High-volume hiring is intense. Candidate volumes are soaring. Budgets are tightening. And recruiters are under pressure to go faster — without losing quality or fairness.
So how can TA teams move forward?
Jess Morgan, Chief Transformation Officer at Instant Impact (ex-Capita, Baringa, Centrica), believes now is the moment for TA teams to take stock — and rewire their hiring process for a world where both candidates and recruiters are using AI.
With decades of transformation experience and a Lean Six Sigma Black Belt to boot, Jess brings the kind of operational rigour and people-first mindset that high-volume hiring desperately needs.
In this episode of TA Disruptors, she sat down with Arctic Shores’ Robert Newry to lay out a simple, pragmatic plan for managing the chaos.
🧭 Reclaiming control in high-volume hiring: What to do when 2,400 applications land in 24 hours — and why the answer isn’t to restrict applications or hire more recruiters.
🧠 Smarter automation choices: How to separate low-stakes decisions (like eligibility checks) from high-stakes ones (like who to hire) — and why your tooling strategy needs to reflect that.
⚖️ Fair, compliant decision-making: Why the EU AI Act means some hiring decisions legally can’t be made by AI — and what to look for in automation that supports, not replaces, your recruiters.
🔧 Landing change with hiring managers: The practical playbook for rolling out new tools — and the one-pager that can make or break success.
💡 Fixing process before buying tools: Why the real bottleneck isn’t just candidate volume — it’s unclear workflows, legacy metrics, and slow decision cycles.
This episode is full of sharp insights, grounded advice and practical takeaways for any TA leader navigating the new realities of AI-enabled hiring. Whether you’re handling entry-level roles or managing enterprise-scale recruitment, you’ll walk away with clarity — and a plan.
Robert: Welcome to the TA Disruptors podcast. I'm Robert Newry, co-founder and chief explorer at Arctic Shores, the task-based psychometric assessment company that helps organisations uncover potential and see more in people.
Today, we are going to explore the impact of volume applications in the age of AI. Recent survey data shows that the average recruiter is receiving 270 % more applications than they were three years ago. Some of this is due to changes in the job market, but a large part is due to the new AI tools that help candidates apply for jobs while they sleep. Tools like teal.ai and ai-hawk. And this is straining and in some cases breaking recruitment processes and teams.
So, who better to speak to about how to deal with this challenge than Jess Morgan, Chief Transformation Officer at Talent Solutions Specialists, Instant Impact? Jess has an extensive management consulting background from Capita and Beringa to Centrica, and Instant Impact leads the technology transformation practice, helping organisations enhance their recruitment processes and drive operational efficiency. And no doubt your Lean Six Sigma Black Belt training and skills are super valuable. So welcome, Jess.
Jessica: Hello, thank you for having me.
Robert: So Jess, let's start with how big a problem is this surge in applications is for some of the people that you work with?
Jessica: Yeah, I mean, there's no getting away from it. It's definitely a problem. We've seen, you know, I was only talking to a few people in the organisation in this last week, and we've seen, you know, things like one of our energy clients, a sort of low-skilled remote role, get advertised. And in less than 24 hours, we got 2,400 applications.
Robert: For one role?
Jessica: For one role, less than 24 hours. You know, there are other anecdotes I tell you across other clients and, the stats that you said at the beginning, I think, point to it. know, TriPad have also released a study and I think it was November alone last year. They had 120 % increase in applications across their platform at the same time of a backdrop where they had less roles being advertised. So that can only really tell you one thing, which is that there are more people applying and at mass scale. So it's definitely something we have to navigate. It's happening more and more. I think… the way that we navigate it is the important bit and trying not to shy away from the fact that it's going to continue to happen. I don't think we're going to break the cycle of candidates using AI to apply for roles. So I don't think we can avoid it. And I think what we have to do is try and design our systems to support it or screen it out or however we want to design it is the only way we can, because it's not going to, it's not going to stop.
Robert: It isn't. And you're right. And, and this is, think, different from the past. We have been in situations in the past, particularly after the financial crisis in 2008-9 where we saw a surge in applications and we brought in systems to deal with that.
