Webinar: Creating an exceptional candidate experience for your emerging talent in the age of AI

[00:00:06] Hello. Good afternoon everybody, and welcome to today's ISE webinar really interesting subject we're covering today. We're gonna support the golden thread, the practical steps you need to take to create an exceptional candidate experience in the, in the age of AI. And I've got two excellent people joining me today. So we're gonna have some really rich content and some really strong discussions as well to go alongside it. Just a quick reminder about how our webinars work. We like them to be engaging as possible, and both Robert and Sarah are both, you know, very engaging people. [00:00:38] So pop any chat points into the, into the chat box, more than happy to get a conversation going. Also if you've got any specific questions you'd like to ask, please plop those in the Q&A. It's easier for me to monitor them if they're in the Q&A box. So any questions you've got for Robert or Sarah or myself, pop them in there and we will get to them. If we don't tackle them as we go through the session, we've definitely got time at the end to cover them off there. In a minute or two, I'm just gonna give you a couple [00:01:04] Of points, I guess from the, from the ISE just about where we see the landscape at the moment, coming from the forums we have. But before we do that, should we get into some introductions? Sarah, would you like to go first? Yeah. Yes, I'll go. For those of you who don't know me, I'm Sarah Wardle. I lead on new relationships we have with customers at Eli focusing particularly on early careers, having had over 20 years experience in the industry. And those of you who haven't heard of Eli it's experience led onboarding technology designed to set your new people up for success [00:01:37] Works particularly well for the early careers audience because it creates this community for your new hires before they've even joined your organization. So Eli, your connections with your business embeds your values and your employer brand and empowers your managers and early career teams to be, to be confident in managing the onboarding process. So really excited to be here today and looking forward to guiding you through this topic. Great, thank you Sarah. And I'll hand back to you shortly. Robert, would you like to say hello? Yes. Good afternoon everyone. So I'm Robert Newry, co-founder of Arctic Shores. [00:02:12] And since January Chief Explorer which I'm reliably informed is one of the coolest titles, out there, but it gives me the opportunity to talk about Arctic Shores mission more than I had been previously. And so, Arctic Shores, for those of you that aren't familiar with us, is a task-based psychometric assessment company. So my principle and philosophy when setting up the company was that you understand people in a more authentic way when you observe how they do something rather than how they tell you that they might do something. And so over the 10 years, that's been our mission [00:02:58] To uncover potential through the way that people do things. We've worked with 350 or so organizations in that timeframe. The likes of Siemens, Amazon, Molson, Coors screen 3 million candidates and help them get into 75,000 or so jobs. And one of my favorite topics to talk about now is AI and the impact of AI is gonna have on recruitment selection. So very much looking forward to talking to you about that. [00:03:31] Fantastic. Thank you Sarah. Thank you, Robert. I'll be handing back to you guys very shortly. So my name is Stephen Isherwood. I'm joint Chief Executive of the ISE. So I'm responsible for knowledge and insights. And just to give you a bit of our perspective from actually all those forums we run from the reports and surveys we do. Just to add a bit of bit of color to start the conversation off. So firstly, thinking about AI, obviously it is a significant disruptor, in the marketplace. What we are finding from employers is actually not many [00:04:06] Are using AI in their decision making process, but actually what's vexing employees is how students are using it. And that's very much a hot topic at the moment. And of course a very evolving topic because it's something that's, you know, that isn't a settled, I guess a settled issue in terms of how students are u are using AI and how employers should deal with it. So in our last recruitment survey that we did the backend of last year 43% of employers said they were okay with students using AI in some way in the selection process. [00:04:43] But there were 32% who actually recommended that students don't. So it's, it's quite interesting from a, if you put yourself in the student perspective, actually, how should they deal with AI in the, in the selection process. I guess there's a certain amount of, uncertainty there. And that's one of the things we're gonna get into shortly, is around actually how employees are responding to that or how they might respond to AI being used in the selection process. But yes, very much a very much a concern. And I guess along with that is the application volume used coming through. [00:05:14] Again, if I go back to our last recruitment survey record numbers of applications per vacancy, which of course for employers is a big challenge 'cause they've got to operationally deal with that volume and make selection decisions. But also, again, thinking about the student perspective you know, that means we were rejecting ever more students than before. Now in some ways, that application volume is a positive development because actually employers have reduced barriers to entry. Using technology much more in selection means that kind of, some of those artificial barriers, IE had to have X number of a level grades or you had to have a two on [00:05:51] Before you could even apply. They've disappeared, but those volumes are really, are a challenge from the employee perspective, deal with 'em, but also I think from the student perspective, actually dealing with those level of rejections and navigating those and our sense that AI is only fueling actually that challenge. So I think that's something we'll, we'll, we'll get into a little bit as well. So of course alongside those volumes, again, another big topics on all the, all the forums that we run are Renee greats. I've got quite a few gray hairs as you can see. [00:06:22] Renee weren't really a challenge, I guess when I started this game. It wasn't something that we, that we, that we thought about, we never would've over offered in my previous roles. 