Ep 41: Using Data and Responsible AI to Reimagine Hiring Success with Barb Hyman

[00:00:00] Welcome to Elevate Your AIQ, the podcast focused on the AI-powered yet human-centric future of work.

[00:00:05] Are you and your organization prepared? If not, let's get there together. The show is open to sponsorships from forward-thinking brands who are fellow advocates for responsible AI literacy and AI skills development to help ensure no individuals or organizations are left behind. I also facilitate expert panels, interviews, and offer advisory services to help shape your responsible AI journey. Go to ElevateYourAIQ.com to find out more.

[00:00:28] Welcome to another episode of Elevate Your AIQ. Today I caught up with Barb Hyman, CEO and founder of Sapia.ai, a company that's been transforming the hiring process for many years. Barb's journey from practicing law to management consulting to HR leadership and ultimately to building an innovative AI platform for hiring is certainly fascinating and inspiring. We dove into some big questions like how can we challenge bias in hiring to make recruitment more fair and transparent? What role does data play in improving the candidate experience? Where does the future hold for HR tech in a world-centric

[00:01:09] increasingly driven by self-service tools and retention-focused metrics?

[00:01:13] Spoiler alert, retention is essential. Barb's passion for shaking up recruitment and empowering both employers and candidates is readily apparent throughout this conversation. She's also a leading voice for responsible AI in the talent and HR tech space. So of course, Barb gets kudos from me for her advocacy in that respect. If you're curious about how AI is reshaping talent acquisition in HR, and the ethical considerations that come with it, you'll want to listen to this insightful conversation with Barb.

[00:01:39] Thanks for listening.

[00:01:42] Hello, everyone. Welcome back to another episode of Elevate Your AIQ. I'm your host, Bob Pulver. With me today, I am catching up with Barb Hyman, who is the CEO and founder of Sapia.ai. How are you today, Barb?

[00:01:54] I'm wonderful. Thank you for asking.

[00:01:56] Nice to see you. We haven't spoken in a little while, but I'm really looking forward to catching up over the course of the conversation.

[00:02:04] Yeah, me too.

[00:02:04] Just to kick things off, why don't you just give our listeners just a little bit about your background and the impetus for starting Sapia?

[00:02:12] I'm sort of a bit old, so I've done a lot of things in my life. I started my career as a lawyer, then moved into managing consulting with Boston Consulting Group after doing an MBA, and then fell into the HR role, was in the seat of head of HR for BCG, and then for a listed company here in Australia called the REA Group, which is mostly owned by News Corp, actually, one of the original disruptors of the classified space.

[00:02:39] And from there started this business, Sapia.ai, really because in both HR roles, I could see that people were everything.

[00:02:48] They were the most important asset that you have, even though they don't sit on the balance sheet and every day they leave the building and you worry about whether they're going to come back.

[00:02:56] And we put so much work and effort and money and people costs into hiring them, loving them, retaining them, developing them.

[00:03:04] It just struck me that there was so little data and science and efficiency in how we went about those tasks.

[00:03:12] I wanted to reimagine that whole reality.

[00:03:15] So that's the journey that we've been on since around about 2018.

[00:03:19] I can speak from personal experience, having used your chat interview product, that it was quite impressive.

[00:03:27] A refreshing departure from some one-way video interviews and some other nonsense that I've experienced in my tenure.

[00:03:35] It was also nice to get some personalized feedback almost instantaneously, which pretty much the antithesis of getting ghosted, right?

[00:03:43] Yeah, I mean, I just, I often think of, you know, the process of finding someone for a role or finding my next role is very similar to dating.

[00:03:54] And at the end of the day, the less asymmetry of information, the better.

[00:03:58] You know, if I'm going to decide on who I'm going to partner with or marry, I want to know as much as I can about them and they want to know the same about me.

[00:04:05] So I can make a really informed decision.

[00:04:07] That's probably the other area of opportunity, you know, where most of our team will often say, gee, I wish we could take this science into dating because it's such a critical life decision.

[00:04:17] And in the hiring space, you know, there's such asymmetry of information.

[00:04:21] You're expected to go and trawl through a career site to figure out whether or not it's the right place for you.

[00:04:26] Very hard to get insight into the manager and the team and the culture, really, unless you go to Glassdoor and even that's going to be fairly deficient.

[00:04:35] And you get to see a resume, which, as we all know, is a pretty poor substitute for really understanding who someone is.

[00:04:42] So there's just an absolute forcity of data when it comes to making the decision on both sides.

[00:04:48] And so we really wanted to change that, but also create the equity that respects that both parties are owning this decision.

[00:04:56] It's not just a one way decision.

[00:04:58] Yeah, absolutely.

[00:04:59] I mean, I like the dating analogy a lot.

[00:05:02] It makes a lot of sense to me because ultimately you are looking for, hopefully, a relatively long-term commitment.

[00:05:08] And you're looking for a win-win kind of situation.

[00:05:12] So I was already married by the time any of these dating apps came out.

[00:05:17] But it just seemed like people complaining about an eHarmony, more thorough evaluation or whatever it is that you do when you sign up.

[00:05:27] People complaining about the time that it takes to do that.

[00:05:30] And it's like, well, look at the potential outcome, right?

[00:05:33] I mean, if you're just looking to go and spend the night with someone, there's other apps for that.

[00:05:40] But we're talking about your career, your professional development, important relationships that are going to help you sort of grow as a person and as a leader.

[00:05:53] And so it just seems like, yes, the juice is worth the squeeze, as they say.

[00:05:59] Yeah.

[00:05:59] I think the other thing that we really wanted to challenge was this very strong bias we have as humans that I need to see you to hire you.

[00:06:08] And obviously in the dating world, we get that, right?

[00:06:11] That chemistry is really important and it comes from that physical connection.

[00:06:15] But in the working world, that doesn't really matter.

[00:06:18] You're not looking for that kind of chemistry.

[00:06:20] And the reality of the way we hire today and the tools like video are it just creates bias.

[00:06:30] It really just, I think, video productizes bias because we all know that humans are biased.

[00:06:35] We're mostly unconscious of those biases.

[00:06:37] And so we will form immediate opinions about someone based on the way they look.

[00:06:42] I'm sure that happens on LinkedIn in terms of who people choose to go and engage with from a sourcing perspective.

[00:06:48] So we also wanted to challenge that and say, you know, why does it matter what I look like?