But now this feels really different. This feels that we've actually got to turn the recruitment process on its head in some ways and re-look at what we're doing and how we're doing it. And it seems to me, and therefore it'd be good to get your perspective on this, that there are two big challenges here. We've just got the volume. So how do you deal with that volume and deal with that volume in a way that's not introducing bias?
And also some of the ways that we used to deal with that volume in terms of benchmarks as well. So what we used to do in terms of screening and the different steps that we used to have in the process. But we also have to think about the whole compliance piece a bit around this too and what's legal and what's not legal because AI is changing uh the candidate applications but also how we as recruiters use it.
So how are you starting with this? So we've got this big challenge, we've got this massive increase in volumes, systems are not set up really to handle this. So where do you start with some of your clients on this?
Jessica: Yeah, it's a good question. Think… I think the reality is it is not possible for a human to absorb that pace of volume increase. Okay, so you've kind of got two choices, I think as an organisation, you either kind of heavily restrict the volume applications you can get, but that probably feels a bit counterintuitive because you want to allow a period of time for the applications to come in to allow for fairness, or you uh augment your team massively and you hire lots of people to deal with this problem. again, know we just say budgets not gonna do that. So, you know, I think a lot of organisations and TA functions are then looking at tooling themselves. So AI tooling, for example, we've definitely looked at uh things like CV and application, AI screening tools to support. Oh recruiters in defining kind of the categories and parameters for CVS applications, that yes, I'd like to look at this pile versus this pile is a bit of a maybe. So we don't auto-reject people. That's obviously not compliant, but okay, I know that the top applicants of this role must be able to demonstrate X, Y, and Z. Can you please put those into a category for me to review first to help me with this huge influx of applications?
I think then what we're also seeing is almost introducing another screening step. So I think traditionally it was CV, call, first interview kind of thing, right? Now I think with the volumes, we then are starting to introduce maybe another screening stage, which could be a number of things depending on the appetite of the organisation. could be a one-way video interview where it's a little bit harder to use AI to completely kind of generate your answers for you. And, you know, some of the tooling can spot that for you and it's built in. So you get a much better flavor for really what the person is able and capable to do for the role. Some additional tooling steps like that, maybe some conversational tooling to sort of screen further down the funnel.
Robert: There's a couple of things to unpick in all that too, how the process itself is going to change. But also what are the tools that we think are going to be used? And so let's go back to the CV screening. what have you seen that works for this? are you thinking that actually companies and organisations, TA teams should be wary about when looking at those type of tools?
Jessica: Yeah, so look, we know that there's a whole sea of… not good practice tooling in this area. And that's, you know, I'm not denying that. think the um due diligence we need to do on tooling is important. And, our organisation have gone through a big exercise with legal teams and making sure that our practice is aligned to various compliance kind of um legislation. I think what you've got to look out for and understand is anything that is going to make automated decisions for you.
a tool that is going to make a decision on a candidate and then reject them without any human in the loop is something that I would be looking to avoid. I think at each stage you still need to have a human to go in and go, yes, I agree with what's been done there. I'm going to have a look. No, I don't agree with this. So there's still a human that makes the decision because ultimately, otherwise you risk losing maybe good candidates that did apply because the problem we've got with AI based applications is they can be quite good quality.
So if you do then just leave it to a system to make the decision, they're probably going to pick the AI em completed applications over the maybe the ones that weren't AI completed. Things like typos, things, you all that kind of stuff.
Robert: Isn't this part of the challenge, though, that you have you've got to set criteria up for that? So any system, if it's not making a decision, it's having to make a scoring or ranking in some form or another. So that's set by a human in order to be compliant. Get that. But then comes the bit of, well, how is it doing the scoring? if, for example, which a lot of the CV screening tools do, it's a word match.
What, how, because candidates have got access to tools now and if they know, look, I'm applying for this job, you've got to have these words in your application. Brilliant. AI tool that I've got, go and put those words in. Get me up the rankings.
Jessica: Yeah. So again, I would be looking for a tool that doesn't just do word match because the other challenge with that is if you don't put the right word in, you let's say you're someone who's got lots of experience in finance, but you don't put the word finance on your CV and they're looking for a finance role, those word matching tools are gonna rule you out. So that's just, that's not right. So again, the more sophisticated tools in this area can do a lot more kind of inference and understand like a wider range of understanding and context rather than just going for one keyword.
