SmartRecruiters' Rebecca Carr on Making AI Work for Talent Acquisition

[00:00:00] Welcome to PeopleTech, the podcast of WorkforceAI.news. I'm Mark Pfeffer. My guest today is Rebecca Carr, the CEO of Smart Recruiters. Recently, they conducted research that found more than a third of businesses aren't sure they've hired the right people. What's that mean for talent acquisition? And what are employers doing about it? We'll talk about that and more on this edition of PeopleTech. Hey, Rebecca, welcome.

[00:00:40] Late last year, Smart Recruiters did some research about talent acquisition, and you found that about a third, 34% of businesses aren't sure they hired the right people once all is said and done. And let's start by asking, what's that mean for people in talent acquisition?

[00:01:03] Yeah. Well, actually, I would argue that that's understated as well. Those are people that are admitting they don't think they hired the right people based on turnover rates that we see within the system.

[00:01:16] So that would be people hiring jobs, those people being onboarded into the HRS, and then those same jobs opening up as replacement or backfill. I think you might be seeing upwards of 50% of people would argue that they didn't find the right person. Now, I think there's some reasons for that, that are symptoms of talent acquisition teams that are a bit overwhelmed by the process and technology that they have.

[00:01:45] Interviews and interview quality is something that is very inconsistent across applicant tracking. So the types of interview data that you collect, the total number of interviews you have to evaluate skill, the use of assessments in the workflow, this is actually declining a fair bit, probably because organizations in the last couple years have been asked to become much more efficient to spend on

[00:02:13] a lot less. This is not just unique to talent acquisition or HR, it's every department in these large organizations are being asked to cut back. And one easy place to do that is point solutions. Unfortunately, you do that, and you lose some of the mechanisms that you might have leveraged to validate skills, validate fit, understand if you have the right person in seat.

[00:02:41] And I think that's a lot of experience. And I think that might be having an impact on the numbers. And second to that, I would say that talent acquisition teams in some of our organizations have been reduced by 50%. So fewer people are managing what is not a 50% lower hiring plan. And that means that you have to do a lot more with less and you skip some steps and mistakes happen when that happens.

[00:03:06] You know, a lot of people have talked about the purple squirrel defining a job that really can't exist or a candidate who really doesn't exist. And how much of that comes into play here where the employer isn't getting what they want, but it's kind of because they don't really know what they want. Yeah, I think a lot of people don't know what they want.

[00:03:33] Yeah, I think a lot of people don't know what they want because they don't necessarily know what they're going to need in one or two years. A lot of these organizations are struggling to differentiate. So they're experimenting with a lot of new technologies and skills in their own product strategies and business strategies. And usually when you start to build different products, new products, you require a different type of person, a different type of skill. You think you know what that is. You go and define it in a job description.

[00:04:02] You might find someone that is a fit for that, but not maybe a fit for what that evolves into in six or nine or 10 months as you learn more about your market, as you learn more about your customer and you learn more about what it's going to take to reignite the growth in your organization.

[00:04:24] So that's true. I also think people get a little bit too specific sometimes and they end up weeding out people that are on the fringes of a role, meaning they have either adjacent skills or 60% of the skills required to do a job. But because the job description is so specific when an AI or a matching algorithm on a LinkedIn or Indeed or otherwise starts to pull jobs up for them, that job might fall farther to the bottom of the list.

[00:04:53] And you might not be getting the candidate pool or a big enough candidate pool of eligible candidates because you've narrowed so refinedly in such a refined way. So it's similar to if you go on Google and you type in a very specific set of criteria, you get specific enough, you have no more results.

[00:05:15] And I think the same can be applied to job descriptions and candidates being matched to those descriptions. It must put talent acquisition people in a terrible place because it looks like they're not doing their job well, but a lot of what you're describing is kind of structural. I'd say it's structural, very deeply connected to business strategy. I think it's also technology burden.

[00:05:44] We've done a couple like internal customer surveys more recently that have suggested that about 80% of the recruiting workflow is administrative. So it's filling out a big form and making sure you get approval from this person and writing a scorecard and setting up screening question criteria and configuring assessments, chasing interviewers, scheduling interviews,

[00:06:10] all the above, all the above, which is not finding people and making sure they're a good fit for your business. And so I think there is an opportunity for the recruiters that do remain in seat to start embracing more AI technology to bring efficiency to that workflow so that they can weed out that noise that is taking away those precious hours of the day from focusing on what's most important.

[00:06:39] It is it's hard to do because most technologies are not enabled to do this, but I have in the last, let's call it six to nine months, seen a wave of new innovations come out of recruiting tech where platforms like a smart recruiters, but like many others have introduced features that recruiters can use to help support them in this moment.

[00:07:02] You'd be surprised at how few people adopt them though, even though they're cool and sexy and people are like, oh, we're going to put a big marketing article out there. And everyone's going to think that there's a big innovator. It's fascinating when you actually see co-pilot adoption. A lot of people are still very stuck in their processes in ways. And if they were to pick their head up and try to try some of these new things, I actually think they'd be able to do their job better and faster and smarter.