[00:06:53] Now, maybe if you're hiring me to go and represent, you know, Tiffany's in that store, you know, I need to look a certain way.

[00:07:00] But for most roles, it's irrelevant what you look like.

[00:07:03] But yet we are still so highly tuned to I need to see you to hire you.

[00:07:08] Interestingly, during COVID, there were a lot of companies that obviously moved to asynchronous.

[00:07:13] Entrepreneurs recruitment interviews and video interview unfortunately went off, but so did other ways of interviewing.

[00:07:20] And tech companies like Atomatic really invented this whole idea of let's just use chat as a blind format to interview you and to understand you.

[00:07:30] And, you know, go really deeply in terms of how we conduct that interview.

[00:07:35] But at no point do we actually need to see what you look like.

[00:07:38] And they didn't go because they were trying to remove bias.

[00:07:40] They just saw it as the most efficient way to assess people through chat.

[00:07:45] And that was for me a really key part of what we wanted to do, which is how do we remove bias?

[00:07:49] Because I think it's not just about social equity and opportunity.

[00:07:53] It's really about missing out on talent.

[00:07:55] And to think that what someone looks like is going to in any way correlate with how they perform in their role is kind of a delusion that we humans hold.

[00:08:03] And so if you can remove that, it challenges your bias, but it opens up a much bigger world of opportunity.

[00:08:09] And, you know, we always want more volume coming in because more volume underwrites more diversity and typically a better quality tool to choose from.

[00:08:17] Yeah, absolutely.

[00:08:18] I think if you cast as wide a net as you can and focus on, you know, someone's actual potential to succeed in the role and potentially future roles, then, you know, what else really matters?

[00:08:34] My concern was not just maybe biases around someone's physical appearance, but it's just the way that, you know, how it could be like trying to detect emotion or maybe even figure out if someone's telling the truth or not.

[00:08:50] So there's a lot that people are going to try to use technology for, but whether that's an appropriate use of this technology is often, you know, questionable.

[00:08:59] Yeah.

[00:09:00] Yeah.

[00:09:00] Right.

[00:09:01] So, yeah.

[00:09:01] So those kinds of things, you know, scare me a bit.

[00:09:04] And I know, you know, safety is more obviously based in Australia where you are, but then, you know, EMEA is a huge market for you.

[00:09:13] I know in the U.S., you know, we've got this patchwork of legislation all over the place and some of those things are focused on, you know, automated employment decision tools like here in my backyard in New York City.

[00:09:25] But then you have other places where they're worried more about facial recognition and other types of technologies that can be, you know, used or abused.

[00:09:33] So what are some concerns that you see as you travel the globe?

[00:09:37] Yeah.

[00:09:38] So, look, we're obviously very fluent with the regulatory environment that is now and is coming across all those different markets because we do operate in all those markets.

[00:09:46] We do have U.S. clients, clients in North America and Canada and obviously in the EU as well.

[00:09:51] And we tend to work with really big companies, listed companies, enterprise who might be hiring 10, 20, 50, 80,000 people a year.

[00:09:59] And so they have a high bar as they should around what is this technology?

[00:10:05] What is the science behind it?

[00:10:06] How do you respect data?

[00:10:08] Do you treat data like the crown jewels?

[00:10:10] You know, give me the evidence around that security posture and so on.

[00:10:14] What I would say from a global perspective in terms of regulation is the U.S. is very messy.

[00:10:19] Yeah.

[00:10:20] The fact that you've got this federal and state reality and you have so many states creates a lot of complexity for business.

[00:10:26] And, you know, we work with clients like Joe and the Juice who are in the U.S. who operate across many different states and the need to adhere to whatever the particular peculiarities are of the regulation in a state just creates, you know, a cottage industry, frankly, of people who are there to provide those services.

[00:10:44] And that's a very challenging market to navigate if you're a company that's operating across all regions.

[00:10:49] In contrast, the EU is very clean.

[00:10:52] And, you know, contrary to what Jan LeCun says around, you know, how constraining it is to have EU regulation, actually, it's incredibly efficient.

[00:11:01] So when the AI Act comes in in 2026, August 2026, the way it works is that we have to go through a process one time that has our technology scrutinized on every dimension.

[00:11:12] And then we get a, you know, we have a CE mark, it's called, and you go into the database.

[00:11:17] And companies who then want to procure this technology can feel confident that there's been a really rigorous process to get you into that.

[00:11:23] And that is just beautiful and clean and makes life a lot easier, not just for vendors, but more importantly, for companies.

[00:11:31] Whereas what I'm seeing in the US, the lack of certainty around the regulatory environment, the fact that you have to navigate so many different states and where they're going has almost just put the shutters down in some conversations we've had.

[00:11:45] It's just become too hard for legal counsel.

[00:11:48] And I also think that the maturity, as a result of that, the maturity of the US market, I think, is quite behind the UK and even the Australian market in terms of understanding how to navigate this regulatory world.

[00:12:01] So GDPR raised the maturity level of companies big time.

[00:12:05] So what we're finding in the EU is it just feels like another GDPR process as in, okay, what do we need to do?

[00:12:12] What are the boxes that we need to tick?

[00:12:13] What are the audits, you know, if they need to happen that need to take place?

[00:12:18] Whereas in the US, there's just sort of an absence of clarity and understanding.

[00:12:23] And so I actually think, ironically, for a country that's really built off innovation, I feel that that is a market that is being very slow to innovate when it comes to using AI.

[00:12:33] I think the other thing I'd say around regulation is, you know, we've been really obsessed, paranoid about the importance of creating ethical AI right from the beginning.

[00:12:43] Because you're talking about people and decisions around people.

[00:12:47] And for me, experience and trust has been such a critical value for us.

[00:12:51] Like trust is one of our values.

[00:12:53] It's one of the product principles.

[00:12:54] And how do you create trust?

[00:12:57] Well, one of the ways is that you're open and you're transparent.

[00:13:00] And you say, here's the data that's being used.

[00:13:03] And that data is visible to you as a user, which is the responses to the free text questions.

[00:13:09] Here's the profile that we're looking for.

[00:13:11] So we use what's called rule-based models, as opposed to classical machine learning models, which don't create for any real explainability and understanding.

[00:13:19] And so you make very deliberate product design choices and you embed explainability within the product.

[00:13:26] And so we've always had those principles to find our product.

[00:13:29] But universally, it's all the same, you know, in terms of what it is that makes for ethical AI.