And so, you know, I've said this on quite a few recent client calls around when you're looking at what you might trial, start small. Have a look at a few products, see what works for you, check it's compliant, maybe give it a go with run roll and just see how it works in a test environment. I think the other challenge you've got is organisations that want to just implement something and go. And I do think that's a high-risk strategy in this sea where we don't really know exactly what our compliance criteria is in the UK yet. And where we've seen it work best is when we've done it on that kind of approach.
Robert: Yes and I think that and we'll come back to that point about compliance and obviously in the UK it's slightly different from the EU, but the EU is a high bar in the EU AI Act and probably a good place to start at the high bar rather than wherever we might end up uh in the UK on that.
And so understanding the challenges, because obviously you were saying earlier, you've looked into this and taken some advice on this, and there clearly is a line that has to be drawn between automation, which is just fulfilling what a human has said in terms of criteria, right to work, or… got to have a finance certification as opposed to decision making, which is clearly where the system is making an inference or it is making a decision to auto reject based on a word match. But it is a bit greyer at that point as to Is an inference still making a decision?
When, what's your take on, you know, if, an application gives you, here's the top 10 CVs out of the 2,400 that have just applied, don't worry about all the others. Here's the top 10 based on an inference and a match to a job description. Is that Is that some sort of decision being made because you've got 10 that are being put forward and 2,390 that are not?
Jessica: I think in that example, it's ranking, and I think it's you input the strength of the requirement into the system, right? So you could say, this is sort of important versus this is almost mandatory. So I think it's then you're still really guiding it. Almost everything you put in there is really human-led. And I don't think that the tool is really doing a lot by itself.
But back to your point about the compliance piece, yeah, we've taken guidance on it because it's a really interesting one. The EU AI Act is quite clear on the two or three areas specifically for high-risk categories like recruitment, right? Where people and decision making and you know, a lot of it is to do with transparency by reducing bias and not allowing a tool to sort of make a full decision on its own. I human must be in the loop at each stage.
And so all the tools that we've chosen basically will allow us to do that and to be involved. I think there are tools out there though that you can almost put applications in and it would put calls in the diary for you and you don't really know how it got there. The problem with that as well, that might help with efficiency, but then when you, especially in the business of RPO, when you then go to talk to the hiring manager who's the client. And also that candidate isn't quite what they look like and you haven't even been part of that process, your credibility is immediately undermined. So I would say we've taken a relatively cautious approach to it. The reason we wanted to get lots of legal advice was to make sure that we weren't breaking any compliance laws, but also to give reassurance to our clients. Because it's candidates who are engaging on their behalf.
But let's see what comes. That's the gold standard at the moment. Who knows what's going to come. I think it's pretty soon that they're going to come out with kind of exactly what we should be doing in our sphere it could be different.
Robert: Yes, and it is rapidly changing from a technological as well as a legal point of view. Although I think the point that you made earlier on this, that there is a starting point that we've got here is just understanding the difference between automation and AI where a decision is being made as opposed to where some information is being from which a human then makes the decision.
Jessica: And some of it, to be honest, isn't even AI. I think we use this big word and actually you look at it you're like, that's actually just automation that arguably we've been doing in different guises for a few years now, but we're just sort of rebranding as AI and then putting in a tool name and going for it.
So I'm, know, there's three or four different phases, right? With what we're trying to do with AI, there's a sort of AI assistance, which is very much your chatbots, you where it's just, might internally have one for HR requests, says, what's the policy on this? And all it's doing is returning information. That's called AI, but I mean, is it? Then there's more AI agents. But again, what we've been talking about, very human-led parameters set by humans. Again, you could probably argue that's largely automation.
Where we get that, I think is the interesting bit, which I think in recruitment, it's very untouched is then the kind of multi-agent workforce, where that really is taking away whole steps of people's roles and making decisions and pushing along work. And that I think is the area that will really impact the workforce. At the moment, think it's freeing up some time, yeah, sure, and people can go and do some other stuff. But when we start to talk about agents actually going and doing big chunks of work. I see there being a real impact on kind of workforce redesign and how do our roles just look different.
