Brainfood Live On Air - Ep413 - Sourcing: Dead or Alive? Reviewing SOSU EU 2026

0:05Isolated still as a result of where they sit. Obviously tons of issues historically, like recent history, tough history. But yeah, capable people with lots of interesting challenges. So very, very educational to be there for sure. So yeah, Sarajevo next year maybe. So we should do it and maybe go to Prague as well. like you know just travel there warsaw it looks like interesting area no reason why not um i'm sad i wasn't i'm not able to go to bruno this for this time because that's happening soon um but again i'm going to be traveling back to apac sort of at this time so uh so yeah tons of great events happening in central and eastern europe uh which is just very promising to see anyway welcome everybody to brain food live on air bringing it to you every friday no fail it's episode 413, in case you're counting.

0:57And today, it's a very interesting conversation for us, because historically, sourcing, the topic of sourcing has been one of the most popular sort of topics, both in the newsletter and also on live conversations like this. But of course, hasn't it transformed for us over the last two years or more? And we've seen ebbs and flows, We've had all kinds of different perspectives as to where sourcing is going to be. And I thought today would be just a fantastic opportunity for us to basically do two things. Number one, review the brilliant Sosu EU 2026 event last week, which was in Amsterdam, where we had some of the best sort of thinkers in the discipline on stage.

1:48I want to hear all about it because I wasn't able to come. but also then to open it up in a wider sort of basis to say, okay, where are we with sourcing? Because I've heard some stories that is dead already. So, okay, what does that mean exactly? Is the technique or the requirement to go and find people, this highly technical way of working, is that still something that we need human beings to do? Anyway, that's the topic of the day. I hope if you signed up for the show, That's what we're going to be talking about. Anyway, let's just say hello to Ala Pavlova, of course. Ala, great to see you as ever.

2:29You were at SOSU, actually, but before we get in there, why don't we just quickly intro yourself. Who are you? What is it you do, Ala? Yeah, I was one of those people who was enjoying sourcing and hacking and scraping and all of the kind of things. So my name is Ala. I'm based in the Netherlands. I'm training TA teams with different things related to their day-to-day challenges like a little bit of web coding AI etc. So I'm learning myself I have a rule booked every week 30 minutes for self-learning. If I can do more I will do more but this is something I highly recommend anyone listening hey dedicate some time for learning because that's I think became even more important than anything else yeah 100 % invest in yourself, we're always busy I'm guilty of this as much as anybody.

3:17Always stuff to do, never enough time to do it. But you know what? The longer you leave the self-development out, I think the harder it's going to be. It's very much you've got to just get into it now because it's going to run further away from you if you don't. So brilliant advice from you, Ala. And of course, you're co-hosting with me today. We're going to review various things. We're going to talk, but we've got so many people to bring on. I just wonder whether we should just get on with the show for us. But before we do that, we always have to thank our sponsors for supporting Brain Food Live.

3:53After all, this cannot happen. These shows cannot happen every week without some company out there saying, you know what, we want to continue to make sure these conversations do continue. And one of the most consistent sponsors that Brain Food's ever had and still has is, of course, MetaView, one of our favorite sort of companies out there. They've been fantastic in defining their initial product interview intelligence. Now, so much more. And one of the most impressive things is how fast they're iterating with new products out there. So metaview.ai, in case you're interested in all of that, do check it out.

4:30But we also have Elise here. So I'm gonna bring Elise on. She can come and talk to us a little bit about things. Let's fiddle with some, there she is. Hello, Elise. How are you? Good. Hey, Hung. Hey, Ella. How are you doing? Very good, thank you. Great to see you, as usual. Thanks for having me on. Always a pleasure. Always great to have you here, Elise. For the folks who don't know you, why don't you quickly introduce yourself? Who are you? What it is you do? Yeah. So my name's Elise. I lead training and education here at MediView. I'm an ex-recruiter myself, so definitely very familiar with the pains of sourcing.

5:10um so i love this topic very much like a catch topic there hung uh is it dead or alive i'm interested to see what the folks think do you know what that is and what we're going to hit sort of the uh we want to be honest but we also don't want to be overly sort of ideological on this like what is the uh what is the scenario at one point we know sourcing was absolute peak um you know this is when there was multiple sourcing events there was everyone going to these things. Sourcing training was the thing that everyone was doing. But then we had multiple hits. It was like, okay, decline of requirements.

5:44We had COVID. We had massive inflation. We had companies cutting back. Maybe you don't need to find so many people. Then we had AI emerging. And then suddenly we have tooling like MetaView, for instance, suddenly look as democratizing what used to be a a very sort of highly expert sort of skill set. So it's going to be a great conversation, but we might as well go there, Elise, because one of the really interesting innovations we've seen out of MetaView, of course, is the production of more sourcing support. So I wonder whether you could tell us a little bit more about what's happening at that side and what the vision is from a product point of view and how it can help companies, generally speaking, who are looking for candidates.

6:28Yeah, yeah, for sure. I guess like I'll start with my two cents if I can throw it in. I don't think, I think it's an interesting world that we're moving into with sourcing. I think there's a really strong need still on the sourcing front. I think the conversation is then now with the rise of tools and as technology gets smarter and as AI gets smarter, it's more how does our sourcing workflow change alongside these tools that are becoming available. So I think in my two cents, the need for sourcing is still there, but it's actually just the way that we source is rapidly changing. And that's what we're working on here at MetaView.

7:08I guess for folks that maybe aren't as familiar with our newest product launch, we have been working on a product called Fillmore, which is now generally available to anyone who's excited about it and wants to access. So Fillmore is the very first fully autonomous AI sourcer. And the way that it works is you as the recruiter, the person leading this search, gives Fillmore your requirements. You say, hey, Fillmore, I'm ready to kick off this role. Here's everything that you need to know. And then Fillmore takes that, digests it, goes away. And your next interaction with Fillmore is ready to go candidates who are ready to speak with him.