[00:13:33] It's, is the data clean and ideally free of demographic data?

[00:13:39] You know, we don't use any resume data.

[00:13:40] I'm, I'm kind of amazed that so many vendors and companies are sort of mindlessly using resumes when we all know resume data is, is sort of pregnant with latent bias, socioeconomic, you know, race, et cetera.

[00:13:57] But no one seems to be bothered by that.

[00:14:00] I don't quite understand that, Bob.

[00:14:02] Maybe you can help me understand that.

[00:14:03] You know, the other is around, you know, that there's ways to, to evaluate and test bias.

[00:14:08] We've always had live bias reporting using four fifths and effect size, you know, at the role family level.

[00:14:14] The other is explainability by using things like wool based bottles, transparency.

[00:14:18] You know, there are just fundamental principles that are almost common sense that you can apply.

[00:14:24] You don't need to go and do a course with Josh Burson to figure out what questions do I need to ask.

[00:14:29] Just go with your own common sense and intellect to go, okay, how do I feel safe with this?

[00:14:34] How do I know what's happening here?

[00:14:36] How do I know what it's measuring?

[00:14:37] You know, where is that data?

[00:14:39] What do you do with that data?

[00:14:40] What is the demographic data that's being used?

[00:14:43] It's kind of pretty fundamental.

[00:14:44] So I feel like, you know, there's no reason not to lean in and learn and just start to ask the right questions or ask any questions.

[00:14:54] And that's going to be your learning process live.

[00:14:57] I completely agree with you.

[00:14:58] I guess I'm wondering if you're seeing more educated buyers in that sense.

[00:15:04] I mean, are they starting to include this in their RFI, RFP processes?

[00:15:08] Are they holding?

[00:15:09] Because I think this is how you become a true or you remain a trusted partner to these clients, right?

[00:15:17] To your talent acquisition teams or talent management teams, talent development teams.

[00:15:21] If you can't trust the AI and the data behind it, I mean, all bets are off.

[00:15:28] You're just grabbing it at shiny objects, trying to maybe simplify your infrastructure or something and, you know, have fewer vendors to yell at.

[00:15:38] Yeah, look, I think it's a bit like the healthy food check.

[00:15:42] I don't know if you have something like that in the US or in Australia.

[00:15:45] You can go to the supermarket and there's a tick that says this has been approved as healthy or good for heart disease.

[00:15:52] And, you know, that AI label, that sort of AI washing is being applied everywhere.

[00:15:57] And so I think there's a lot of confusion and conflation of what is really AI.

[00:16:02] The simple definition is think about it as a human, which is the thing that we do really well as humans is we learn.

[00:16:10] Sometimes we learn based on biased data points, but ultimately we have the ability to get better at something, you know, whether it's, you know, how we practice a sport.

[00:16:20] But we can improve and we improve based on the experience and based on learning.

[00:16:25] And so AI is fundamentally in place if it is able to learn and if all it's doing is just automating from one step to the next, look at a resume and find the keywords and then automate the right keywords to the next stage and give them a score.

[00:16:41] You know, there's not a lot of AI really going on there.

[00:16:44] There's not a lot of intelligence.

[00:16:45] It's pretty simplistic.

[00:16:46] But something which is learning from data is really what I think is the true pinnacle of where you get the value from using AI.

[00:16:55] So an example is around, you know, our technology is using language data and natural language processing to effectively DNA you.

[00:17:03] It understands your skills, your competencies.

[00:17:08] You know, are you a good problem solver?

[00:17:10] What are your communication skills like?

[00:17:11] Are you someone that we can rely on?

[00:17:13] Are you high accountability?

[00:17:15] You know, maturity, resilience.

[00:17:16] All of these qualities that don't exist in a resume is what we're able to discover through this chat conversation.

[00:17:22] And it's much like what you do in an interview.

[00:17:24] You know, when you interview me, you ask me open-ended questions to test my experience and you're extracting meaning from that.

[00:17:32] You're figuring out, wow, Barb's got amazing reflective learning and she, you know, she's good at this but she doesn't sound like she's very good at that.

[00:17:39] Effectively, that's what we're doing.

[00:17:40] And, you know, you start with a hypothesis of what you think great looks like.

[00:17:44] So, if we take a sales role, a lot of recruiters think that the best salespeople are team players, extroverted, you know, into team kumbaya and, you know, but are they?

[00:17:56] And what you're able to do when you're using a true data science tool that is capturing a data profile is you go, well, let's actually look at the profile of success of those who've been high.

[00:18:08] So, we hide these people because we believe these qualities matter.

[00:18:12] Now, we can actually see that the people who actually get into productivity fast and who are continuously meeting target and exceeding target are actually a different personality profile.

[00:18:22] Wow.

[00:18:23] So, we got it wrong.

[00:18:24] How do we teach the AI to recognize that reality and start to optimize for that profile?

[00:18:30] And that's the power of AI is that you get better, which no recruiter ever has the opportunity to do because you never get that level of feedback or data.

[00:18:40] And so, what that means for our clients is that we're driving turnover reduction using data.

[00:18:45] The ability to say we're seeing 80% turnover, now we know the profile of success.

[00:18:50] And by the way, we don't just know it as a role.

[00:18:52] We know it as a role in that store, in that state, in that particular configuration.

[00:18:57] You know, you're hiring in airlines.

[00:19:00] We work with Qantas Air Canada.

[00:19:01] And the experience of being on a plane that's an international versus a regional versus a domestic versus, you know, Asia is very different.

[00:19:10] And we all just think of a ubiquitous cabin crew role.

[00:19:13] And so, that power to customize at a local level in terms of the actual role context based on data that is objective data, it's not a performance measure, it's a true profile of success, which we see, you know, sales and retention data being very clean, is the power of AI.

[00:19:32] Right.

[00:19:32] Not how many interviews you were able to schedule in a day, right?

[00:19:35] There's very little intelligence in that process.

[00:19:37] All you're doing is automating a faster status quo, you're not actually getting better.

[00:19:42] Because ultimately, what companies want is to hire less.

[00:19:45] Yeah.

[00:19:45] You know, how do we not have to hire so many people?

[00:19:48] Because we've done such a frigging clever job of figuring out what is the true profile of success.

[00:19:53] And that's what we optimize for.

[00:19:55] Hi, I'm Stephen Rothberg.

[00:19:56] And I'm Jeanette Leeds.