Robert: Yes. And let's dive into that a bit. The agents coming because there's a lot of talk about what roles are going to be around in the future for TA and what's not. And part of the role in TA was to manage the process. And you've talked about keeping humans in the loop being important to remain compliant. So where I can see some around the scheduling, but you're obviously thinking and feeling that the agentic AI may actually do that first sift. things like, I assume that they might even do a, because I've seen it and it'd be interesting to get your take on this, AI avatars that will do an interview rather than a human doing the interview set in questions and parameters. And
and then that information I suppose will be sent back to somebody, but is that the kind of thing that you're talking about, that the human doesn't have to be on the phone or doesn't have to look at the video interview?
Jessica: Well, or, you know, there's different layers of it. I think that could ultimately be the longer term view maybe where you do have more avatars doing certain process stats, but actually a lot of what we're talking about in sort of agent AI recruitment is more like this orchestration layer that will almost perform a lot of the tasks for the TA individual. And so they're really just a checkpoint.
So agent goes away and does, I don't know, of sourcing or whatever, comes back, says, you happy with this? Did I do this right? And then you kind of go, oh, no, actually, I wish you had gone to that job board, you should have done this. And it learns, goes away, comes back then presents to your shortlist. Are you happy with it? You know, so really the role of the TA person doesn't go away, but it changes. And then potentially, I don't know, more time later down the line for the final few candidates, who knows, or to work on more HR strategy, who knows?
Like, I don't know where that will go, but I don't know if I think that roles will cease to exist but I do think fundamentally they will change in terms of purpose and the way we deliver recruitment will be different. I was just looking this morning and the Society for Human Resources Management think was saying in 2023, 26 % of organisations were using AI and recruitment, 2024 that doubled to about 56 % and I think this year it's gonna be something like 75%. So it's really, you know, it's coming.
Robert: Although, as you mentioned earlier, I think a lot of people are saying they're using AI when in fact they're just using automation. We have a little bit of a challenge around all of that, but I like your point about the orchestration piece in this. And so it comes back to if the role of TA is to orchestrate on this then actually one of the key roles for TA and to some extent possibly even then for their advisors is how do I get good advice on this orchestration piece so I don't end up drifting into a decision-making area without realising it because the point that you mentioned earlier is that you can think, oh, I've just got this little feature that's appeared on an ATS or somebody has said, oh, it does scheduling, but actually also it can do ranking and scoring.
And so the challenge for TA is, well, how do I interrogate and know? what is going on behind the scenes. Because if you don't know, and in TA we weren't trained to be algorithmic specialists, how do we address that challenge we might be walking into a tiger trap?
Jessica: Exactly. And I think this is where a new type of role is spinning out in TA, I think, which is, know, quite good timing for my sort of industry background know, like transformation type uh AI tech consulting roles where, you know, um organisations are starting to have this feeling that, you know, AI is coming or we want to be part of it or what do we do? As you say, it's not our background. And there's this sort of phrase that keeps going around. So lots of events I've been to recently, which is fear of being obsolete. Fobo.
I'm sure you heard that at but it's real. It's sort of like, you know, we need to get on board with this, but where do we start? And so lots of people kind of either kind of speaking to consultants or internally hiring a role, which is let's look at our processes, which is our tech stack, and let's see where the opportunities are to leverage what we've already got. Because actually a lot of the software, as you say, ATSs, they do have little bits of AI automation built into them, but people just aren't leveraging them. I see that classically.
It's almost like 101 of going in for a diagnostic to a client. And I'm like, well, let's just, before we just wipe out your entire tech stack, let's just have a quick look. See what capabilities you actually have. actually that's quite a cool piece of tech you've got there. Are you using, you know, and actually they're not because they don't understand and they need some advice on how they then build their processes and their change agenda around it.
So it is, you know, it's, it's a bit of a project really to uncover kind of where you're on use AI, how do you implement it? How does it impact the processes around it? Because there's no point really in implementing it, but carrying on with exactly the same process because it's meant to, you know, 10X your output and change the way you do stuff. But you need somebody to help design that I think. If you don't have that someone in your organisation who's naturally minded like that. And then what do do with that time released from the team? What kind of strategic projects could they work on or can we do more volume and quicker and all sorts of things?
Robert: Yes. And there's a lot of talk around how all these AI tools are going to free up all this time. We can come back to a bit later as to then where does the value of TA come in. But let's just go back to the how do you do the transformation?