'cause you'd be scared they all turn up [00:06:34] And you've got too many people for the business to deal with. But actually, I know that's quite common for employers now, sometimes over offering our data again shows that 7% on average is the level of Renee. That's where a candidate has accepted an offer and then declines that offer at a later stage. Ghosting in the selection process. I think that's very much tied to, to the volume of applications candidates are making. So employers tell us around 10% of their applications, when employers try to get back in touch with them [00:07:07] To arrange interviews, et cetera, about 10%, they just never hear from. Again. So there's some big challenges and big themes there in the, in the selection process. Also at the development end of the spectrum. So as people go through the onboarding piece and start with organizations there's lots of conversations around work readiness work readiness affected by the pandemic. 'cause students of course, have had some very different experiences over recent years about how they might have done part-time work or the things they've done extracurricular, but there are some long-term trends. I'm always interested in the work [00:07:44] Of Bobby Duffy who's coming to our next development conference. 'cause he talks about delayed adulthood. You know, people leave education later. They start work later in life. They stay at home longer, all that means, and actually I think if we talk, we use that dirty word average is that actually your average 21 tenure year, 22-year-old is probably a different place. So they will have been 10 or 15 years ago then, and that's playing through as well. So it's something we definitely need to think about in terms of how we transition people through the selection process [00:08:17] And into work. And then of course, overarching all this at the moment lots of employers tell us in fact everybody in our sector tell us, actually, I won't say it's just employers. You know, we are in a pretty tough economic climate at the moment. There is low growth, budgets are under constraint, vacancies are holding up pretty well according to our data. But teams are operating on tighter budgets, which means that workloads are greater. And of course you add that challenge of budget. Add to that, the challenge of dealing with the complexities driven by AI [00:08:50] And some of those other forces we've talked about. And it feels a pretty tough market to be operating in at the moment. Most definitely. So, some key themes there around use of AI, what's happening with the candidate experience, what's happening as candidates come through that selection process and join in organizations. Lots to get stuck into that. And who better to take us through that than Sarah and Robert? So, reminder, please use the chat. Please use the Q&A. Sarah, would you like to take us from here? [00:09:18] Yeah, so actually I'm gonna hand over to Robert Takeaway, not past the buffer or anything. Sure. I'm gonna take you over to Robert to start us off. Brilliant. Thank you Stephen. Sort of great setup on some of the big challenges. And I'm gonna focus on the first one, which is the use of AI by candidates. And there are two elements to this. One is volume and the other is integrity. And today, I'm, I'm going to in focus on integrity and share with you that actually this is, this is a bigger problem than many of you probably realize. [00:09:59] And so there's a lot of talk about what the impact of AI will be, but I don't think there is as much realization as to how much is going on a amongst the student population in terms of the level of AI that they're using and how they're using it. So here's just some stats from a survey we did a year ago. And the first one is not particularly you know eye popping in any way but the second one is, and so 86%, this was already a year ago, described themselves as proficient in the use of AI. [00:10:40] So it's not just that people have access to gen AI tools and that they're using them regularly, but actually they're becoming really good at using them, and that's only going to get better over time. And we know that they're using it students using it for coursework. And so it's no surprise that 60% of them nearly are now starting to use it as part of the recruitment process. So the activity level is high there. And how then are students using it and what's the implications for integrity. So if we go to the next slide one of the things [00:11:22] That I like to share with people is that chat, GPT is like a calculator. So here, now you have an example if you click on it, thank you, Sarah, of ChatGPT on a phone, on a student's phone, just saying, you know, imagine you're applying for a job and you've been asking jt which many people do or have in their recruitment process. And chat g PT will help you in the way that you do that. So this is, this is getting round any of those copy and paste deterrence that are in some applications just using your phone. [00:12:00] And you can see taking a picture here of the instructions and immediately interpreting those instructions. It's then being moved onto using, that's the start button to the next bit on this. So what are the options that you have to do in this scenario? It's, it's already read those, it's given the scenario, it already knows the instructions, and now it's looked through the options and it's given you the most likely action that you should perform and the least likely. So it's not just able to make a selection. It's actually able to provide some reasoning behind that. [00:12:41] And it gives you the reasons to why this is the most likely action to perform. And this is the least likely. And if you wanted to be more sophisticated, you could easily have added an additional prompt into chat GPT here, which is go to the x, y, Z company career site and review the behaviors that they say that they're looking for, and use that to help answer the SJT questions. So it is that easy to be able to use it, and it doesn't require any great sophistication. So that's one challenge. The next one video interviews. [00:13:24] The, there isn't a early careers recruiter that I'm not speaking to at the moment that isn't saying, Robert, it is a nightmare at the moment with video interviews 'cause people are reading the answers out of ChatGPT, when they're responding to them. Now, that is at the lowest end of the scale. One of the examples you see here on the left that did the LinkedIn round a couple of weeks ago was that the character on the bottom left there is actually an AI avatar. And the face had been superimposed on top of the candidate there. And the recruiter spotted it. [00:14:06] And the recruiter pointed out this is the second time where they had interviewed somebody and they didn't feel comfortable that the person that they were interviewing was actually the real person, that they were talking to. Now, if you thought that was bad enough, I'll show you the next video now. And if you click on it And you can see I'm currently participating in a Zoom call, but what if I tell you that I'm actually not in front of the camera right now, and this is an AI generated live screen that moves in sync with my voice. [00:14:39] People generates real-time video of talking