7:55So essentially that means Fillmore takes that whole top of funnel process for you. So with your brief, Fillmore will go and do a ton of research on each candidate it's looking at. It's doing really deep research. There's a number of agents that go in and dig into that. And then that means that the way that you can engage these candidates is super personalized. candidates are getting hyper-personalized messaging because you have a swarm of agents that have a heap of hours and resources to do this research that just humans physically cannot do. And then so from the engagement perspective, it's making candidates more engaged because they're getting these hyper-personalized outreach and comms.

8:37You're getting more intentional candidates as well because Fillmore is doing all of this research for you to make sure that the candidates that it's tapping into is relevant to you. But then Fillmore's also bringing those candidates back. So it's saying, hey, Elise, I've found Ella. Here's why I think she's great. But before I actually book her into your calendar, you tell me whether you like her, yes or no. And then so I'm still the recruiter. I still get to say, yeah, Ella looks great. Or actually, you know, we're going down a different path. So it is a completely new way of sourcing. But I think at the top level, it is still sourcing and there is still power within sourcing.

9:10And the need for sourcing we see is it's still there to find top talent. I guess that's kind of how Fillmore works. The biggest thing to call out as well, and folks might be wondering, is it is almost like a co-worker. So it's not a sourcing platform that you jump in and work this interface or jump on a web app. It lives within Slack. And so that's kind of the key difference. Of course, at Medivy, we have a sourcing agent. If you want to do more filter-based sourcing or agentic sourcing, kind of like open source, like ChatGPT or Claude, that's available as well. But then you have the option if you want that fully autonomous sourcing flow, you then have the option for Fillmore, which then lives in Slack as a co-worker.

9:53Is it only living in Slack or is it in Teams as well? You know, sourcing, we're cool cats, everyone's on Slack, but you know what, vast majority of people, I think 90 % are still on Teams. So where does Fillmore fit into that duopoly, Elise? That is the hot question. It is, at the moment, it is just Slack. Teams is definitely on the radar and will come at some point. We don't have a timeline there. But yeah, at the moment, it is just for Slack. It lives in Slack. Yeah, really interesting. Folks, is that not really interesting? I mean, I think we're moving past this era of having to log in to different websites.

10:34How many dashboards are out there? It's one of the big reutilization issues of new technologies. They've got to learn a new website. You forget your password, all this type of stuff. It just collapses usage. But here we have AI agents sitting in workflow and doing your recruitment for you. Super interesting. Sorry, Elise, go ahead. I was just going to add to that. I guess the way that we see it is now it is really you have these agents where I'm meeting you where your work already gets done. So exactly as you've just said, it's not a new workflow. And we've seen this a lot, like for folks who work on like technical recruiting or are very close to the engineering side, that we see this world already alive for engineers.

11:16You know, most software engineers are no longer writing code. They're working with agents that live in Slack, that live in their workflows, and they're orchestrating the agents. And that's kind of the world that we're building here with Fillmore as well, where the recruiter is kind of that orchestrator and still in that driver's seat whilst you've got that agent support from the sourcing. I guess like some numbers as well, I guess like some really high numbers to kind of throw out just to put into perspective like how impactful this is. In the last 30 days, so we're using Fillmore across all of our roles here at MetaView.

11:53In the last quarter, Fillmore hired a third of all of our sales hires within our team. So Fillmore now is officially the recruiter with the most placements second only to our founding recruiter, Shiv, which is super cool when most of that kind of came in the last quarter of hires. So that was like eight hires that Fillmore did for us just in one quarter. And then when you put into the amount of work that the agent's doing, in the last 30 days, Fillmore has done 86 ,000 hours of recruiting work and the way that it can do that is because we have this swarm of agents you've got all your 21 25 different agents that are going off and doing this deep research work um it's just like pretty pretty significant the amount of impact that it can have is there any way we can take a look at this at least uh without disclosing any you know confidential information or breaching any kind of basically i don't want to see candidate data on screen but is there a sandbox thing that you could show us at all?

12:58There is. What I would say, if anyone is interested in seeing film more live in action, I would definitely suggest book a demo with our team and we can take you through more of like a tailored workflow. What I can show you though is I want to show you like some of the case studies that we've seen so far to show you like some of the data on how it works. Right, let me make sure this works hard because we usually have some issues here. There we go. I think some of my teams in the chat that's going to share these, but these are some of like our our case studies that we're seeing from customers on roles that Fillmore is working.

13:30Let's click in. This is actually kind of one of my favorites, which is just a software engineer hire. Any technical recruiter knows that software engineering hiring is hard. Here, let me just call out top lines. We hired one software engineer within 13 days. So that is 13 days time to fill between Fillmore sourcing this candidate and offer being accepted. And the way that you can see that like broken down, like the speed of that is it took one email and then the candidate replied within 27 minutes. Fillmore, as soon as the candidate replies, you don't have that human delay anymore. Fillmore just books in the call straight away once you've got that approval from the team.

14:10And then again, just like expediting this process of really being able to move people across. There is so many other case studies. I won't walk through all of them, but I would suggest if you're interested, jump on here and just take a look. We've got a senior manager, again, a more senior role for demand gen in marketing. We all know senior and exec hiring is a little bit more tenuous and takes longer. Here we have a 30-day time to fill. So just some really, really impressive stuff. I don't have an actual kind of sandbox that I can show you live here. I think we've walked through it maybe in another session we've done and hung.

14:47But I would say to anyone watching who is curious and is like, show me, I think the team will pop in the demo link. And I would definitely say, just book in a call with our team and we can show you exactly what it looks like in how it would fit into your process. Yeah, that would be fantastic. Go ahead and do that, folks. Just check it out because this is the sort of thing which causes conversations like the one we're about to have to be set up because it is something that I think is going to just really shift the amount of focus that sources and recruiters previously were doing. I'm reminded of our good friend, Glenn Cathy, one of the legends of sourcing.

15:27And he basically, he might've even been one of the people that helped define the discipline back in the day. But he talked about sourcing fundamentally, what was his quote? I'll never forget it. It's basically the human beings interrogating online systems for human capital information. So human beings interrogating an online system for information that leads to a candidate. Now, if you think about fundamentally what that is all about, that is just a lot of human hours typing in retrieval scripts or retrieval sort of instructions to an inefficient system that's going to try and match it to a bunch of candidate records or candidate information and then pull it back to you via some sort of interface.