Because you've got this on the one hand, we're going to start small, we're to have a little try out somewhere now see that we understand AI, we understand actually what it means, where it fits in the process. And then at the same time, we've got to turn the process on its head because it's not fit for purpose now. The process was designed for a very different world that we live in. So how would you advise somebody that is thinking, I can have done a little bit of experimentation with AI. feel I'm comfortable now with the orchestration piece where it's low risk and I can implement it. But now I need to change my recruitment process because if you tinker with lots of little experiments, you'll end up with a Frankenstein's monster. And also unintended consequences.
Jessica: Exactly. I mean, it needs to be carefully structured is the reality of it. We've got, and I've got quite a agile approach to these things because you're gonna, there'll be areas where mistakes are made and you go, actually, maybe we should change that a little bit or, but, and you don't wanna be too regimented on it because as I said, it's such a green fields area. There'll be some tools you might try and go, no, not for us, that didn't work. And that's okay as long as you haven't committed your whole organisation to it straight away.
But I think once you know a couple of tools, potentially all sort of tech strategies that you want to to go with, it needs to be a structured, like, program of work effectively, where you look at your sort of all design, you know, are people doing the right roles to go with these processes? Do you have the right technical expertise to run some of the orchestration? Who's going to implement it? Yes. Because otherwise what happens is it will fall over. The other thing I think is really important to call out that is so easily forgotten, and you know, I've come into this many times doing management consulting because it's a similar thing, it's a change programme.
You know, you're saying fundamentally we're to change the way we do this part of the process or how we interact with candidates to people who for their whole life as TA professionals, as you say, have run it one way. Yeah, we've done it in one way typically, probably quite manually. And actually it's been part of the amazing part of their role is to really interact with every stage of the journey and spend time reviewing everything.
So you can't just assume that that… that lands and is really quickly adopted. So that is a really big part. think at beginning of any of this, if you're planning to do something quite fundamental to your process, how do we get people on board? How do we train them to use the tooling? What do they do with their time? And that is the piece I think is quite often missed. You get really excited about tech and transforming and let's buy it, let's do it. But it will never, you know, I think there's some stuff around 85 % of AI projects will fall over in the first year because it's not run as a bigger change piece.
Robert: And I think that's so important because I mean, you come from a management consulting transformation background. And one of things that always surprises me is, and I suppose it's sort of natural human psychology around all of this, which is, oh, let's get a bit of tech. And I've been sold by the vendor on this that it's going to be transformed. I'm going to get a 10x uh return. Then I'll put it in and I'll expect the rest of the process just to get better and to work as opposed to, on, well, we've understood how the tech can work a bit, but actually now we need to change the process. need to get, I often talk about the golden thread from here's the output, a quality hire, and we can have a whole discussion about quality of hire and what that looks like in the future, because it's gonna be different from the past.
But then you've got to work back to the application piece on that. And if that golden thread as a process does it doesn't matter how much tech you put in and AI you put in there, it's going to be automating a process that isn't fundamentally got a golden thread and working in the first place.
Jessica: Exactly. And in our, in our line of work, right, we've got so many stakeholders to keep on board, the candidate, the hiring manager, the business, you know, all the, these other kinds of stakeholders. And so that's another place I think is really important to start on any of this you know, is this ultimately going to deliver us quality of hire? Is the candidate going to love us throughout the journey? Because we also know our brand, we want to get the best candidates, but also is it going to be easy for hiring manager? Because again, if you create a process that is really hard for the hiring manager it's going to be difficult to make it land. Yes. But yeah, there's a whole skills piece there as well, because if you're expecting the hiring manager to interact with quite a tech-enabled process that potentially they weren't before, you cannot assume that that's going to be easy.
Robert: Absolutely. You make such a good point there because I've seen quite a few cases where TA gets very excited, understandably, about something that's going to really improve their efficiency and the value that they provide to hiring managers and then a hiring manager has not been brought into this process or as a group they've not been brought into it and so they're looking at going, well, how has this made your life easier? It's not necessarily my life easier. And then it starts to fall apart.