[00:19:58] And together, we're the co-hosts of the High Volume Hiring Podcast.

[00:20:02] Are you involved in hiring dozens or even hundreds of employees a year?

[00:20:05] If so, you know that the typical sourcing tools, tactics and strategies, they just don't scale.

[00:20:11] Yeah.

[00:20:12] Our bi-weekly podcast features news, tips, case studies and interviews with the world's leading experts about the good, the bad and the ugly when it comes to high volume hiring.

[00:20:23] Make sure to subscribe today.

[00:20:24] Do you see a maturation in terms of the types of metrics that teams are tracking?

[00:20:31] Because there's just so many organizations, probably still the majority, that are doing what we know is not the right approach.

[00:20:39] We're just trying to starting.

[00:20:41] Your first gate is broken, right?

[00:20:43] Like the matching of this AI generated, in many cases, CV with perhaps now AI generated job description.

[00:20:53] And then you've got this cat and mouse game, which is getting you not very far because that's not...

[00:21:01] Unless you're hiring for a journalism position or something where your CV or your cover letter is actually a writing sample for the job,

[00:21:11] they might have outsourced that task to someone else so that they could put their best foot forward, right?

[00:21:19] So you've got to...

[00:21:20] I don't know.

[00:21:21] I just think it's a really strange bet to place and put so much emphasis on that everything comes through this faulty first stage of evaluation and decision

[00:21:35] when you know that the skills that are really going to show you that your quality of hire is going up

[00:21:42] and people are more likely to stay longer and be more engaged and things like that.

[00:21:49] And then as AI, of course, gains more sort of competency in some things that used to be the domain of humans.

[00:21:56] These are the skills that really matter.

[00:21:58] And those are the skills that you're talking about, right?

[00:22:01] Like having someone just take their time, make sure it's their own brain,

[00:22:05] you know, talking to their fingers and typing in, you know, these thoughtful responses that show you

[00:22:13] do they have the human, you know, power, durable skills that are going to get to that win-win we were talking about before.

[00:22:22] Yeah, look, I think it's really critical that you start with what's the business problem we're trying to solve.

[00:22:27] You know, something I agree with Josh Berson on, which is to fall in love with the problem.

[00:22:31] And I think too little time is spent on that.

[00:22:36] And too much time is dedicated to just layering on the next technology that looks really cool,

[00:22:43] that's sold by a really awesome salesperson.

[00:22:45] I think HR is massively overstacked.

[00:22:49] And I think they're in for a big awakening because platforms are going to become redundant very quickly.

[00:22:56] You know, you're still going to have your workdays and success factors.

[00:22:59] And then you're going to layer on APIs because layering on platforms is incredibly expensive and risky from a security and a data governance perspective.

[00:23:08] And no matter what you say about the integration, it's never as great as what you would need it to be to make it seamless for your people.

[00:23:15] And no one has ever said, bring me another platform.

[00:23:18] I can't wait to spend my time in Workday.

[00:23:19] It's just so outdated.

[00:23:23] And the beauty of, you know, generative AI is that it can be a human-directed experience as long as it's laid on something that's reasonably intelligent, you know, and accurate and guard-railed around hallucination.

[00:23:37] And so I think we're in for a world where, you know, what I say to our prospects is don't buy any more platforms.

[00:23:43] Like you've got success factors.

[00:23:45] Wait, because technology like ours and others are very quickly moving to APIs where you can feed in to deliver a lot of the solutions or the outcomes.

[00:23:55] And you don't need to have layers and layers of platforms.

[00:23:58] So I'll give you an example.

[00:24:00] Mobility.

[00:24:01] A lot of companies are investing in mobility platforms.

[00:24:04] And why are they doing that?

[00:24:06] Because they obviously believe, as I do, that retention is the North Star metric for HR.

[00:24:12] You know, if you have good retention, on average, it shows that you made a pretty good decision in the hiring.

[00:24:18] You've got a decent culture that people want to stick around.

[00:24:20] You've got a decent manager cohort who are okay at, you know, keeping you feeling engaged.

[00:24:25] And you've managed to help people find that next role and stay and grow within the business, right?

[00:24:33] It is the ultimate North Star metric.

[00:24:35] Not engagement.

[00:24:37] Not speed to hire.

[00:24:39] You know, I think ultimately retention would be one metric that every CHR should report on.

[00:24:45] And, you know, how do you go about solving for that?

[00:24:48] Well, HR is doing what they've always done.

[00:24:50] They're trying to DNA people and then orchestrate those decisions themselves where, you know, you are being asked to self-rate your skills.

[00:24:59] What we know from 5 million people doing chat interview, 47 countries, is people don't know themselves really at all.

[00:25:05] They just aren't able to truly understand their skills and articulate them.

[00:25:11] So asking people to do that is a recipe for, you know, a pretty poor data set.

[00:25:15] The other part is we know men and women think about their skills very differently, right?

[00:25:20] So, you know, I might go, there's no way I'm going to say that I'm great at X until 100 people have told me I'm great at X.

[00:25:26] Whereas you being a bloke will most likely put that in before anyone said that to you.

[00:25:30] So you're going to end up with, you know, inadvertent sort of almost adverse impact.

[00:25:35] And then somehow use that data set to figure out who are the next people that go into that role.

[00:25:40] Where we're going and where I think technology is going is we've already built a beta version of this.

[00:25:46] So you can try it on our site is everyone gets their own career coach.

[00:25:49] It has a conversation with you using it's a combination of our IP and data and Claude to help you understand your skills and then automatically skews you to the next role in the business.

[00:25:59] You don't need to do anything in HR, right?

[00:26:02] Like it's self-service.

[00:26:03] And the same with 360s.

[00:26:05] Why would you ever run a 360 process?

[00:26:08] It's torturous for both the people that have to fill it out and the people that get it back.

[00:26:13] I would much rather go forward slash 360 in Slack or Teams.

[00:26:17] It automatically, the chat goes and engages and has conversation with my peers, my one-up, my reports.

[00:26:23] Feeds it back to me and says, hey, Barb, you know, do you want the long version or the short version?

[00:26:27] Coaches me on it.

[00:26:28] You're done.

[00:26:29] And maybe I do that every month because I'm really needy as a manager.

[00:26:32] Or maybe I do it once a year.

[00:26:34] You know, HR doesn't need to get involved.

[00:26:36] So I think the whole world of HR is the operating model is changing.

[00:26:40] They're going to become a lot more recessive.