16:12I think it's very obvious that AI is going to be able to replace all of that, in which case, what is sourcing going forward? That's the question we're going to be talking about in the next couple of minutes. Elise, thank you so much for joining us and giving us a demo on that. I'm sure people will be very interested in it. Stay with us as you're coming into, I'd love to bring you back to have a chat once we've had some of the guests on to get their feel on it. Because I think you're going to get a lot of people super impressed and super interested. But then maybe the same people will also have a sense of, oh my goodness, what does this mean for me going forward?

16:48And how do we handle this type of existential crisis, which artificial intelligence is having on a lot of us here so at least brilliant to have you on the show let's keep you there um okay ala pavlova um let's get you yeah you just removed this but i wanted to ask her a question but maybe i can do it uh later um you know it's it's if if you ask me uh what's going on with sourcing um like seeing the engines progressing uh giving you opportunities to really uh maybe reduce the amount of time spent on just quite simple work, right? Like just trying, as you said, like retrieve information. It's very hard to sell, by the way.

17:31So sourcing is very exciting. So you do use a lot of creative ways to find people. And probably the most creative part is to outreach, to reach out to the candidates. But the question I actually wanted just to maybe save for later for Alice is for the sourcing data. I will be very curious to see the efforts that the agent is doing, because in sourcing, the KPIs were always something that I was struggling with. Like you have to source for 200 profiles to close one role, like or per day you have to look into like 40 profiles, select like 10 and reach out. So I'm very curious if the agent can also collect a little bit of them, like sourcing data.

18:16That could be very helpful. um yeah but i do think that sourcing is um yeah has been automated in a way how it was done before uh it's um yeah it's not gonna be the same and i don't see that many jobs um uh popping up for pure sourcing a person for for a sourcer um and i saw erin comments as well so that some people who really enjoyed sourcing and didn't want to move to recruitment or something else So they got into recops or talent intelligence. So they're finding also a little bit more like a technical type of jobs so that they can use their sorcerer as a mindset. That's what I can tell. Being a sorcerer in mind, it's very, very helpful to do any kind of job.

19:07So you're basically very resourceful. You never give up. You have a very niche, hard role to close. you're going to still operate with agents the way as your sourcing mind works and you can get the best profiles out of it so i don't know if it makes sense it totally makes sense um but i think it is a case of look that the skill set of sourcing of sourcers i think is is going to be very persistent um we do need more technical people in ta as a as a as a as a as a as a as a going forward, but it may well be those technical skills are going to be applied in a different way compared to where we previously were candidate hunters, you know, generally just finding candidates.

19:52Guess what? Maybe there's still space for that, I guess, in a way, because AI is going to just go through, I wouldn't say the standard stuff, but you can imagine where there isn't a certain sort of roles where the information is going to be very light. You can imagine things like executive positions where information is deliberately obscured. You're trying to find an email address of a CEO. That's not going to really happen. That's been blasted off the internet. So it may well be sourcing as it was. We'll still persist in those spaces. But in general, operationally stuff, hard to say whether it will stay that way.

20:32Okay, let's bring on some guests that were on so soon and get their view on it. So soon, by the way, in case you were not aware, is the preeminent sourcing summit. It probably is one of the ones that really put the discipline on the map. It was happening last week. I want to review that first with the people who were there and then talk about these wider topics. Folks who are also on the show, we're going to do round robin, off and on, because obviously I don't want to stop stuff on the screen with too many people at one time. So please bear with us if you're on the calendar event and not getting invited.

21:08Let's do some intros here. We've got Guillaume Alexandre. Guillaume, why don't you quickly introduce yourself, sir? Who are you? What it is you do? Hi, I'm Guillaume. I'm French. I live in Switzerland, you know, Swiss watch valley, so surrounded by watchmakers and uh i do sourcing uh i do training uh on linkedin recruiter because yes there's a big platform there uh and i do uh records more and more so i i'm the technical right hand man to global head of talent acquisitions pretty much so that's fantastic stuff and by the way no need to say you're french guillem very obvious that's the case that's it brilliant stuff we have Wim here as well Wim great to see you man can you quickly introduce yourself what it is you do I'm Wim I've been recruiting for like 15 years now I do sourcing and I do some training although a lot less good than Guillaume's training

22:11did he ask you to say that? he did not tell me to say it So yeah, wonderful to see you, man. And always great to get your contribution here. We have Chloe here as well. Chloe, can you quickly introduce yourself? Who are you? What it is you do? Hi, everybody. I'm French too, but I live in Ireland and I am your token rec-ups person here. I do a lot of rec-ups and security. Ask me about Candidate Road. Great stuff. Okay. Welcome, everybody. I've just shared everyone's LinkedIn on the chat stream there. on Crowdcast. Please do follow all the panelists. They're fantastic. Let's review SoSue. So how was it, folks?

22:54What were the talks about? Was there a general theme around the day? What's your thoughts? Wim, let's go with you first. I had four great themes. First of all, I missed a bunch of talks because I had a lot of sidebars with a lot of cool people. But there was a lot of AI and implementation. There was quite a bit of doom and gloom in future of sourcing. And there was a lot of people talking about sourcing on social platforms and things that cannot be automated. And where do we need to look there as well? I think that were the three. And then like a fifth thing, which was security. And like, where is this whole thing when it comes to laws and GDPR and so on?

23:38Yeah, really interesting. Let's pick out some of those points there, Wim. You mentioned sort of there was a sense of doom and gloom. So a little bit of, you know, this existential fear, which, by the way, is not unique to sourcing. If you speak to any professional outside of even recruiting, they'll say the same salesperson, software engineer, project manager. Everyone's thinking, hang on, you know, I used to do this. I used to spend 80 percent of my time doing this. And now Claude is doing it or some other product is doing it. What am I doing at that time? So it's very, very common. No one feels completely comfortable with the current situation.