Jessica: Yeah. And you've got the trade-off of all the different stakeholders, but then the end game, of course, to be more efficient, to be able to do maybe more with less is also sadly the backdrop of quite a lot of this as well. So, yeah, it's an interesting one and you know we're keeping close to all the lessons learned across the different transformations that we do outside of Instant Impact, management consulting-wise, you touched upon, the successful ones are where you've got probably someone a bit dedicated to seeing the kind of program of work through. So from a people side, from a process side, and a side. Because all three things need to kind of run concurrently and be looked after by somebody who is not just sort of like side of desk, whilst they're also doing X, Y, Z. And that's typically where we see successful deployments.
Robert: Yes, and it would be interesting to get your take on this. A uh job role in TA now, in one definition it's called a TA enablement manager. I've also heard it referred to as a rec ops manager, but fundamentally the requirement is that point person whose job is to go, right, what? What are the actual needs here? What are the processes here? Let's get that all agreed, stakeholders uh aligned all of this, and then we'll go and look at the tech. We might be looking at existing tech and tweaking some of that. Is that the kind of thing that you think is going to be really important? Because you can't change this overnight.
This is going to be over several years. so you're going to need somebody whose job is not being taken with firefighting. I've got another 10 wrecks now. I need them hired next week. Off you go. You actually need somebody who's focused on this transformation.
Jessica: And probably quite a central resource. You can work across, well, for know, for our PO businesses, obviously you've got lots of different client teams, somebody that can be like a relatively impartial view across all the processes and just know best practice and come and try and kind of uh implement it and not get kind of bogged down with client specifics or whatever it might be because the bigger picture here is, you know, what as an organisation do we think that we should leverage and transform?
I think the other bit, you know, not to forget as well is the strategy. I think there has to be alignment at a senior level on what is the strategy around AI and tech. Do we want to be really tech-enabled? Do we want our people to be leveraging technology in every rec that they do? And we're going to get rid of Excel and we're going to say we only use the ATS or whatever it might be. And everybody has to be brought into that because it's quite painful to move from old world to new world. you need some senior impetus to drive buy in and then yes, a role to drive it. And then of course the business has to be bought in part of the change program.
But I think that strategy around AI as well, because I speak to clients sometimes and you know, maybe the TA leads like, yeah, definitely want to try this, this, this and this. We need it. um My team are crying out for it. And then I might speak to somebody else on the leadership team who's really nervous about AI. We're never going to do that. We're breaking all sorts of compliance laws, never going to happen. And so I think setting your position as an organisation on this is what we believe we are comfortable to do in terms of AI and our principles and our principles. And everyone's aligned behind that and the same with transformation em is a good place to start.
Robert: Yes. And you make a great point about, you know, the stakeholders is now includes a data privacy officer in a way it didn't before. I we've obviously got GDPR, but under the EU AI Act, similar kind of fines if you're found in breach of that to GDPR. And so a key stakeholder now is your data privacy officer. So you can have a strategy to go and adopt AI everywhere, but if they're not on board with how you are working within the confines of whatever legislation or jurisdiction you want to work with and you haven't understood those, then the whole thing, your whole strategy and your whole tech roadmap can be thrown out the window.
Jessica: Yeah. And the interesting thing is almost every company you speak to now does have a DPO. Yes. That didn't used to be the case. It used to be you might go, you might kind of call on a fractional DPO from a different company once every six weeks and just check a few things. It's so fundamental.
I mean, it's like one of the first things that I did when I joined my role was to kind of formalise somebody in my team as the DPO. We had DPO support, don't get me wrong, but as that key person and in my team, because a lot of what we're doing is driving tech transformation AI. We have to know inside out what is okay, what's not okay. Speaking to the lawyers all the time about it, making sure the business is trained on what's okay and what's not okay. So actually, it made sense for it to sit in my world rather than be this kind of outsourced.
support I didn't really know anything about. uh But yeah, mean, the EU AI Act is interesting because it's seen as the gold standard. So it in theory should be what we will aim for, but no one quite knows how it's going to be kind of legislated for in the UK.
Robert: Yes. And so, you we'll have to figure that out as we go along on that one, but there's no harm in preparing.
Jessica: I think you're better off going high, you know, and then being able to row it back if you wanted to.