to you based on your audio inputs, which can be streamed directly to Zoom. How about that? You know, we kind of think, is this out of, you know, black Mirror or something? No, this is, this is something that you pickle is something you can go and pick up from the internet, and now you can stream your video as if you were talking to the camera and sitting down when in fact you are wandering around. So this technology's out there. Does it mean that everybody is cheating? Absolutely not. But what it does mean that it's incredibly easy [00:15:14] For some people to cheat. And if it's incredibly easy for them to cheat, then it will become more prolific. And so the argument that I often get presented with is, oh, Robert, there was cheating all the time. And so the integrity of the process that we have in recruitment is no more or less vulnerable today than it was yesterday. It's just not true. The integrity of the process is much more vulnerable now because you have all these tools out there that make it super easy. [00:15:47] And there was just another one recently as well from Columbia University that hit the news at the end of last week. A guy called Roy Lee had a, you can either put it as an earpiece or you can actually have it as an invisible, overlay on your video interview, giving answers to technical coding questions. And Roy had managed to get a summer internship with Amazon, Google, and Meta by using a tool to give immediately the answers live with an interview on a coding challenge. So the things that we thought were sacrosanct and not vulnerable are increasingly so. [00:16:35] And we need, we need to be aware of that. So if we go to the next slide. Now this is, this is a piece of research that Al Bourne my old friend Al Bourne who used co-founder of Sova now working for on Marty as a consultant. And he looked into this and just did some of the numbers on this. 'cause he wasn't convinced the assumption that a lot of the psychometric assessment companies out there are saying, well, we're not seeing the averages changing very much, and we've got hundreds of thousands of people going [00:17:15] Through our assessments, and the averages aren't really changing. So this idea that cheating is going on, or manipul manipulation is going on is just not evident in the data. If you just sort of go back a second on this one, the reality is that if you look in the orange there, those are people who are manipulating using AI and they will give themselves a half a standard deviation uplift. And so it's normally distributed. The assumption is that anybody who's using AI will automatically get the best result, and therefore it'll be obvious is not the case. [00:17:53] The evidence shows that people who are using AI to manipulate their results, it's normally distributed. So if you look at the red line here, and this was Al's point about this, which is the cutoff that many people, you know, one standard deviation's about 70% of people who are doing a psychometric assessment, who will be screened out, 30% moved onto the next stage. So if that is the typical cutoff, if you look at the orange, in there, those people are now already a disproportionately large element within that group of people who are now being moved forward to the next stage. [00:18:31] And if you click on the next bit on this one, Sarah, the hi, his modeling of this showed that 30% of the candidates who get to the last selection stage could have used AI to enhance or support their results to get there. So we're not talking about a small percentage here. We are talking about small percentage who can then become a large percentage because of the advantage that they're getting. And so this is much more significant than probably many people realize and why. And I put my last little sort of spin on this [00:19:12] Before I hand over to Sarah on this, why is this so important? It's because as we think about Renee and that, and Steven talked about that, it's become much more of a, of an issue. Now, when you look at this Roy Lee example from Columbia University, he got three job applications by using or sorry, job offers by using AI to enhance his results. [00:19:42] So he was able to find a way to cheat the system and get multiple offers and that will then play out in terms of Rene's in time as well. And so this is important not just in terms of the integrity. Who are you hiring? Are they as capable as you expect them to be? But also the people who are using AI to enhance their results are doing it in multiple job applications and are then giving themselves the pick of the crop to then move forward, which will only then increase your NA rate. So there are all sorts of implications from, [00:20:22] From understanding whether the integrity of your process is as strong as you think it might be, and it probably isn't. Over to you. Sarah. Thank you. That was a perfect transition because I wanted to touch upon this impact of hiring agan attrition rates. So both Steven and Robert have highlighted these as being, you know, a really big issue. And many of you on this webinar will be experiencing this and wondering, you know, what can you do about it? Recent data indicates that the concept of graduates remaining on accepted job offers is becoming, [00:21:02] You know, more and more prevalent as you've heard. I think a number of different STAs from different places have slightly different numbers. But as of November, 2023 the Renee rate in the graduate markets to the 12% with certain sectors experiencing even higher rates bright networks, that was interesting here. They reported last year that 57% of students would reject an already accepted offer if a more appealing opportunity arose. And this was up from 50% in 2023. This is an enormous number of people who employers may have thought were safely hired you know, in the bag only [00:21:40] To be rejected the last minute and left, you know, pretty much high and dry. Also very popular stat, often quoted, but LinkedIn reported that 22% of turnover occurs within the first four, five years. So once someone has actually joined they aren't always in that long either. So this is really frustrating for employers. But it's also costing far more than they realize. And I want to take a quick example here actually, because we can start to calculate this real cost of re eggs from first year attrition. We know it's a problem, we've all agreed [00:22:19] That's a problem. But if we assume for this example that the cost of, you know, per hire of a graduate earning an average annual 13 salary of 35,000 pound is 4,758 pounds. It's based on ISE data. And this business is hiring 200 graduates per year. This is a completely made up business, by the way, just for the example. The cost of hiring is going to be about nine 50,000 pounds, which is enough money in itself. [00:22:49] We need to also take into account a dropout rate. So if we take a conservative about rate