24:17But tell us a little bit about the sentiment. I mean, the key question, is sourcing dead or alive? How did you read the room, Wim? Did more people think it was alive or dead? What are your thoughts? I think, I personally think it's still alive. But I know a few people were talking about pivoting, looking at what else we can do and so on. But I think, yeah, it's a normal thing for the times. And everybody's looking at AI, looking at what they do and what comes out of it. And everybody's still very much, I think, searching for the right solution and the right way to do the things that they want to do.

24:52Because a lot of the things I saw were somebody has cobbled something together and then everybody was taking notes of like, oh, this is how I'm going to try and do it next time. So I think the jury's still out if it's dead or not. And I think there's still a future for sources, very much. Yeah, that's good to hear. Guillaume, let's go to you on that. What's your take? You gave a talk, I think, didn't you? I mean, I'm not sure whether everyone gave a talk on Panely and all everyone attended, but you gave a talk, and I guess you've got some opinions also as to the big question on the future of sourcing.

25:27What's your take, man? Well, I opened Sourcing Summit this year, so it was an opening keynote, and the first thing I said is, if your job is identifying data on a database, saying yes or no, whether you're going to contact and contact with empathy, a machine is going to replace you, all right? If this is sourcing to you, this will be automated on public data and own data, et cetera. This being said, there is plenty to do. Like you would need to be a fool not to see the elephant in the room. Like your sponsor just said 86 ,000 hours of recruiting were done by a machine. Okay. I've met last year the engineers at MetaView.

26:22By far, and trust me, I have seen engineers at the Big Blue, etc. Super great. Okay. Super smart people. Everything that can be done will be done to automate sourcing. And that is fine. That's what I said. It's fine. You have to, just like DHH, the guy who created Ruby on Rails said, choosing code by hand is just not an economically viable work anymore for developers. For sources, identifying data, saying yes or no, and contacting with empathy. is not or will not in the short term be a proper entire role, I would say. This being said, it's also an exciting time. We've seen a sourcing summit, people going to a depth and airing you should catch up on Keti's work on GitHub that I've never seen before, like super impressive technical work on the sourcing side.

27:24So yes, it still exists.

27:29very interesting talks about what to do with those systems because there is a difference between what you want to do what the machine says it's going to do and what it actually does and what the outcome is going to be so there will still be the need for technical people definitely and I think we are the geeks of the recruitment trust me recops is just something that is scratching the surface talent intelligence converting what people want in terms of data and being able to gather it and get insights out of it is just what we've done forever just now we can do it at scale so no no plenty of exciting stuff but let's be honest if you ask me there will be less recruiters and sources in the future as there is now because we are not developers development is elastic you can have as many development systems as you want recruitment is less elastic but it's fine there is still plenty to do what do you mean by elastic guillaume i mean if you have a team an it team that is able to do 10 projects in a year if tomorrow they can do 600 fine, everybody will find don't worry, every organization will find 590 more projects to be done okay, if you are and even more for internal recruiters able to do not 30 or 40 hires a year but 270 it's less elastic, the company is not going to need to hire that many number of people just that there will be less need for people simple as that yeah yeah i see what you mean so essentially there's almost infinite demand for for more engineering work simply because you could always engineer new things even if it is like building an entire new division right and building a new totally let's go and do that metabue itself might be a good example of that given the fact they started off as like a single point solution now they're a multi-point solution who knows what it ends up to be you have more engineers you can keep doing it.

29:41But what you're saying is recruitment doesn't work that way simply because there's usually a headcount sort of target. Very few companies have ever recruited to infinity. I mean, maybe in the weird boom times where there's so much weird money coming in that, yes, you could just hire without limit. I remember maybe early in the 2010s, there was a few companies seemingly that way where they were on that unicorn trajectory and it was like hire as many engineers as you possibly can um but that was a very short sort of period of time that's ancient history now and you know once you've hit or approaching the headcount limit maybe you don't need to hire so many anymore okay and you'll always have bottlenecks you can identify let's say with a magic machine a million people contact a million people but you still somebody still has to go in an interview and shake some hands and there will be a bottleneck of the number of hiring managers able to interview, et cetera.

30:42So there will always be bottlenecks in recruitment. Yeah. It's all fundamentally may not scale if a human being is still interfacing. Ali, you want to say something? Go ahead. No, I mean, with mentioning hiring managers as a bottleneck, I'm not sure what you see in teams currently. I see more hiring managers involved and more curious to try also like agents and saying, hey, can you explain me how can I do this automation that I can see like, you know, CVs faster. I saw hiring managers having a lot of, you know, fun using AI to get some information, to put like nice dashboards and to have to make some sense out of it.

31:29So I do see a little bit different engagement and involvement and excitement from the hiring teams to be involved and becoming even more pretended to be recruiters themselves and sourcers. Well, that mission creep is one of the things that AI is doing. It's basically breaking the barriers down between different functions. We're seeing it in software development where product managers are now doing a lot of prototyping. They're starting to do a lot of building. We even see sort of managers doing the building and saying, hey, I've vibe coded this whole thing. You're going to use it. You know, so that's happening.

32:05Engineers are looking at this and thinking, hang on, they're creeping on our turf. But actually on sort of in the context of recruiting, something like Fillmore, for instance, why is the hiring manager not using that? And do you should not not that shouldn't the hiring manager be the optimal user, actually? Because they're the person that is doing the hiring. They're the person that actually can do the calibration. No need to give it to a recruiter to then translate the meaning or the motivation of the hiring manager. That person is right there. So maybe it's an expansion of who does the recruitment.

32:40More people do it. Maybe less specialized recruiters required. Okay, Chloe, I'm going to you on this. What was your vibe on the event? You gave a talk also, I think focusing on a different area maybe of work that recruiters need to get involved in. But I guess to answer the two questions, what was your general sort of sense check on the room with regards to the future of sourcing? And where do you see sort of the pathway for people who are currently enrolled? um that that's loaded questions now that two people have talked already um like my like i really enjoyed sozu that was my first time and i suppose we talked a lot about ai and what it involves for people is uh sourcing alive or dead i think it's alive is it alive and well probably not um what i think is that people who are very good at that job will be lifted by all the new tools, AI, automations, and whatnot, people who are mediocre will disappear and the role will disappear with them on that level.