Robert: Exactly. Exactly. That's certainly the advice that I give. And so that's fascinating about then the process and the transformation and, you know, people have to think about that different point person to make that work. And then, what about the candidate in all of this? Because we're doing all of this to create efficiencies, we've got to do more with less, we have just got to deal with the volumes.
But often… I talk about the moral contract uh that used to exist between a candidate and a recruiting organisation, which is I put all this effort into your recruiting process, you're gonna put me through these various steps. And my expectation, therefore the moral contract, is if I put all the effort in, I expect to see somebody, human, respecting that effort. So if we drive all this efficiency and we make it more automated, and increasingly we use AI to screen a CV, however clever that is, or we use AI avatars to do the initial phone screen on this. Are we breaking the moral contract with the candidate on this and they're just gonna be going, well, if you're just gonna be a computer-automated process, then I'm just gonna use whatever tech I can get my hands on to try and break your processbecause we don't have a moral contract anymore.
Jessica: Yes, I think in some roles and in some processes, yes. So I think the first thing is there is no blanket approach to this, right? There isn't a, right, well, once we decide we're use AI, it has to be on every single role, every single client, and that's it. Don't care how personalised we wanna make it. I think you've got to target the right kind of roles for sure. Typically high volume, as we discussing, and maybe the slightly more kind of like entry level or lower skilled all these kinds of roles.
What we are seeing actually is that that position can be quite a misconception in some of the roles where maybe like the next generation are typically the people who are applying for them. And where we've used AI and there's a conversational chatbot as a bit of a screening slash FAQ tool, which is essentially WhatsApp. And they can ask some questions, we can ask some questions, but it's 24 seven. And we get really positive feedback because people might be Saturday night and they've applied for the role and they immediately hear back.
And they're thinking, oh, I wanted to ask about the salary, but I didn't want to ask in the email to the recruiter, but I can just ask this chatbot, cool. And they're way more comfortable with that. And then already on Monday, they've got an interview scheduled, for example. So they prefer it. And we're finding that they are going through the process quicker.
So it feels just like a more pacey process. Typically, they might get more personalised feedback because… know, recruiter will get access to that information from, let's say, one of the tools and be able to give real reasons to a candidate because they're not having to manually figure it all out themselves from an application process. In some roles, we're actually seeing the... It actually can make it much better.
They prefer it and we get quite good kind of...kind of NPS scores through the process, I would not recommend doing it on uh potentially some really niche executive hiring type roles where, yes, it's all about building a relationship from day one because maybe your application volumes aren't gonna be naturally that high and you actually want to headhunt or whatever it might be. So I think you have to pick the right roles.
But I think it's commonly expected now for… definitely for the next generations coming through to interact with technology. That it's going to be automated in some way. And that's really common, know, they're using it in universities, they're using it in schools, they're using games, gamified assessment type stuff. So I think if you get it right, and you still obviously have to have a human at some point, I don't think it necessarily has to be a negative for a candidate.
Robert: And I suppose the trick to all of this is being clear.
Jessica: Yes, honest about it. And honest about it.
Robert: That's right. And upfront that you are going to be dealing with an AI type of tool for the first bit, but you'll get more feedback from it. You'll get faster movement through the process if you're the right fit. And then you'll get to a human.
Jessica: Yeah. And the other funny kind of anecdote, I guess, from where we've seen it working, part of transparency is you have to put on the sort of job ad. AI will be using this process or whatever. And in there it says, if you disagree with this and you want to speak to somebody, here, call this number or here's an email address or whatever it is. I don't think you've had one person do that. Not one person.
Robert: They've just accepted, this is fine. You've been upfront about it.
Jessica: Exactly. So that's interesting. Now I'm sure some people will say, well, that you probably already alienated them so they just didn't bother. But… We haven't seen sort of a mass drop off or anything concerning to the point where I'd be like, right, we need to stop doing this, but we have been quite targeted about where we use it and how we use it.