before day one of 12% based on data from GT I, now this could be lower or higher depending on the industry. I know it's higher in industries like engineering and energy. But if we take that as a kind of conservative estimate, I mean to think about retention rates, well and again, these will vary depending on industry location, role type, that type of thing. But again, if we take a conservative estimate of just 10% of graduates leaving within the first year, [00:23:25] And again, it might differ depending on what industry you're in. But for the purpose of this example, all of these things are true. Then the actual number of hires this example organization has to make per year is actually 244. [00:23:40] So cost of hiring is 44 people who didn't even start or who leave within the first 12 months is over hundred thousand pounds. You know, the business is spending over 200,000 pounds on hiring graduates who either don't start or leave within the first year. And this is just the recruitment costs. To be honest, it doesn't even include the training and onboarding costs once someone's joined and all the internal effort that goes into someone in the first year of their career. So we obviously want to avoid this. And there are obviously many ways to do this. [00:24:15] But if we create a better golden thread, as we started talking about at the beginning of this presentation, and we're going to a bit more detail about this, throughout this webinar. If we create a more engaging onboarding experience as well. This example business could easily half that dropout rate and reduce the attrition by just a fifth. I've used those numbers 'cause that's actually the average reduction our clients do using Eli. So it's a good kind of rule of thumb. We think admin time costs of a very conservative estimate, you could reduce maybe by around a hundred pound or higher. [00:24:50] You know, it could be a lot more. And with a more prepared and engage graduate, you probably expect their time to productivity to be reduced by reduced a lot more than a day. But for conservative reasons, I've left that as well. If we did all of those things and, you know, there's a lot of ifs there, but if we did all those things, this business could save over a hundred thousand pounds by having a more engaging onboarding experience, having this golden thread and therefore reducing dropout and attrition. So I thought this was a useful exercise to do [00:25:22] And I would recommend that all of you know, go away and play with your own numbers because it does help to plug in your numbers and see what effect they could have. [00:25:33] Another challenge that Steven highlighted at the start Bobby Duff and Buff Bobby Duffy's talked about it too, is this impact of delayed adulthood or this concept of delayed adulthood. This is where young people are taking longer to reach these traditional milestones than previously. Perhaps home ownership during employment, leaving home [00:25:56] Reaching financial independence, changing career expectations. All of these things are having a real knock on effect. For us as graduate recruiters. You know, there's an increase in Renee turnover of grads more uncertain about what they want to do. We have to spend more time building up relationships and connections with our future grads and, you know, try and use this longer lead time that we often have between offer and start date to help build these students and grads are needing more support and help with transitioning from school and university life into the working world. [00:26:34] And we can no longer assume, as Steven said, you know, that people will know etiquette, work etiquette, you know, etiquette of meetings and writing emails turning up on time, dress code even things like virtual meetings have different rules at school and at work, which most of us would know, you know, cameras on or off it's, it's very different and the environment's extremely different. But this long lead time that we often have between offer and start date offers like the perfect opportunity to give support before they join and keep them engaged. [00:27:09] You could even offer the opportunity for parents to get involved and engaged during this window, which would reduce stress and anxiety especially for that younger apprentice audience. [00:27:19] So I guess before we go into a short poll I would ask you all, you know, would you rather do all of this before someone joins you or have to spend their first few months in their role getting them up to speed? So to think about there gonna jump into you, a call now because we've talked about quite a lot of different challenges and it would be really nice to understand, you know, who we've on here. And if you can highlight your biggest concern in early careers recruitment right now. We've got several different options. [00:27:53] If your this challenge isn't there, then please feel free to let other and write the answer in the chat. So hopefully we've got that up there. Is that up there, Wally? I can't see it personally. [00:28:14] I'll go through them slow while we're waiting. So we've got volume of application it constraints interview and assessment process, length of onboarding compliance with AI, legislation, number of reneges or under-prepared new hires. [00:28:33] And then I can see that in there now. So really interesting to see what you guys, are struggling with and see if this aligns with our other people on the call as well. [00:28:51] That's probably way of me seeing what's coming in here while he isn't there. But yeah, I can't rely on you. The poll is still going on, so when it's done, I'll just yeah. With you. Fabulous. Thank you. That's great. We've got some coming through actually on the chat as well, which is really useful. [00:29:15] Steven, which is the one you are seeing most? I don't want to influence people's answers at all, so maybe we could wait until we finish. Well, it's like, I don't, there are an eggs bit is just something that, I mean, you'd always have one or two people, there are some problems close to the point. Mm-hmm. But whenever I hear employers talking about, so I've seen Rachel's comment in there about, you know, defense, you know, my background, professional services, regulated industries, you know, there's stuff people don't just walk through the door and you know, you've gotta kind of do stuff beforehand [00:29:45] And then, and it's amazing how just hearing employees tell stories of people who just don't turn up on the day. And it's interesting, I did a tried to do some press on this a few years ago now, but the pushback from the media was, well, you can't tell students not to Renee. It's a free market kind of. Yeah. You've gotta point the challenges of course, that it too close to the why employers can't back backfill those positions. So actually, you know, jobs go un fill [00:30:12] And then you've got your answers up now. So we