33:46But they can bring other stuff. And what you were just talking about with hiring managers, maybe using the tools, I think you should call Diane on that later on because our talk would be very relevant to what you just said. And I forgot the second question. No worries. I mean, people are saying, hey, I've got to get Cathy on, I'm going to try and get everyone on, folks. So this is going to be one of those ensemble events. I think they've increased the limit of the amount of people that actually share the screen. So there will be a massive crowd scene at the end of this. But, yeah, really interesting.

34:19I think we get a general sense check, don't we, folks, as to where we're at with the sourcing world. I want to draw everyone's attention, by the way, who's watching this on Crowdcast. There is a poll on the right-hand side of your screen. Just click on that. I just wanted to basically get your take on this. Question is very simple. Sourcing, more or less important for you? More, click yes on that. Less, about the same. Don't know. Other, please comment. Let's get your views on that. Okay, very, very good. Right, excellent stuff. Let's talk about sort of the, there's a very interesting question coming in on LinkedIn.

35:02Who is that? It was some guy, I forget his name, Dominic, was it? Sorry, I clicked off the LinkedIn from there. But he talked about actually Fillmore. And he was talking about, hey, can it spot defakes? Can it help you with candidate fraud, et cetera? I can't answer that question. At least maybe you could. But I just wonder how that sort of element comes into the role of sourcing. So previously, it probably wasn't part of the job. It was finding the candidates, doing some prequel. Let's get them in the pipe. Let's do all this. the vetting happened later, right? It happened at the back end of the funnel.

35:38Tech assessment was conducted by the software engineers and then admin or someone else would be doing background checking. Now background checking, all that assessment stuff maybe coming to the top. Is that more work for us in some way? Or does that also go to agents and are agents better than humans? Any thoughts on that? Anybody can answer this. I have thoughts. because I do a lot of that. I do a lot of that. I work with security on detecting what we call fake candidates or candidate fraud. What was at the end was making sure that the person was passing the reference checks and did not like that they were employed.

36:18These days, we're faced with candidate fraud at the beginning of the funnel and that does bring in an extra layer of work for the recruiters and the sourcers because they may be facing someone who does not exist that might be a fake persona altogether that might be identity theft that might be identity leasing and those are like three different uh things that you may be faced as a recruiter or sourcer and that's an extra layer of work for you if you have a tool that will give you signals you still have to check the signals that's an extra step that even 18 months ago we weren't thinking about yeah so it's an addition work actually um someone's got to do it um we hear a lot about defects we hear a lot about candidate fraud that noisiness um is going to bleed into that pipeline who else is responsible for that i mean it's got to be us right um i mean imagine you can't just subject the hiring manager and have that hiring manager suddenly become an infosec um go ahead it's everybody like ideally you part like me as recups i partner with security i have trained the recruiters and the sources and I have also trained the hiring managers and interviewers because everybody might see different red flags and as just one red flag, it's just like, ah, yeah, sure.

37:36But if you've got three back-to-back and you line them up, then everybody who is in the hiring process chain has their little responsibility in there, but it's extra work for everyone. Yeah, so it's annoying extra work, but it is extra work. Is that something that were cool to do, Vim? What's your thoughts on that? I mean, it doesn't even have to go that far because as Eli and I were having a discussion in the chat, even LinkedIn profiles that aren't fully filled in or people that haven't fully optimized their LinkedIn profile are now being designated to the dead zone of the algorithm. where we used to tell sourcers, don't start on page one, look beyond page one, because page one gets spammed a lot by everybody.

38:20And those perfect optimized by AI profiles, they're going to show up a lot, but people who haven't done that or who just were like, I'm on a job, I'm okay, I'm making enough money, I'm not going to update my LinkedIn. Those will be shown less and less and there'll be more and more hidden within those algorithms where we as sourcers, were also great at looking at a profile and looking at a potential and looking at what people weren't saying but might be doing. Oh, this person is working at this company. I know they're using this technology because I've talked to enough candidates. So there is a hidden...

38:56We're slowly making the pool very murky. And for everybody, there's a lot of noise. And it's very easy to go poison that well, which is a really great article and a deep dive if you want to dive into it, on AI that it's very easy to just... poison information streams and it doesn't have to go that far but even just not optimized profiles are slowly falling behind and i think there's also a part there for sources to look at yeah really good point again i psych our friend glenn kathy he or he was very very clear one of his early sourcing strategies was actually to search for incomplete profiles um because he realized that they would be missed by standard recruiters they would they would be ignored and actually that would be a person that no one else is speaking to.

39:43So why is that not a great candidate? AI agents, maybe they all converge to a mean. I mean, with respect to all of the products out there, eventually, maybe they're all looking for similar signals, I would guess. We've seen convergence on LLMs generally, right? So we can assume that might still be the case on the application side, in which case, who is hiring the people that are marginal? They may actually be the exceptional talents going forward. Okay, Guillaume, have you got a thought on that? Plenty. You published an article in the Brain Food of Pierre-André Fortin lately about how LinkedIn profiles are actually very much empty.

40:25Okay, there is plenty to be done. Like any, if we had to talk about AI and sourcing, et cetera, any AI is going to base itself on the fact that there is data out there, be that fraudulent data, be that real data. The thing is, what I see is a lot of empty and missing data, and especially the LinkedIn. And one of the criteria is LinkedIn doesn't show you a profile. You see a thousand results on LinkedIn Recruiter. So if your search is 20K, you see 5%, first of all. Right. So what's in the 95 percent? And within the 95 percent, I can guarantee you that the people who did not connect on LinkedIn lately are officially considered as dead on LinkedIn.

41:18So they as soon as they reconnect, they will get messages from recruiters. OK, but it's not in LinkedIn's interest to show people that are not active on the platform. but as a recruiter do I care whether my candidate is actually active on a social recruitment platform professional recruitment platform no there's zero means so the how you search if we're talking about LinkedIn the how you search LinkedIn the how it works how the filters actually and what I said in my talk is any search that is based on filters is going to fail because of how the filters have done okay I can prove it um but yeah you have a problem with lacking data with filters with how AI is done.