[00:26:42] All of those products and solutions that sit from the beginning to the end will become obsolete.

[00:26:48] You know, never do an engagement survey again.

[00:26:51] And so I think we're in for a really radical change in the HR operating model.

[00:26:57] Are HR leaders ready for that?

[00:26:59] Not at all.

[00:26:59] Not at all.

[00:27:01] And the only thing that I can do is say, look, focus on the end goal

[00:27:04] and don't invest in another platform.

[00:27:08] Do you target just like people leaders?

[00:27:12] Like, you know, a line of business leaders and things like that?

[00:27:15] Because I think part of your point is this is not just an HR problem, right?

[00:27:20] Everybody's got a hiring.

[00:27:22] If you're the hiring manager and the leader of hiring managers, you know, this affects you.

[00:27:27] I mean, you wouldn't put this all on HR because you guys are the ultimate decision makers as hiring managers, right?

[00:27:36] So you've got to enable those folks and understand where their current efforts are misguided or falling short.

[00:27:43] Yeah.

[00:27:44] Look, I think where we still very much engage with HR, you know, reality is the business doesn't have time.

[00:27:50] Yeah.

[00:27:50] And they are entrusting HR to do the right thing.

[00:27:53] Where I'm pretty focused is to actually get the whole HR leadership team around the table.

[00:28:00] You know, one of the things that we've had clients say to us is it's too narrow to just think of this as a TA solution.

[00:28:06] Because what you get when you bring in a true data-based technology is you start to change everything in HR.

[00:28:14] So, for instance, our platform allows you to see what are the skills, strengths and weaknesses of people we're hiring.

[00:28:20] That can now go into L&D.

[00:28:22] So it becomes data-driven L&D programs at, again, a very granular level.

[00:28:27] It might be at the cohort level, at the regional level.

[00:28:29] And rather than L&D living in a world of, you know, guessing what they think the skills are that they need to develop.

[00:28:36] We now have data that can show you from a recruitment marketing perspective where the quality of people that are applying is weak.

[00:28:43] That can go to the marketing team.

[00:28:45] We have data around where the diversity of talent is below your fair share.

[00:28:48] So if 20% of the community is this cohort, you know, let's say female and you're only attracting five, that can then feed into your recruitment marketing efforts.

[00:28:59] So it changes the way you do everything.

[00:29:01] It massively accelerates your ability to be data-driven across all of HR.

[00:29:07] Talent management and Ninebox, you now have a data profile, a talent profile that tells you what someone's strengths are so that you can create a more objective conversation.

[00:29:16] So I definitely stick with the CHRO, but what I try and do is help them to recognize that you've got to work more as one team because this data has fluidity.

[00:29:27] You know, when you capture data of people on the way in, it changes what you do, you know, all the way through.

[00:29:32] And so how do you continue to leverage that data set to help make better decisions from end to end?

[00:29:38] You know, HR's drowning in data, Bob, having said that, but none of the data is useful.

[00:29:43] You know, what is that engagement data really going to help me do differently?

[00:29:46] What is that exit survey data going to help me do differently?

[00:29:48] You have to create a linkage between the personal profile and then how do you optimize the decisions around hiring, promotion, mobility, et cetera, based on that profile.

[00:30:00] So it's a muscle that is very slowly being built, really using data to drive, inform, shape decisions and shape your HR tech state, which is truncating, massively simplifying, I think, over the next few years.

[00:30:17] Too much money has gone into HR tech, I think, by VCs and people are overwhelmed with tools and technology.

[00:30:25] And that is a consistent theme I'm hearing, which is how do we consolidate our tech stack, which I think is a really good thing.

[00:30:33] Yeah, I agree.

[00:30:34] I think HR, they have an absolutely critical role to play here.

[00:30:40] And I think you're, it sounds like you're equipping them with what they need to maintain that, you know, that proverbial seat at the table, right?

[00:30:50] Like this is the insights that we've been looking for forever.

[00:30:54] You know, there's insights buried in this data.

[00:30:56] If we can clean it up, if we can understand the impact of, you know, connecting, you know, disparate data sources, if we can understand what is really making people successful, you know, this changes, this could eventually change our whole approach to how we build our talent strategy and sort of future-proof the organization.

[00:31:17] Yeah, and I think also because we are very deep in certain sectors, so travel and tourism, retail, hospitality, et cetera, we're able to show you benchmarking data.

[00:31:27] And so for companies to see that, wow, we've got a ratio of one recruiter for 1,000 hires versus our competitor.

[00:31:35] I mean, we always anonymize it's 10,000 or the next three are 10,000.

[00:31:39] What's wrong?

[00:31:40] Why is that?

[00:31:41] Well, that's because you've got a centralized op model rather than a decentralized op model.

[00:31:44] Why do you need a centralized recruitment model in a world of, you know, high volume hiring?

[00:31:49] You want your hiring managers to earn that decision.

[00:31:51] Technology can do a lot of the work for them that your recruitment team is doing.

[00:31:54] So instead of starting with you don't need as many recruiters, it's saying here's how your peers are performing.

[00:32:00] Let's unpack that and understand why.

[00:32:02] And we have obviously a lot of insight around why.

[00:32:06] And that creates a level of almost accountability for HR performance that doesn't exist today.

[00:32:12] You know, you go and ask the CFO of a business and say, how do you know if your HR function is doing an incredible job and is doing it in the most efficient way?

[00:32:21] They have no idea.

[00:32:22] And the HR function doesn't know either.

[00:32:24] So that benchmarking has really led to some significant aha moments for both prospects and customers to go, wow, how do they do that?

[00:32:35] We should be able to do that.

[00:32:36] So that's, you know, in my consulting days, that's the one thing that drove change of behavior.

[00:32:41] You know, if a consultant comes in and says, we believe you can be more efficient, no one wants to hear that.

[00:32:46] But if you say your peers are more efficient, they have no choice but to lean into that.

[00:32:51] That makes a lot of sense for sure.

[00:32:53] I'm just curious about, I've seen some other studies where people have shown a pretty disturbing, I'll call it, you know, level of people who are basically just so desperate to find work that they'll do whatever it takes to make sure that their input is optimized, I'll say.

[00:33:14] Hey, it's Bob Pulver, host Q podcast.

[00:33:17] Human-centric AI, AI-driven transformation, hiring for skills and potential, dynamic workforce ecosystems, responsible innovation.