Have got our answers up over the top. That is volume of applications, which we did talk about right at the very beginning and closely followed by under-prepared new hires, which again, we talked about too. And a few o other things have been mentioned in the chat. Budget constraints as well is a high one. Is that kind of what you guys were expecting? Steven and Robert? Is that, I mean, that to me, I'm not surprised about the volume. That's, that's something that we're all concerned about. [00:30:45] I definitely expected it to be volume and, you know, that's fueled by both the change in the job market, but also the ease of which it is to apply now with some of the AI tools out there. Yeah. What about you, Steven? Yeah, and I linked the volume to the under-prepared piece as well. I got sent an academic piece of research that was done recently. The numbers involved was pretty small because it was interview based, but the theme in it was around [00:31:22] Students using the application process as part of the career decision making process. So if you think historically the idea was, you know, I research a business or an organization a career, then I apply. So as an interviewer, you expect them to show why they want to work there. And I think it's linked to the volume where they have to do so much volume. It's been like, well, let me see if I can get through the first stages. Mm-hmm. And if I do, then I'll put some effort into because, you know so they haven't got the time to do that. [00:31:53] So there's definitely a, I think a link between those two that's under explored. And for us as a sector, I think this is one of those wicked problems whereby we're gonna have to get our heads around it because there isn't an easy answer when I talk. Some employees are saying they're, you know, they might be increasing some of the cutoff points or they might be shortening windows and they don't really want to, because that then affects diversity of applicants. But then if you are just filling this hopper with thousands of people, you're gonna say no to. [00:32:23] Mm-hmm. So yeah, I don't have any magic solutions to give, but I think it's something as an industry we're really gonna have to try and tackle and get our heads around. Yeah, absolutely. Yeah. Well, good news is, well, we're gonna move on now. Some of those technical steps that we can do and that Robert and I think are relatively straightforward things that you can start doing, maybe, you know, even tomorrow, to create this a more scalable candidate experience. So I'll hand over to you, Robert, then. Brilliant. Thank You. So I'm always keen to make it practical and all of this, [00:32:54] And so I've got three areas here. So if we go into the first one around this in terms of things that I think everybody, should be doing. So the, we talked a little bit about, you know, a golden, AI thread here. And I think this is so important, thinking about from the time that somebody goes onto your career website and they want to know not just how they can use AI, but also how you as the recruiter are using AI. And I think Shu Smiths have done a great job on that one. [00:33:40] I've, I've done a podcast with Samantha Hope, who's head of early careers there. She talks about some of the changes that she made, but it has to flow all the way through the recruitment process. There's no point tinkering with all of this and thinking, oh, I've given a little bit of candida advice, on the website. Important though, that is, but as we just were talking to there, well, what about if you are using application questions? How are you using those and what sort of guidance you are you giving on those? Then you've got, well, how do I do that first sift? [00:34:16] Then you've got video interviews as well. And I think there's ways that you can guide candidates on that. There's also the type of questions that you ask them, but I think it's how you set them up for success. [00:34:32] And rolls Royce, I spoke recently to the brilliant Ellie Long, who's global head of emerging talent at Rolls Royce. And her philosophy is how do we set candidates up for success? And Steven alluded to this earlier, if we are just a rejection machine on this, and we are not giving good guidance all the way through to help students work out whether they are the right sort of person to be applying in the first place, but also how they should apply and how their strengths and capabilities can be brought out in this process then we're not doing the fairest job that we could do [00:35:22] As recruiters. And I, and I really like that philosophy, which is why I talk about as a golden thread. How do we not just think about this as a rejection machine, but more about how do we set candidates up for success knowing that there is a great volume out there? And so we've gotta do it in the most thoughtful and fairest possible way. But that thinking all the way through where and how can students use AI, but also how you are using it within your process. These are the two probably most important pieces of guidance [00:35:59] That I think every re particularly early careers recruiter should be working through for their upcoming campaigns. So that's the first bit. The second bit is, I think as early careers, and in fact anybody in recruitment, you should have your own AI principles and requirements. And too often, I'm sort of hearing at the moment, and it's understandable that when I talk to early careers teams and say, well, what is, what's your AI approach and what's your AI strategy? And they go, oh, well, it's not our decision, Robert, that's governed by it. And they've given us a bunch of [00:36:42] Ethical AI guidelines. And so we're following those. But the it, you're looking at this entirely from an employee perspective, whereas in ta we've gotta deal with candidates around this. And so it's really important that we're thinking through what are the implications of AI, not just in our own internal use, which the IT teams are giving us guidance on, but externally, how might candidates use it? How might we use it to assess candidates too? How might we as TA teams use technology to support us with the efficiency we're gonna need? You know, Steven talked earlier about [00:37:26] Shrinking budgets. We saw that in some of the poll results too. So you've got increased volumes, shrinking budgets, you've got to find ways to help TA teams go through this sort of tsunami of applications. And so there will be an increasing interest in using AI tools around this, but it has to be very carefully done. And I've heard all sorts of stories about people, well, I've just put in a bunch of applications into ChatGPT and I got ChatGPT to rank them, and it did a pretty good job. It's just full of bias potentially in there. [00:38:04] There is black boxes in sorting that's going on, you know, and when it comes to the EU AI Act, you're in big trouble if you are using AI