42:03Okay. And therefore, what results actually come up. So there are ways around it, but it does take time and you have to master your tools at least. Yeah, it's very interesting. I see a future, guys, where maybe AI tools will end up just generating, I wouldn't say the same candidates, but generally generating sort of from a certain pool of candidates, complete profiles, active people, et cetera, they'll optimize those folks because it correctly identifies them as, you know, the highest sort of bang for buck type of results. But then there's going to be a bunch of people that are either ignored or kind of de-optimized that may still actually be great candidates, may even become better as everyone hits the ones that are already overexposed.

42:53So it's not something, you know, a good or bad candidate is not necessarily a characteristic of the candidate. It's a characteristic of the market and how it actually fits in. And there were frictions in the number of people you could contact before. Imagine if tomorrow all the recruiters of Geneva who have exactly the same position as a Java developer have all the same AI tool and can just send an email to all the Java developers of LinkedIn or any database or whatever. or even everybody, okay, they can send a message. Their personal interest is to say, I don't care. I'm going to contact 15K people.

43:32And out of that, I'm going to make a placement. After three months, what happens? Every single Java developer of Geneva got emails for every single position by every single recruiter who got the job. So the relationship between the client and the external recruiter is going to be even more important because you don't want everybody to have access to the requirements. Because can everybody spam everybody? Yes, it will happen. But then that's called spam. And therefore getting the relationship and the people who are going to make the people reply is going to be key. Yeah, it's about engagement. It's shifting that way.

44:10How do you get engagement? You know what? We spent a lot of time, maybe a decade's worth of time talking about how to write better copy. It's not about copy, guys. it's about your relationship. It's about how you're perceived. If you have a strong relationship, you could write an extremely crude and crappy message and still get a response. Okay, folks, I've got to bring some more people on screen. Stay with us, Guillaume, Wim, and Chloe. We're going to bring you back on a big ensemble in a bit, but I want to get on some of the other folks that have been waiting patiently. Marcel, for sure, one of those.

44:47By the way, who else went to Sosu and wants to come on screen and have a chat with us? Do make yourself available. Let me know. I'm also going to bring on Chad Horn, by the way. I don't think Chad did go to Sosu, but I'll bring him on anyway because I'm interested to get his sort of like head of function perspective on what this entire world looks like. Okay, I'm going to try and find who else do we need to bring on here? i will need diana diana is she there dian no let's see if you can find her i'll bring her on i can't see oh i can't i can do it myself okay yeah i think so i think you can do it um okay let's bring marcel on marcel wonderful to see you how are you i'm fine man yeah great shirt bro i always dig your shirts man uh really wicked um marcel for the folks who don't know you Can you quickly introduce yourself?

45:42Who are you? What it is you do? Yeah, Marcel, I'm from the Netherlands. People call me the BFG, the big friendly giant. I think a fan from South Africa came up with that phrase. I've been in recruitment over 20 years, and I help recruiters and recruitment teams to become builders. So RecOps, or I call them AI recruitment engineers. Yeah, fantastic. And Marcel, you were at RecFest, weren't you? And again, I'm going to ask you the same. Sourcing 7. Oh, sorry. Yeah, you're at Sourcing 7. I beg your pardon, guys. So I just wanted to get your, I guess, same take. Like, what was your general vibe on it?

46:21The question is, look, is sourcing at this point of, like, major pivot? Are people kind of thinking, right, what we did before, we've got to do differently? And if so, what generally are they going towards? Yeah. I didn't present this year. I passed because I was doing some other stuff. But last year, my presentation started with sourcing is that. So this is already, I think, last year, sourcing is that. What I mean by sourcing is that manual sourcing is that. So people are still grabbing buddhium strings or whatever. they need to upskill their knowledge and to do something differently because there are a lot more agents and AI tools out there, et cetera, et cetera.

47:16So it's already happening two years already, at least one year. So that's my take on it. No, I totally appreciate it. And you actually have lived that journey a little bit. um about more than a little bit because at one point you were very prominent in in teaching sourcing and and building a business actually and build a sort of teaching people how to do this but latterly you've pivoted your focus and now it is about automation it's about teaching people how to be more basically better user more ai enabled tell us a little bit about that uh that journey when when you when you decided personally uh to make that shift oh that's really a good question I was really doing some automation stuff back in the days when I was starting my own business, but it was more if than that, a start of automations.

48:11And I was in 2022 already testing some AI tools with Jasper and Wider.me. I was really intrigued by those tools. And then the 1st of December, the day before JTT came alive, and I got a message from one of my Dutch colleagues. He said, this JTT, you have to test it. I just tested another tool. Let's test it. So my first instruction, I didn't know it was a prompt, but my first instruction was, write me an email to a Python developer in Swahili. It wrote an email in Swahili. I couldn't read Swahili because I can't read Swahili, but I had to do it translated to Google Translate. It was really, really good.

49:01Of course it was full of, I hope this message finds you well, etc. said, well, okay, I thought this is the next game changer. I already put all my effort in it and got around it. And I think in January 2023, I have already my own first online masterclass about how to use JCPT in recruitment. So that's my journey. And yeah, now I'm teaching people really advanced stuff, but also very basic stuff. And I'm building, I'm stuck myself. I just delivered to a company, bespoke ATS for them. They want to switch off their own ATS. And they called me a couple of weeks ago. I need an ATS, can you build it?

49:56Yeah, I can build it. So I built a complete ATS for, I think, 15 recruiters are working on it right now. yeah that's crazy and we've seen sort of examples of that that'll be a brain food live going going down the pipe like what what can you build with ai in such a way can you replace your tech stack with it um because that's going to be a very interesting conversation to have okay very very good um uh marcel stay with us on here i'm going to bring on chad here for a second there he is uh chad horn um wonderful to see you chad um for the folks who don't know you can you quickly you introduce yourself who you what it is you do yeah hi hung hi folks great to see you all on this

50:37Hung Lee: call um i'm chad i am based in dublin ireland work for service now i have the privilege of leading the global talent sourcing organization for us um which for a company that's 30 000 strong is quite significant so i've been listening with intent on some of the opinions here so i have my own views but um yeah happy to share whatever's good let's share your views um chad because i wanted to bring you on specifically for the for your head of sort of department sort of perspective um i mean people are saying hey listen ai can definitely do a lot of this work i think we can all accept that that's not in contention um but how do you see this sort of affecting you know numbers of jobs, the quality of jobs, the type of things people do in sourcing now in a function like yours?