[00:33:26] These are some of the themes my expert guests and I chat about, and we certainly geek out on the details.

[00:33:31] Nothing too technical.

[00:33:33] I hope you check it out.

[00:33:35] Are you seeing a similar pattern?

[00:33:37] Like people are just ignoring, they're not using their common sense or reading your very clear instructions where people are trying to bypass that?

[00:33:46] Yeah.

[00:33:47] Yeah.

[00:33:47] So in terms of the percentage of people that see that and then change their response so that it isn't caught by GPT and they offer their original response.

[00:33:56] Yeah.

[00:33:57] So firstly, what I'd say is the rate of cheating is way lower than what everyone imagines.

[00:34:01] You know, years ago we built a plagiarism detector because we're working in graduate, you know, campus hiring and recruitment teams were paranoid, obsessed about how much plagiarism was happening in assessments.

[00:34:13] And at least for our assessment, you know, it's one way to think about what we do as an assessment is it was never more than 3%.

[00:34:23] But from talking with organizations, you'd think that it was 50, 60, 70.

[00:34:27] Now, I think that's due to a few things.

[00:34:30] One, the fact that this doesn't feel like an assessment, it just feels like a chat and I can take my time and the questions are what you would ask.

[00:34:38] There's a level of psychological safety and all the fear has gone from what traditionally comes with a scary gamified assessment or a traditional psychometric assessment where I need to do abstract, you know, numerical reasoning.

[00:34:51] And I'm going to ask my really smart sister who's studying math to help me do that.

[00:34:55] Yep.

[00:34:56] This is very relatable.

[00:34:57] It's what's called high-phase validity.

[00:34:59] So I think that has created a whole different set of behaviors.

[00:35:04] And not timing it is a really critical part of that.

[00:35:07] We often get challenged to say, well, you have to time and how can you test someone without timing it?

[00:35:11] But the whole point is we just want to get to know you and for you to get to know yourself.

[00:35:16] And unless the job requires you to act, you know, like that in a role, technology isn't used to high air traffic controllers where you need to make decisions in the moment.

[00:35:25] Yeah.

[00:35:25] You know, that speed is not important, right?

[00:35:29] What's really important is that you get the right profile, someone who's got incredible service orientation, who loves working with people, who wants to help people, you know, whatever the role requires.

[00:35:38] Yeah.

[00:35:38] And so we've seen the same with generative AI is that we're just not seeing the levels of people using it as you would expect.

[00:35:47] I think because, again, the experience just feels really safe and easy.

[00:35:51] We've just released our second version of what we call artificially generated content detector, which now has, I think, a 94% accuracy.

[00:36:00] And so we're very confident about its ability to pick it up.

[00:36:03] But the ultimate measure, Bob, is, you know, we're not hiring like we're screening and ranking and you're hiring.

[00:36:10] Yeah.

[00:36:10] And so if we put someone in front of you, which you're evaluating as the human, usually the hiring manager, and you go, this person is not at all what I expected based on, you know, the score and the profile.

[00:36:21] We'll see that in the data, you know, today.

[00:36:23] And so our measure of success is that we're actually effective as a screening tool across the board, whether it's detecting people who might be inclined to cheat or not.

[00:36:33] Eight out of 10 people that we put in front of you, you hire.

[00:36:36] You know, we had a client event last week and a really big luxury clothing retailer here said they had 40,000 applications for Christmas highs.

[00:36:45] They put all of them through Sapia.

[00:36:47] They didn't look at a resume.

[00:36:48] There was no human intervention.

[00:36:50] They took them through to assessment center.

[00:36:51] They were hiring a lot.

[00:36:53] So they had a lot of assessment centers.

[00:36:54] And they expected a 50% offer rate.

[00:36:57] They had an 80% offer rate from that.

[00:36:59] And so that's a blind screening process.

[00:37:02] So, you know, it has to work.

[00:37:04] Otherwise, why would you pay for something that where you only get a 50% yield?

[00:37:08] So that data point is a key one that we track from a validity perspective, real time for our clients.

[00:37:13] And they get to see those results too.

[00:37:15] And our benchmark is 80%.

[00:37:17] If you're hiring less than 80%, then we don't feel like we're doing a good enough job as a tool.

[00:37:22] That's a pretty high bar.

[00:37:23] Well, it's what you would expect though, because again, if I come back to what are we solving for?

[00:37:27] We're solving for, you know, time really, which is how do we ensure that you don't

[00:37:32] have to meet 10 people to hire one?

[00:37:34] Like that's such a waste of time.

[00:37:36] That was always for me the thing that I saw in recruitment is, you know, how much time, invisible

[00:37:42] time, the productivity tax of recruitment.

[00:37:46] And that's what we're trying to solve for.

[00:37:47] And that's not tracked.

[00:37:48] No organization tracks that.

[00:37:50] No one's going in and saying how many interviews are in Microsoft Outlook that we've done.

[00:37:55] You know, what is that cost to the business?

[00:37:57] That is the biggest cost of hiring.

[00:37:59] So if we can make a more effective conversion, because you're not having to do as many interviews,

[00:38:04] like, and then obviously you're not having to do as many replacement interviews.

[00:38:07] That's the ultimate measure of, I think, Sapia's impact.

[00:38:10] That last part, I think this is where I was going to go is, you know, you're also slowing

[00:38:14] the revolving door of the people that did accept an offer, but then two weeks later, they're,

[00:38:21] you know, they're gone or a month or whatever.

[00:38:22] So I don't know if it's different for high volume hiring or just more traditional, you

[00:38:29] know, knowledge workers or what have you, but are you replacing the application process

[00:38:36] or are people applying, going through that cat and mouse game and then, you know, encountering

[00:38:44] Sapia?

[00:38:45] So, you know, we're typically living within the workday success factors of the world,

[00:38:51] items of the world.

[00:38:52] And so whatever that process is to get in from a career site is the process.

[00:38:56] We don't affect that.

[00:38:57] If you're not using one of those, you can go straight from a LinkedIn job to Sapia.

[00:39:02] But generally we work with enterprise, so they usually have that system.

[00:39:05] But we are definitely replacing the 25 questions that usually sit in an application form and

[00:39:10] upload the resume.

[00:39:12] It's basically knockout questions within the ATS.

[00:39:15] Do you have working rights?

[00:39:16] Are you over 18?

[00:39:17] You know, do you have a responsible service of alcohol license?

[00:39:20] And then you get to Sapia.