that isn't transparent and auditable within the way that you are using it. As a final piece on this, if you want to go and look I did a an interview with the amazing Lauren Gladwell, who's head of AMEA apprenticeships for Amazon, has been very thoughtful about using AI from a TA perspective. And there is a blog on which you can see the link here on the ISE website about how Amazon have created their [00:38:51] Their own principles and requirements on this. So only a 15 minute or so read really worthwhile. And then the final bit the advice that I like to give, you know, we talk about the golden thread, [00:39:05] And it's really important that when you've got that golden thread as a process that you've worked through from the moment you are attracting candidates and how you are giving them guidance on good use and bad use of AI all the way through the process to onboarding that you design the process, and then that you are using technology elements then to support that process to make it scalable, repeatable, auditable. And where I see a lot of people starting to think is, oh, I've got a volume problem. So I'm gonna chuck a piece of technology [00:39:44] Into the sifting as a way of addressing it, as opposed to, actually, I've gotta really rethink the whole of the process here. I've gotta get that golden thread sorted out, and then I will go out and should go out to look at what elements of my process that technology can improve, but I'm doing it against a process that I'm comfortable with. So those are my three practical tips for everybody to think about in terms of their next recruitment round over to you. So thanks, Robert. I've, I've been tasked with the, my fourth, piece of advice. [00:40:21] And I suppose, you know, my section is more on the onboarding side, but I think there's, it's really important to be able to create personalized content and learning. So I'm not just talking about technical learning but life skills, you know, all of these challenges right at the beginning, what we were talking about, if we prepare people with more life skills, they're gonna be ready for the workplace. You know, tailored content builds knowledge understanding where individuals need it most. It'll embed your culture and your values and prepare, you know, these new people for work. [00:40:58] In, you know, improving the early performance, which is so important, and reducing anxiety, which is, which is ever increasing. Training and pre-learning allows candidates coming from a school environment to familiarize themselves in this new environment before day one. And this is, this example here is NatWest, you know, organizations like NatWest do this for their, for their work experience, interns, apprentices, and graduates. And it's just an essential part of their recruitment and development process all the way through, no matter what stage of early career they're joining. And they see it as a, you know, integral part of, [00:41:36] Of how they prepare a new joiners, start work with them. And my fifth example we actually haven't talked much about it much today, but I think these, this element is so crucial because a lot of the stuff we've been talking about, it is very theoretical. But in actual fact managers are aware a lot of this responsibility, sit, responsibility sits. So my last tip would be to ensure your managers are supported. That's, that's something we can do. So the importance of managers can't be underestimated. They're often the most important person for a new hire, [00:42:19] And someone they want to engage with before day one. But managers as we know, are very busy people, often hard to get hold of. And they're often not experts in onboarding specifically. So try and ensure your managers have this tailored support knowledge and reminders they need to create and engage in onboarding experience. So, you know, it sounds obvious, and you guys are probably doing a lot of this already, but giving the manager training and development so they can provide their new hires with a better experience. You know, creating a simple checklist, you don't even need technology necessarily for this. [00:42:57] You know, you can just create a simple checklist with alerts ideally, but keeping the manager on track you can see from the example on screen the manager has a personalized experience too. You know, they're learning how to do onboarding properly. They can keep track of their new hires, see how engaged they are they can connect with them and manage their onboarding experience really easily. So I think not direct making the importance that the manager is kind of my fifth. And the, as you can imagine, lots and lots of things you can do in this space [00:43:32] To make that easier for yourselves. And I'd just like to finish really on a, on a couple of stats. If onboarding has done well, following the concept of this goal of thread and employer performance will increase exponentially. Not just the performance of new hires down to training, but also this effort that your graduates put into their careers with you. And that's something we all have to remember when we're going through all this process. [00:44:02] So, yeah. I guess let's go to Q&A. It'd be great to see if we've got any questions come up, from the Q&A or the chat. Steven, We have got a couple, I had a few lined up myself, but I think we'll go straight to the audience first. The ones that paid. It's essentially your start. You just finished with that actually. I was I was never able to find it again years ago. I read a report that somebody had linked onboarding that onboarding experience and about how, and to retention whereby, you know, if it's all about [00:44:33] How you set the mindset, if you set the mindset of somebody's view of an organization wrong, they might stay for quite a while, but their mindset is always taught on that experience anyway. Mm-hmm. And we dive diverse. So a question, so let's go to the first one from Charlie. And I, this is probably for you first, Robert. So Charlie's asked do you have any suggestions for being able to spot the use of AI in recruitment with regards to Neurodivergence? So Charlie says we've seen lots of LinkedIn posts where people think they've been rejected for using AI when they haven't, they think it was down to, [00:45:04] Neurodivergency. Any thoughts on that, Robert? Absolutely. And it's a very good question, Charlie. And sadly there was a story during the rounds a few weeks ago that I came across where a lady had applied to a law firm and was told that her application had been rejected, because she'd used AI and she hadn't. And she was then ringing around other law firms to ask them do you use AI detection tools? 'cause if you do, then I'm not gonna apply because I've already been badly burnt by this. And she was neuro divergent. And so the, the challenge around all [00:45:57] Of this was that, okay, the false positive that had come out about use of AI here was causing some real damage and concern amongst students. And the reality is that detection is just not reliable. You get as many too many actually I think the best tools are about 70% effective. 