51:27Hung Lee: Well, firstly, I'm really optimistic still about the future of whatever we end up calling talent sourcing. I don't think the word sourcer in this current guise will stack up in the future for the reasons we've already discussed with AI agents and automation helping us with top of funnel activity. I mean, I speak for myself with our global team of, I'd say, talent experts, actually, in terms of they really know the market, they really know the talent, but also they really have the intelligence. And they are leveraging AI and using their intelligence to influence business decisions. So we've moved our sources further to the business and less close to requisitions.

52:09Hung Lee: So requisition support can be done, absolutely. But we build talent pools and pipelines in advance of the business needing them. And we share that influence and that information with the business. And it often influences location strategy, comfort benefits, profiles, criteria, and all of the above. So I think the human judgment and expertise and sophistication is absolutely required from talent sources in the future. Less about building a top of funnel pipeline for recruiters or the business to consume. And more about fixing the most important talent problems that the business has and using the expertise and the talent insight and the intelligence that you bring live from the market to really influence those decisions.

52:53Hung Lee: And that's what we're seeing in real time at ServiceNow. Less about building the pools and working on requisitions to fill those wrecks. Arguably, the recruiters and the AI agents can do that for us, but we're working further upstream, closer to the business to fix the biggest problems that they have. This is like talent pooling, community building. Remember talent communities back in the day where, you know, Wouldn't recruitment be great if we just had the time to build those relationships with these hard-to-hire people? And then when the time came, we put them on speed dial. And then, of course, we've all been just smothered with this just-in-time requirements, a completely novel requirement.

53:31It needs to be high yesterday. We have to go and try and find a way to do it. And, of course, that consumes all of our efforts. We couldn't do it anymore. So, yeah, maybe recruitment gets better in a way. So it becomes more like talent pooling, talent community, stuff we've never been able to do. Reminds me a little bit of like how a sports club might build a network of people or a network of talent that they then kind of would pipeline without necessarily needing to have a requisition there. You're going to have a conversation with someone and you're not necessarily going to process them. You want to just build a relationship.

54:08You want to get them, you know, to the point where they can be, where they see you as someone who's a friendly face and a known face that can work well. Does that change the sort of skills that you're looking for? What are your thoughts? Anybody can answer that, you know?

54:26Hung Lee: Yeah, I mean, absolutely it does. I mean, you know, the skills are, I've acquired from a talent source, not what they were a few years ago, right? you know you're bog-standard searching the pool using linkedin and other searches to find a talent is is arguably replaced now so what it is is that business front front team um persona uh the ability to solve problems to bit to influence to to tell a story with data and insight and intelligence that you're building from the real market and acting as that real talent strategist i'm really i say working closely with the business to fix those problems um in the most critical impactful and high value areas we can't do all as you rightly said in these conversations our teams are reducing so the teams that will remain have to be the the experts that people got the capabilities and the mindset importantly to move into this new world where we're influencing and we are those real talent advisors and talent strategists um which i think the whole of ta will end up moving to if i'm if i'm honest with you um and that includes recruiters as well yeah really interesting i mean it might be i mean marcel let's bring you on and it's it sounds like recruitment might be going back full circle, you know, like way back when, I mean, I'm going to, you and I are probably of a similar generation, Marcel, started doing the recruitment work a long time ago and, you know, prior to sourcing existing as a technical thing, that the role was really relationship building, it was having a high EQ, it was being able to be, you know, a very somewhat likability, like your ability to be useful and likable to somebody, build a relationship that would be make you effective recruiter.

56:04Does it go all the way back to that side of things Marcel? I think so, I think so, because I'm also previous Google and previous LinkedIn so I was also there before LinkedIn was even actually there. So you have to maintain relationships and to really network with your network and to do a lot of manual stuff. So the manual stuff of course can be automated way. But there's also a talent pooling is really a thing people always forget to do properly. I see a lot of companies who start immediately on LinkedIn searching and forget, for example, to look at their own ATS and they have a ton of information about candidates who they previously already spoke and they can reach out to them.

57:08and they have phone numbers and everything. And they always forget to do that first part. And if you do Telepop, putting really the proper way, I think it will be the next level, I guess. Yeah, I think the quality of the job might get better, you know, in terms of the enjoyability of it. Even though, because as you say, a lot of the stuff that we were doing, the mental effort of doing the sourcing side It was enjoyable in a way because it's an investigative thing. But building those relationships, I think, remains the premium thing that people will pay more money for, essentially. That might be the decisive thing.

57:52I can see a future where a TA person becomes a highly valued asset if they are an expert relationship builder or domain expert within a particular field. you might think, okay, I'm going to hire Ala because you know what? She's brilliant at networking with these gaming developers. I'm not only hiring Ala, I'm hiring basically her network of gaming developers. And that's much more valuable than even her formidable sourcing capability. So yeah, really, really interesting. Okay, I'm going to bring everyone on for an ensemble final round robin here, folks. Let's just do all of that. Elise, if you're still around, I'll bring you back in as well.

58:36Let's go around and just say, you know, where are we going forward with sourcing? Can we give someone who's listening in on this some tips? Like what's the one thing you'd recommend they do? Let's imagine they were in a sourcing role. That means the classic sort of Glen Cathy description of it. They were a finder of candidates. They were interrogating systems. They were doing Boolean searching, doing all of those things. now AIs here, now Filmores here doing 80 % of that. What do they need to do going forward? Chloe, your thoughts first. Keep learning about anything that interests you because that's probably going to be your next gig, whatever happens.