[00:07:31] And it might quash some of their concern and lack of confidence around performance despite all the business challenges that might be circling around them. How much of it is fear? You know, I think there's still a lot of people out there who are nervous about AI tools and how it's going to impact them. And, you know, this talk of increasing efficiency and talent acquisition has got to have some people looking over their shoulder. Yeah, no, there is fear.

[00:08:01] There's fear by the individual persona of recruiter or hiring manager, but there's also fear at the corporate level around making the wrong investment. AI is a bit pricey. A lot of people have budgets for it, but it's quite the leap of faith. And one of the things about AI that we hear from a lot of our customers is first impressions really, really matter.

[00:08:22] And it's the same if you were to go on a website and start to converse with a bot around, let's say you go to your banking website and you're asking questions or asking it to open an account. Or you call a line and you get a bot that sends you through a program where you're picking two and then entering in a number. The second that it starts to not work, you find yourself yelling into the phone, no, go back, representative, representative, because you're so frustrated.

[00:08:50] And if you can imagine that same application to a candidate workflow, there's a fear. You get that wrong in the candidate experience. You could lose those really good people that you actually do need to bring in. So that is as equal a fear as maybe my job is going to be displaced.

[00:09:10] I actually think recruiters that are starting to embrace these technologies are being more swiftly elevated within organizations because they're demonstrating a behavior and a mindset that CEOs like myself want to see. I look at my organization all day long and go, be more productive, figure out how to do three jobs in one.

[00:09:31] Well, if you're a recruiter and you're figuring out how to do three recruiting jobs or the coordinator job, the recruiter job, the talent management job all in one, you look superhuman. You're an example that actually I would double down into. And so it feels a bit counterintuitive. But you do see examples of this in some successful customers.

[00:09:57] I would say there's probably one bigger looming fear, though, which is regulatory. There's a lot up in the air, a lot of gray area right now as it relates to how local federal requirements will put pressure on vendors and companies to be more transparent, more explainable, more auditable. They might not even allow certain applications of AI. And you don't want to be caught in those crosshairs.

[00:10:24] So you do see a group of people that are stepping back and going, well, let me just wait and see how it plays out. I think, though, there's a difference between high-risk AI products and lower-risk efficiency AI products. Gen AI is a good example to help you write a job description or help you schedule an interview is not a decision. The agent itself is not making a decision. It's taking an administrative action.

[00:10:53] Those I would actually encourage people to lean in very quickly, too. If you're going to sit back and you have a lot of concern, AI for discovery and things like that is maybe where you might be a little bit more hesitant. But I give it only 9, 12, 18 months before I think the world starts to see a lot more clarity on what's acceptable and what's not and what you should be asking of your vendors in order to reduce risk for your business.

[00:11:22] So where are your customers in all of this? Where's their comfort level and their progress? And what's smart recruiters doing to sort of help them along? I think that the comments you make about fear are prevalent even within our customer base. I think they're getting more and more comfortable as we've started to release more efficiency bots into the system. You see people start to pick those up.

[00:11:53] But I would argue only between 25 and 30% of our customer base is fully embracing AI, leaning into design partnership, rolling it out globally. You see a lot more people starting to experiment in pockets. I want this department to try it and see if it works. Or I'm going to do this small POC in this particular country. And vendors like smart recruiters are leaning into that moment, helping them to roll those out, really get comfortable with the idea.

[00:12:23] Because we know, or at least we feel very confident in the technology we're bringing to market. We know if they see success, if there's measurable impact, if there's good value there to end users, they're going to roll it out much broader. And that is going to only benefit us. But smart recruiters, you asked the question, what are we doing very specifically? First and foremost, we're focusing on a very specific type of hiring as a starting place.

[00:12:53] AI, very broadly, is going to solve a lot of problems for recruiting. But recruiting for an intern from a university versus recruiting for a CMO or a project manager at your corporate headquarters is very, very different.

[00:13:11] And to my point around first impressions matter, we have decided to really focus a lot of our energy into high volume, high turnover, very specific types of like lower skilled roles.

[00:13:28] So think hospitality, retail, manufacturing, healthcare facilities, things like a store manager or a delivery driver, where the requirements are very, very specific. The profile is a bit broader. And the evaluation is really around fit and maybe some one or two very small skill requirements.

[00:13:56] We're leaning into those workflows, which there are many businesses around the world that struggle with high volume hiring. And we're focusing on making hirers more efficient and productive. As an offshoot of that, candidates end up leveraging a lot of our AI technology.

[00:14:13] But the persona I'm solving for is how do I help that H&M store manager hire as many people as they need to hire while not taking away from their day job, which is they need to run an H&M store. And how do they leverage AI to be really productive and efficient in doing that? And so we're starting there. And Winston, our new hiring companion, AI hiring companion, is focusing on that workflow.

[00:14:41] And as we continue our roadmap throughout the year, you're going to see us extend to more complicated hiring workflows, more complicated roles. Just, again, bringing efficiency and productivity to the hirer.

[00:14:55] That's helping them find candidates faster, helping them make decisions with more information, helping them schedule interviews, helping them understand what stores or attributes of their hiring workflow are inefficient as compared to others. And they could make some improvement. And just using AI to really fulfill the mission that we have as a business, which is to make hiring easy, to make hiring faster and smarter.