30% are giving you false positives. So it is just, you can't have that many candidates when you've got the volume that many people are dealing with who are being unfairly, rejected irrespective of neurodiversity on this as well. But I think the big challenge is the AI detection tools [00:46:44] Are built on models that are neurotypical and therefore they are looking for things that are unusual and inherently that's gonna be a problem for people who are neurodiverse and therefore being compared to a neurotypical algorithm. So I think detection tools are largely ineffective around this. And being really clear about how candidates can use AI is important because here's my final thought on this one student asked me recently if I copy my application quite often, they get 500 words of why would you like to work at this organization? If I copy my answer into ChatGPT [00:47:32] And I ask it to help, just make it a little bit more sync, succinct, use a little bit better language, and then I copy it back into the application form, is that use of AI, can it be detected and is it illegal or cheating? And the answer is no. It is, it is simply using a tool no different from Grammarly to help you improve your application, but people don't know. And so that is using AI, but it's using AI in a perfectly acceptable way. And that's why clarity on your career site is so important. [00:48:09] I just quickly jump to another question on here, Robert theme, I'll come back to you a minute, Sarah, on neurodiversity in onboarding. But somebody asked the question, have you got any examples of assessment activities that are unable to be hacked by AI? Or is nothing AI proof? Well, there's so much change that it's difficult to actually predict what actually might be on the market two weeks, hence. [00:48:32] So the, there is an element now of, you know, it as I showed in the presentation, AI can read images. So the assumption in the past was, oh, well if it's language based then it's at risk. So we'll just put some images in there and it can't read images. Well, it can read images now and it can create images. And so the simple answer to that question is behavioral complexity. So what AI can do very well is take a set of complex information and give you a single answer. But what it can't do is take a series of images [00:49:14] Or complex informations and give you a series of behavioral responses that you need to do, especially if you each response is to react to something that is changing as a result of you intervening. So what we are seeing, and we're seeing this trend now towards tasks, and I think job simulations tasks where there is behavioral complexity in the assessment are be, will be the only way that you can be sure that AI is not manipulating it. Got it. Sarah, can I just come back to the neurodivergence and the onboarding piece? Is it changing, it is the greater understanding that [00:49:54] That employees, recruiters have of neurodivergence changing the way that they're approaching on onboarding? Yeah, absolutely. And to be honest with you, I think it's helping everyone in the onboarding process because what a lot of our clients are doing, and a lot of people I speak to are doing, are just making the onboarding process a lot clearer. And actually to Robert's point, you know, just giving people clear instructions about what's expected and what's not expected not just in the application process, but during onboarding too, isn't just gonna be helpful to people. Neurodivergent people. It's also gonna be useful to everyone [00:50:33] Because everyone is different and everyone needs to have really clear instructions. So, you know, we work with an organization called Autism Unlimited, and they although they're one of our clients, they've also helped us, to really understand the process someone with autism would go through during an onboarding and what they want, what clear instructions they expect, you know, take away anything that might be misunderstood or misinterpreted and give people clear I suppose timelines and workflows of what they need to do and when, and what they need to expect, not just online, but what they, what they're going [00:51:11] To expect when they actually join on day one who they're going to be meeting. Give them an opportunity to ask questions and connect with people before they start to ask those difficult questions and make them feel, you know, more comfortable. So yeah, in that's probably a long answer to a short question. Yes, PE employers are much more aware but there is still a lot more to be done and I think it does, you don't have to go the whole hog and do everything. It just simple changes to your process [00:51:41] And giving people instructions solves a lot of those problems straight away. Got it. Yeah. And it is again, and it's also great, see from our perspective, I agree this to lots more to be done, but actually again, you go back five years ago, you know, your divergence really just wasn't on the agenda for most employers. An interesting question in, I thought this actually an anonymous attendee asked about probationary periods and actually will the increased use of AI, meaning employers make more use of probationary periods? I've always been a bit reluctant around probationary periods [00:52:17] Because I think if you are gonna hire somebody, you're hiring them and hiring behind. But I dunno, do you have a thought on that? Do you think that is possible? It's, I can see why would happen. It's just delaying ev all these decisions are being delayed and delayed, aren't they? It's like the candidates are not making decisions until just before they join and now employers aren't making decisions until after someone's talking. It's, it's almost things are happening later and later. What do you think, Robert? So I do think that's gonna be a problem. [00:52:47] But it's, it's of course the outcome of a process that has become flawed. And so the way to deal with the problem of people not being as capable as you thought they were is not really solved by increasing probationary periods. It's, it's getting the recruitment process right at the, at the beginning. [00:53:15] If anything, I think I'll make it worse. 