think about.

Aaron Elder [00:14:40]:
Totally. And I think especially when times are tough. Right. Especially when the pressure is on, people will frequently confuse change with progress. I was talking to a recruiter and they already have a tool that they have a. I don’t know if I would name names, but, like, they have a tool that lets them do video interviews embedded inside their emails. Great. But they were talking to their coach, and this coach uses a different tool, and the new tool costs five times as much as the old tool.

Aaron Elder [00:15:04]:
But it’s new. And the point was that you’re not using the tool you have today that does the exact same thing because you didn’t take the time to go look. And literally I got on the screen share with them and I went into the KBR article, I’m like, it actually supports the thing you’re trying to do. It’s just right here, you just didn’t know it. You were going to go waste a bunch of time and buy a whole new tool to go do this thing. I guess my point of that is as you’re evaluating AI solutions, you need to understand that there will be an adoption cycle and a cost for it, that the benefit will never be quite what you budgeted for. So always apply a discount to that and then see if the math works out and if the level of effort makes sense. There’s also lower barriers to entry to doing this.

Aaron Elder [00:15:43]:
You can sign up for these tools today that are arm links with your ATS or other processes and do kind of a copy and paste integration for now and at least measure the ROI to get a sense of this is directionally heading to where you want to go and that way you can try a range of them. One thing too. Another seed I’ll plant with you is realize that there’s a lot of these startups that are popping up and they’re offering call summarization and outreach. Da da da. Behind the scenes, they’re really just repackaging these large language models that are being developed by these mega corporations and mega organizations. They’re actually providing very limited utility on top of those. And those things are moving fast. Everything’s moving fast.

Aaron Elder [00:16:23]:
And so I guess what I’m getting at is you might want to try before you buy as you think about implementing those things because people are kind of grifting right now on selling the excitement and selling the change over progress. Companies that didn’t exist a year ago have just started up and they’ve slapped a UI on top of AWS something. You know what I’m saying?

Kortney Harmon [00:16:43]:
Yeah. Erin, I think you answered all of this letter as condensed version as we possibly could, and I think it gave our listeners definitely things to think about and to process before jumping on the, you know, the shiny tech syndrome where they’re ending up spinning cycles and spending cycles. We don’t want them to do that. So thank you very much.

Aaron Elder [00:17:04]:
Great, thank you.

Kortney Harmon [00:17:07]:
I’m Kortney Harmon with Krill Eight. Thanks for joining the full desk experience. Please feel free to submit any questions for next session to fulldeskrilate.com.com or ask us live next session if you enjoyed our show. Be sure to subscribe to our podcast wherever you listen, and sign up to attend future events that happen once a month.