[00:15:25] We're just making sure we start small so we really get it right. And we provide value to those customers in that particular workflow and can build upon that trust with more and more features and functionality over time. When employers look at this whole situation, their satisfaction or lack of with talent acquisition and the kind of people that they're getting, it strikes me if I was an executive, I'd say, well, that's a problem.

[00:15:55] What should employers be doing about it to solve that problem? Specifically to solve the quality issue? The quality issue and the sort of roadblocks or hiccups you've mentioned. Yeah. For one, I think employers do need to focus on sort of that core crux of an issue we talked about a little while ago, which is business strategy. Where are you going and who are you going to need?

[00:16:23] What skills are going to make you a more powerful business as compared to your competitors, a more differentiated business as compared to your competitors? That question is unanswered for a lot of the businesses we work with. And without answering that, no matter how different your workflows are or efficient they are, it's going to be difficult for you to change the quality of the people and the right people that you bring into your business.

[00:16:48] The second piece is data cleanliness and normalization and integration. AI is pretty powerful, but it requires good data. It can't read context out of thin air. It doesn't know all the stuff that's offline. It only knows what you give it.

[00:17:10] And it is pretty shocking, actually, how unclean a lot of the data sets around skills, job profiles, positions, job descriptions exist within organizations. Investing a lot in figuring that out and having really clean job profiles. If you have a project manager, what does that job profile look like? What are the skills of the people that are in your business already that have that job?

[00:17:39] Like, what is the career path for that? What was this like? Where did those people come from? Understanding that and then feeding it allows agents to make more intentional recommendations that are going to be very suited to your business. But just saying I need a project manager is far too vague of a prompt, so to speak, to get the right thing. And unfortunately, a lot of people just go, I need a project manager.

[00:18:07] Three people get returned or five people get returned or a job description gets returned. They get an inbound of candidates. They're overwhelmed, so they just fire or they hire, rather, the first or second person that they interview and they like. And the reality is that was far too broad of an articulation of what's required to really narrow yourself down to the people that are truly going to be the best fit.

[00:18:31] So it really starts with your data and junk in is junk out in these moments. So I do think it's important that investing in, you know, a consistent language across your systems, consistent data entry tools that drive adoption processes that are integrated into your end user workflows so that they use your tools. They always have clean and consistent inputs. This is going to be important, I think, to getting it right.

[00:19:01] And then, of course, there's adoption of AI tools within these workflows that I think help along the way. Do you think that the business executives, the company leaders sort of understand all that, that this isn't necessarily something that technology is going to solve for them? They still need talent acquisition skill to oversee the technology. Or are people under the impression that if I install something with AI, everything's going to get better?

[00:19:32] Yeah, I think they understand. I think the most sophisticated ones do. And they have started to invest in taxonomies and things like that within their organization. That stuff takes a long time, though. I think if there's anything that business executives have failed to do, it is invest in the infrastructure to clean up that mess faster. It's just one of those projects running in the background. But the tech is here.

[00:20:02] The value is here. And we're almost waiting for a lot of these businesses to catch up to that clean input that's going to provide the value that they seek. I do still think that most sophisticated organizations, I'd say all organizations, still understand the value of evaluating candidate fit and validating skills.

[00:20:26] You see a lot of businesses that are actually quite fearful that candidates are lying more frequently in applications because they're using AI to enhance their CVs. And in doing so, they're exaggerating maybe the skills they have. So if anything, it's forcing them to emphasize those conversations a little bit more. And for that, you need talent acquisition teams. You need recruiters.

[00:20:54] I just think that the role of recruiter is going to change and become more strategic. It has, unfortunately, I would say, evolved over the last decade or so into a very administrative role. And those people don't become recruiters so that they can clean up workflows and fill out offer forms. They're connectors. They're networkers. They're intuition about who's going to be a fit for the hiring managers for the role is what makes them great recruiters.

[00:21:23] And those skills are going to continue to be invaluable. I think that their role will evolve to also starting to think about the skills that will be required for the future and how you go and find those skills and validate those skills and thinking more strategically about organizational design and planning and less about interview scheduling and management. And most of the executives I speak to see that as well and are vocal about it.

[00:21:51] If anything, they just don't know where to start. And they're looking for some support, like the blueprint. What's step one? What's step two? What's step three? Because I've never done this and I've never seen this before. And frankly, no one has. We're in a new world, I think, over the last couple of years here as AI ML tools have really been introduced. Well, Rebecca, thanks again for joining me. I hope you'll come back sometime. Of course. It was great to talk with you.

[00:22:20] Yeah, great to speak to you as well. My guest today has been Rebecca Carr, the CEO of Smart Recruiters. And this has been People Tech, the podcast of WorkforceAI.news. We're part of the Work Defined Podcast Network.

[00:22:45] Find them at www.wrkdefined.com. And to keep up with AI technology and HR, subscribe to Workforce AI today. We're the most trusted source of news in the HR tech industry. Find us at www.workforceai.news. I'm Mark Pfeffer.