59:18Yeah, you've got to be curious, right? You've got to be one to do it. Let's not be overly wedded to any professional ID, by the way. again I was here before sourcing existed as a professional identity it emerged it will demerge talent acquisition as well emerged at one point I mean it was just a task that HR did at one point before it became its own professional identity so the idea that sort of tasks become jobs and jobs eventually might fall back into being tasks that's just the normal flow of things let's not be overly wedded on that just think about what your core qualities are and what it is you good at doing and see whether you can apply those in, in whatever sort of where the demand is going to be.

1:00:01Okay. Wim your thoughts on, on this. Try new things. And most importantly, try building new things and see what works for you. Cause I noticed there's so many interesting tools. There's so many different solutions. Everybody like every week, there's something new coming out. Somebody figured out something new. So yeah, see what works for you and what works in your workflow. Because it's not because one company says that they have a solution, that it works for everybody. No, absolutely. You've got to not only be a great user, you've got to be a builder as well, right? I mean, you might as well, especially sourcing.

1:00:38Sources have probably got least sort of excuse because they have built their skill set as technical people. So you actually ideally optimize at this moment to dive into it. Guillaume, your thoughts? The same, build shit. Like if you're just a user, you'll be replaced because a vendor can, then you can just get a user. You need to be at the frontier, like a forward deployment engineer, at the frontier between AI and the tools and the business problems, understanding the business, knowing what sourcing is, what recruitment is, what people mean, et cetera, and translating this together with tools.

1:01:19But if you're just a user, you're probably out. build stuff. I'm on 2.4k commits for somebody who's never done developments ever in four months. I'm building. Get on with it. And there is an existential pressure as well. I mean, we don't want to put that overly too much, but we have to understand that there is urgency to the matter. Okay, Marcel, your thoughts? What's your advice? Yeah, what everybody said, but on top of that, I think there will be room for not traditional sourcing, but there will be room for more sourcing because I see the whole application tsunami wave coming towards us of more and more applicants who are using AI or fake applicants are using AI.

1:02:18And so maybe it's not good that you open a rack online, but that you source for the people instead. Yeah, absolutely. Do it more quietly, right? And that should lean into more sourcing capability, sourcing demand, to be honest with you. Okay, Chad Horn, your thoughts on this? I mean, I guess you're probably advising your team about this right now, right? Yeah, yeah. So what are you telling them?

1:02:42Hung Lee: Business curiosity. Now, AI fluency matters, the tools will come and go um understanding the business problem and how to fix it i think it's a doable skill that will last long long long more than some of these ai tools out there yeah brilliant stuff okay i found what you ala having heard all of this stuff like what's your take on it and uh what's your advice i was the one who was doing the closing keynote so um and it was very interesting because what I did, I went on this networking times, talked to people and asked them, like, is it something you're missing? Is it something that you would like maybe to hear at the end of the second day?

1:03:23And it was interesting. Everybody asks how to build skills. So they want to build skills that will automate part of their jobs, like, for example, setting up sources strategy, creating Boolean searches, maybe doing the calibration as well um and what i did at the closing keynote i just shared my screen and i uh went people through like like basic steps like this is where you find it this is how you do it and i think this um this is still very relevant that a lot of people in a lot of companies are in very different sourcing um life relationships because for some companies none the eye is available for some companies they do have um budgets for uh recruit LinkedIn seeds uh some companies um yeah like recruiters never like probably have time for um sourcing and this kind of thing so I think um we're living in a very like um uneven type of um companies and conditions that we have um one thing that's still I think is very important and still alive is the outreach i want to challenge um yeah it's probably not here so met the few teams so let me try to see if i can train your agent because my response rate is still close to 90 percent and this is something that's um is still very exciting part of the job so that you get through the noise and so you you know when when everybody sends the message to the same candidate i want to be the person that I get the reply.

1:04:58And most likely, it's something very special that I put into this message. So the outreach is still the game. So I think we can still compete in a good way. So, yeah, and so the message to everybody still stays the same. We're living in the times when learning is not optional. You know, you just have to put yourself there and try things out because falling behind. So it's really kind of, you know, the feeling that a lot of people have, but I would like to tell you, hey, don't really look up on anyone else. Create your own path. Take one, two per day, take one, two per week, take one, two per month, but take something.

1:05:39So you will be further away. So that's probably was the thing. Yeah. The keynote ending part. No, I appreciate it. Also go to the YouTube Academy. Also very good. free stuff this is the thing there's so much help out there first thing to do if you don't know where to start just ask one of your AI of choice and it's going to give you that so develop that muscle memory and get through it ok brilliant stuff folks very very interesting moment for us it is there's urgency but there's also excitement there's opportunity there's also risk that's life folks what more can you ask for Let's go get it. Okay, thank you, Chloe Morrison.

1:06:24Thank you, Bim Darmans. Thank you, Guillaume Alexandre. Thank you, Chad Horn, Marcel van der Meer, and Ala Pavlova, and everyone else who paid attention to this conversation. It was fantastic to have this chat with you. And great that SOSU was such a great success last week. Let's make sure that community keeps on going, no matter what we end up calling it or what we end up doing. So fantastic stuff, everyone. Thanks for watching. We'll be back next week. We're going to be talking, we're going to start a new series, which is going to be called How AI Has Changed My Job. And we're going to get different, basically, professionals to come in to talk about it.

1:07:02I'm going to get teachers, I'm going to get doctors, I'm going to get lawyers, I'm going to get random people who do a job, and they're going to just tell us, okay, how AI has changed it. Next week is going to be software engineering. Of course, it's one of the areas that has had the biggest transformation. So we're going to speak to software engineers and ask them, okay, take us through it. What has changed the last four years in your career? How is engineering today, software engineering today, different from what it was when you trained for it five years or so ago? So make sure you sign up.

1:07:30Follow the channel on Crowdcast if you're watching this here. If you're following this on LinkedIn or my LinkedIn, go to my page and just follow me there. Basically, I broadcast or live stream from that page every week as well. Okay, that's it. Let me let you go. I will see you next week, folks. Thanks, everyone, for watching. Thank you everyone! It's always fun.