[00:39:22] And the idea being put it all in the chat so we can ask questions around availability,

[00:39:27] you know, weekend availability, Christmas, whatever else is important for you.

[00:39:32] Do you speak three languages?

[00:39:34] These languages, because that's important.

[00:39:36] That can all go into the chat.

[00:39:37] It's both a structured interview from a science perspective in terms of what we're evaluating

[00:39:43] and then the ability to capture other data.

[00:39:45] And you can filter on that.

[00:39:47] So I want to only look at the people who are scored above 70, who can speak French, German

[00:39:51] and Spanish, and who have a forklift license.

[00:39:55] Show me that list.

[00:39:56] And then that turns up.

[00:39:57] And we have what we call a skin that lives within the ATS so that you don't have to actually

[00:40:01] go back in and out of the ATS to look at the profile of all of those people and very

[00:40:07] quickly move them forward.

[00:40:08] And then we've just introduced what we call live interview, which is the ability to move

[00:40:13] that person through to automated scheduling, but more than scheduling.

[00:40:18] So for me, live interview, which is our latest add on is not just about the scheduling.

[00:40:23] It's about how do we make that last event the most valuable event for both of you?

[00:40:29] So we have intelligence about the candidate that we share with the home manager.

[00:40:34] Hey, Bob, here are things you might want to test when you meet with Bob.

[00:40:37] And here's her profile if you want to take a look at that.

[00:40:40] We ask the candidate for feedback that then feeds into hiring manager feedback, right?

[00:40:45] So part of the challenge we find with enterprise is thousands of people are doing interviews

[00:40:50] who are terrible at doing interviews and don't know that they're terrible.

[00:40:53] So the ability to coach them on how to do a better interview through capturing that feedback

[00:40:57] at scale and using Gen.AI to summarize it, you're improving the capability of your hiring

[00:41:03] managers to be an interviewer.

[00:41:05] If you've asked for disability information or diversity, and I say I'm in a wheelchair,

[00:41:10] that will be fed through to the hiring manager so that you'll be told, hey, you should meet

[00:41:14] Bob on the ground floor.

[00:41:15] She's in a wheelchair.

[00:41:16] You know, it's really trying to bring, again, all the best of and the relevant information

[00:41:20] to bear on that final event so that that live interview is effective for both parties,

[00:41:26] which is way more than the scheduling.

[00:41:28] So we haven't built all of that out yet, but we've started to, and I'm very excited about

[00:41:34] that.

[00:41:34] And again, we're trying to bring a human lens to that final step of the journey to picking

[00:41:39] the right job and picking the right person.

[00:41:42] Well, and just overhauling the whole hiring experience, it sounds like.

[00:41:46] I mean, because you're hitting on all these other pieces that I think are adding tremendous

[00:41:51] value for all parties, right?

[00:41:54] You're streamlining the process.

[00:41:55] You're giving more insights to all of the stakeholders in the process.

[00:41:59] You're increasing the quality of hire.

[00:42:03] And the interview, any insights you can give to the interviewer or interviewee is just making

[00:42:11] the whole process better and potentially less bias.

[00:42:16] It's kind of symmetrical value.

[00:42:17] I think we underestimate how important it is to dignify the candidate, help them, guide

[00:42:24] them on this process.

[00:42:25] It's astonishing to me how poorly candidates continue to be treated by consumer brands, which

[00:42:31] usually the candidate is your customer.

[00:42:34] And you can quantify if you're a grocery retailer and you spend $2,000 a year in a store, that's

[00:42:41] probably really lowballing that number.

[00:42:44] And I go, I had such a terrible experience.

[00:42:46] I'm going to go to the competitor down the road.

[00:42:48] Is anyone quantifying that right?

[00:42:50] So I think that's why we love working with consumer brands because they get it.

[00:42:53] They get that experience is a consumer experience.

[00:42:56] It's not just a candidate experience.

[00:42:58] I think about in here in New York, the Department of Labor runs a lot of career fairs and stuff.

[00:43:06] It's like, come on, come all.

[00:43:07] We're going to meet at the county center.

[00:43:09] There's going to be these 20 employers coming or whatever.

[00:43:13] Or bring a smile and 20 copies of your resume.

[00:43:16] Like, oh, we're doing that now?

[00:43:18] We're going back to not just resumes, but paper resumes.

[00:43:22] Bring your resume so someone can hold it.

[00:43:25] So look, if you want to try to do that, and there's probably a lot of people there, and

[00:43:29] maybe think of some questions on the fly you want to ask them, like, great.

[00:43:33] But you could probably just set up a couple kiosks with a couple of Sapia licenses and just

[00:43:39] have people take some time.

[00:43:41] I guess since you have unlimited time, it might be a little challenging.

[00:43:46] But it could be a take-home.

[00:43:47] If you're interested in these five companies, just send me.

[00:43:51] Here's my email.

[00:43:52] Yeah.

[00:43:53] The other thing is if you're advertising it, put a QR code on it, and the QR code can be

[00:43:58] scanned and be directed to Sapia.

[00:43:59] Absolutely.

[00:43:59] I think we work with organizations hiring on campus where, you know, why wait until

[00:44:04] people apply?

[00:44:06] Why not get people to just scan?

[00:44:07] And now you've got a ready-made talent pool that you can go and ping from, and you can

[00:44:11] say, here are the best.

[00:44:12] Let's go and schmooze them.

[00:44:13] So we make sure we win them when it comes to, you know, the recruitment process kicking

[00:44:17] off.

[00:44:18] It so changes the speed with which you can get to great talent fast if you think about it

[00:44:24] as an experience that people want to do rather than a recruitment process.

[00:44:29] And again, you know, we tend to work with transformational leaders who want to really think differently.

[00:44:34] And, you know, that's one way that is super powerful.

[00:44:37] Now you can send this link out to everyone on campus and say, whoever wants to do this,

[00:44:42] you're going to get feedback.

[00:44:43] You're going to learn.

[00:44:43] It's going to help you in your career.

[00:44:45] And then the downstream impact is the company gets a qualified, you know, pool of talent that

[00:44:51] they can go and work with.

[00:44:53] Think about how that changes the game in terms of getting to talent first.

[00:44:56] Well, plus the time, like even if, let's say you took an hour to answer, you know, these

[00:45:02] three rich text, you know, questions, the alternative would be for you to go customize your resume,

[00:45:10] send it into the application process.

[00:45:12] And then if it's a workday, fill it all out again.