'cause if you end up losing more people, well a, your sunk costs increase 'cause you've got them into business. But also the experience for both the candidate, the disruption in the organization, it's right that, yeah, that's, that doesn't, but yeah, but it might be The sooner you, yeah, the sooner you get that sorted, the better. So whether that's in the recruitment phase or the onboarding phase, that's when you need to, you know, to get it right and then you shouldn't be having those challenges you know, ongoing after they've started on that. [00:53:46] So does this mean, do you think that one of the other, of course the testing selection onboarding will all evolve but do you both think that actually we might end up with more face-to-face back into the process? We're actually, you know, doing things live. I mean, every time we do a session on this, I'm always gob maed by the example. So I was, I had not seen that example before Robert, about the person who actually stood over here while, you know, in my mind, avatar as look, just look fate. You couldn't have that looking like a real person. [00:54:14] So, but I just wonder if actually, you know, we might end up going back to more and model where recruitments done on campus and you'll do much more face-to-face. I dunno, any thoughts on that from either of you? [00:54:25] Well, I think that there is a move to more face-to-face. You've just got a commercial restriction on that. You know, we've, we just talked earlier about we've got budgets under you know, constraint now. So you to suddenly say, oh, rather than putting people through a video interview, we're gonna invite them into assessment centers and treble the costs of a, of an assessment center. You've got whole social mobility thing about giving people expenses to be able to come in person as well. So I think the in-person element is going [00:55:07] To have more focus and become more important, but equally because of the commercials constraints, how we filter that top of the funnel down to the people that we want to bring in is also going to be really important to get that right to, And I would say from an onboarding piece as well, it's that mix. You need to keep it a human onboarding experience. You still want to be talking to real humans that could still be on video, but using technology, I suppose, to bring people together. So you might still have a great onboarding, you know, [00:55:40] Welcome day where everyone comes in and meets each other, but then you go off and you're given your groups to go and chat in and you are given the content that you need to go and have a look at. So it's that, it's a, it's a mix. I think you are never gonna escape the face-to-face and you would never want to, because, you know, we all agree that meeting face-to-face, it adds that extra element, but I think it's how you use technology to implement that and help it happen. It's gonna be what develops. [00:56:08] Sure. One last question on the, on the Q&A and this is about anonymized applications. So if I read the questions right, it's about actually, how can AI policies match you know, the drive to have anonymized applications? Are there challenges around from candidates that might not want to show their appearance in a video interview because they're thinking about discrimination bias? Any thoughts on the AI perspective on that anonymization piece from either of you? [00:56:42] Well, I mean, I think it's, it's, it's an interesting one. I don't think AI will necessarily, enhance that from the candidate point of view. I think it's part of an AI policy, I think as a policy is very important and I'm a strong advocate as an adjustment that if, or an accommodation that if somebody doesn't want to have their, you know, video on when they're answering an asynchronous video interview that's perfectly reasonable to ask for you then have the challenge of, okay, well had we give them guidance not to read stuff or not to read stuff that isn't somehow [00:57:24] Bringing out their personality. So guidance I think, becomes really important there. Where I am seeing a trend and I worry a bit about it, is people introducing AI to create a word match. So I've got a job description and I need people with this experience and skill and you know, it may even be a competency and I can see it being used in early careers too, and therefore I'm going to look for, in an application form, has somebody use that word that then, and so this idea that it's anonymous and it's just word matching is better, actually, [00:58:06] It has different elements of bias. It's, it's really easy to game that sort of word matching approach too. So the, like all these things as you start unraveling it and thinking, oh, could this be a benefit? It could be, but it depends on how it's being implemented and very much, you know, the checking for bias and this whole thing about transparency and auditability. Yeah, I think it's less relevant for onboarding. But I think giving guidance goes back to that thing. If you've give, if you've got your set rules [00:58:40] And guidance as to how to do things, then people can't go wrong. You know? And it's fair because, you know, there's element of fairness. I think that's really important as well. [00:58:51] Yes, it is. No, I think it's probably a good point to end on 'cause you are at the at the hour, you know, that's what we're all trying to do, isn't it? To be fair and, you know, match the right person to the right. Job. And it's finding ways to do that. Yes, there are efficiency constraints, but also, you know, we are dealing with people human beings. There was an interesting article published by the ftt, which I thought was quite interesting, which was about students used AI and it said that actually maybe actually what [00:59:19] Academics might do in the future is say there's a piece of AI produced content, make it better. And I thought that's quite an interesting way of, way of using I quite like that. Yeah. Yeah, I like that too. And I think you've, you've just gotta accept that AI is used and therefore how do you, how do you assess a student when they have access? Like we do, we've adapted mass exams now to the fact that people might have a calculator. Yes, no, exactly. And you know, onboarding wise, you can, you can actually give AI training as part of your onboarding [00:59:49] To get people ready to use AI in their role because that's what they're gonna need to do, aren't they? Definitely, yeah. Fantastic. Sarah, Robert, really thanks for pulling this webinar together for us. Been really great content. Lots of comments in the chat, people saying they really appreciate the content. We have recorded this. If you're registered for the webinar, you will get a link, I think it comes through this afternoon or maybe tomorrow with the webinar and details. And also, you know, I'm sure Roberts and Sarah will be more than happy to answer any questions you have of them, if you would, [01:00:19] If you would like to get in touch. I'm sure this is a subject we'll be coming back to time and again over the coming, well, I'd hate to predict how far, you know, we'll be trying to work out how to use all this stuff. So thank you very much for me and as good Robert. So thank you very much and thanks to the audience for lots of really interesting questions coming through. Thanks everyone. Thanks Everybody. Thank you. Goodbye. Have a good afternoon.