[00:45:16] And oh, by the way, once you submit it, you might not hear anything back at all.

[00:45:20] So how does that help you do better next time, right?

[00:45:24] Yeah, look, I don't think there's a battle to win candidates over.

[00:45:27] Candidates have loved it and won over.

[00:45:30] It's the recruitment team that wants to run things as a process and quite transactional.

[00:45:36] I mean, the other area as well for us is, you know, why go to market again for talent when

[00:45:41] you've already got a database of talent?

[00:45:42] So why don't you just use search in our platform to say, can you find me people who are really

[00:45:47] strong on these dimensions, who can work Fridays and Christmas, who were scored above a 60?

[00:45:54] You know, it's effectively, we've built a CRM without meaning to.

[00:45:58] You shouldn't have to run a recruitment cycle again.

[00:46:01] You know, so how do you just kind of disrupt the machine and go, we may not need that machine

[00:46:06] anymore.

[00:46:06] Like we can solve the business need in different ways.

[00:46:11] Let's reimagine how we do that.

[00:46:13] That's the power of using a technology that's truly AI based.

[00:46:17] Yeah.

[00:46:18] Where you have that data set.

[00:46:20] Anything that you've been particularly interested in, in terms of, you know, this particular topic

[00:46:26] or this particular use case has caught my attention?

[00:46:30] So much.

[00:46:31] So much.

[00:46:32] Pick one that's either like fascinating or concerning.

[00:46:35] Look, I think the fascinating one is this idea of moving away from platforms.

[00:46:40] And how do you, and I think for HR, it's, it changes the whole role of what, and the data

[00:46:48] that I get to see.

[00:46:49] Like right now, HR gets access to a whole lot of data.

[00:46:53] Engagement survey results, you know, exit survey results, performance management.

[00:46:57] What happens if none of that exists?

[00:46:59] And people are just getting on with it.

[00:47:01] And the technology is enabling them to do that, you know, through clever chat based tools that

[00:47:06] are helping with 360, helping with learning, helping with development, helping with identifying

[00:47:10] my next role.

[00:47:12] You know, that's very challenging as a mindset, right?

[00:47:16] Like what is my role?

[00:47:17] The other thing that's challenging is the data that gets retained.

[00:47:20] Because what I say to HR leaders, and they don't like to hear us, is why can't I delete

[00:47:25] the data once I've shared it?

[00:47:27] If I'm going to give feedback on Bob, I want to delete that afterwards.

[00:47:30] Because the end goal is that Bob learns from that.

[00:47:33] And he grows as a leader.

[00:47:35] Why do you need to retain that data?

[00:47:37] And so the power of data ownership going to the employee, I think is the next phase of

[00:47:44] where we really need to move to.

[00:47:46] Where I have control over my data, I can just press delete.

[00:47:50] And you're the leader.

[00:47:51] I want that report.

[00:47:53] I want to retain it.

[00:47:54] No, I don't want the organization to see that.

[00:47:56] Why do they need to see that about me?

[00:47:59] So suddenly HR becomes sort of invisible, like as in the way things are working is quite invisible.

[00:48:05] That whole question around data privacy, data ownership, you know, when I'm giving the

[00:48:09] data as the employee and that data is helping you, why does HR need to be in the middle?

[00:48:16] You know, that is a conversation that's really people are very confronted by.

[00:48:20] Because I think the whole point about generative AI and what we're building in FI is it gives

[00:48:25] people agency.

[00:48:26] I don't need HR anymore.

[00:48:28] And so why do you need the data?

[00:48:29] Because I think data control, data agency, data ownership, data privacy are such big

[00:48:36] issues and will become bigger and they should be, you know, and that forces HR to think about

[00:48:41] what's the end goal rather than, you know, tracking what's going on along the journey.

[00:48:46] Let's just focus on measuring the end goal, which might be retention.

[00:48:49] It might be, you know, employee brand advocacy on LinkedIn.

[00:48:53] Not how we get there.

[00:48:55] You're no longer orchestrating that.

[00:48:57] The individual is orchestrating it, not HR.

[00:49:00] Yep.

[00:49:01] That's really intriguing.

[00:49:03] I've had those conversations sort of externally, like for example, in the context of like freelance

[00:49:10] marketplaces, right?

[00:49:11] Everyone wants to be sort of contingent and fractional and all of this.

[00:49:15] And at some point, the projects you've worked on, the skills you've attained, the certifications,

[00:49:20] all of that would be nice if you, you know, owned all of that and you had that sort of control

[00:49:26] and portability.

[00:49:27] But I hadn't thought about it inside an organization.

[00:49:30] I'll have to think more about that.

[00:49:32] The very last question I had for you, Barb, was just any advice for folks trying to get

[00:49:37] started with AI literacy and readiness?

[00:49:41] Any advice?

[00:49:42] We've put out a lot of papers around that.

[00:49:44] You can find them on our site, you know, AI bias guide.

[00:49:47] I think the other is TriFi.

[00:49:49] It's on our site.

[00:49:50] It's free.

[00:49:50] So you get to see how clever technology is these days and play with notebook LM.

[00:49:55] You know, when it first came out, I was pretty obsessed with it.

[00:49:58] And you'll see how learning is completely reimagined.

[00:50:01] Like why would you ever put people through a one-way training program again when you can

[00:50:06] experience something so much more powerful like a podcast to learn?

[00:50:10] You know, I've put all of our clients' career sites in there as a tool for them to learn

[00:50:14] about what it is that they're strong on, what are they weak on.

[00:50:17] I mean, it's extraordinary.

[00:50:18] So that way people start to realize the power of this and they get excited rather than being

[00:50:24] fearful.

[00:50:25] You know, how do we get people excited about what this can do?

[00:50:28] Not scared.

[00:50:28] Absolutely.

[00:50:29] Love it.

[00:50:30] Barb, it has been a pleasure as always.

[00:50:33] I can't believe how time flies.

[00:50:35] Yeah.

[00:50:35] I really appreciate you taking some time to talk to me and catch up and a lot of insights

[00:50:40] for my audience.

[00:50:42] So thank you so much.

[00:50:43] It's a pleasure.

[00:50:44] Thanks so much, Bob.

[00:50:45] I'll chat to you again soon.

[00:50:46] Thanks again, Barb.

[00:50:47] Thanks everyone for listening.

[00:50:48] We'll see you next time.

[00:50:49] Bye.