Robert: Fascinating. what about, because I know, and we work together, we'll be working together on a client on this, about how you set candidates up for success when using technology on this, because video interviews in high volume, One is something that a lot of people will be using but we're suffering from a kind of sea of sameness of everybody reading out from a script in order to Get through that stage. So if you got any tips and advice from having done this as to how you avoid that sea of sameness
Jessica: I there's um, there's a bit of education, isn't there with the with the invite to the video interview? I think that you know the example that you're giving you know that there's going to be a bit of a video for those candidates that says, here's how to do a really good video interview. And it basically says, don't read off, don't read off a script because it's really obvious. You know, of course have your pointers. We like that you prepare, but actually it comes off really nice if you're authentic. know, so giving some tips about basically what will probably alienate the reviewers versus what reviewers like. I think in terms of then, if the question would be around, how do you know it's not an AI avatar answering the questions for someone, because those things exist.
The technology is pretty good at basically being able to identify that and identify people, try and leave or whatever it might be and can call it out to you. So that's also a really good way of making sure it's the right person answering that questions. And I've used those platforms for other clients before as well. And it's very easy to spot when people are just reading line by line, because you can see their eyes, the answers are kind way to um buzzword bingo, you know, and you kind of think, well, people don't actually talk like that when they're talking authentically.
So yeah, I think give guidance. The other thing is giving time. So we tend not to be too strict about how long people have to answer it or not being able to rerecord it, but equally not too much time and not too many rerecords because.
Robert: Yes, you've got to get a balance between that. And I think it's just being sympathetic to the candidate's situation on that one because I feel that if you don't give good guidance and it's partly through an email invite, it's partly through the introduction when they do it, but it's also, think, through the career site as well, then candidates don't know that they're falling foul of something.
So if you're dyslexic or dyspraxic, you might feel more comfortable in front of a screen is if in some cases people don't feel comfortable in front of a video screen and they want to read something out because they feel nervous when they've got a camera pointing at them and so we can't just rule somebody out because they've read something because that actually might be an aid, we have to be clear. Yeah. I have to declare it.
Jessica: And what's cool about the tools is obviously you don't have to, call it video interviewing, don't we? But then it does have the option to do voice voice voice. Yeah. If you say someone's uncomfortable with it and then also to put in, you know, if you've got sort of sort of special requirements, it would mean that maybe you'd have slightly longer videos or as you said, no video, whatever it is. So yeah, giving people the space to sort of say that and not be ruled out is obviously really important for the sort of bias angle as well.
Robert: Yes. So we've got to give good guidance training to TA teams, we've got to give good guidance training to candidates on this and we probably have to do the same for hiring managers as well, don't we? you get involved in that too as to how… we explain to them that AI is gonna make life better, but they've got to come on board with all of this.
Jessica: I think that is probably one of the more challenging areas to make it feel valuable to everybody. So we've just had a bit of a company kind of away conference where a big theme was AI and I did the big, here's the macro picture on AI and here's all the cool stuff we could do, but actually on a day-to-day basis, how you as an individual see the value. And it wasn't talking about all the big kind of cool AI tools. It was talking about how can I compose this proposal in one second by typing a prompt in or the X-ray.
And the examples we used, we made sure were quite tailored to different individuals around the business, depending on their role was. And it really landed well. We ended up doing a hackathon where everyone had to generate and build a product in the space of about an hour, which is obviously hilarious. It's not nearly long enough, but you know, it was to get the excitement.
And I think um for hiring managers for our clients, I'm recommending do transformation workshops, it could be led by asks with them, work with them, but even internally, you run it and you get people excited about it and you start to spot opportunities and then you do have to teach people because...If you have never used it before, it's not natural to you to just sort of start leaning on AI to do this, this and this or to trust it even. Yes. And it's definitely a kind of hand-holding process, I think, to show someone the value. And then for us, think where I've seen the most impact on our organisation has been when we can then show the results.
So, you know, when we've been able to do case studies for clients where we said, right, well, we did roll out this technology here and we changed the process a bit. We got the team on board and look, they've managed to reduce their time to hire by 50 percent, they've got a candidate engagement score up or for however many, and we're running more roles but we haven't had to massively augment the team. All those are great um case studies that you can't really get away from and that's when I then get people sending me messages going, oh can I hear a bit more about this? Can we try it? em So it has to sort of like show its worth as well, like proof in the pudding kind of thing.
Robert: Yeah, yeah and I think having those workshops to start with to give people confidence and that is really important because one of the things I came across is almost a shadow use of AI by hiring managers where they go, oh, look, I've got these candidates here. I've got the CVs now. um I'm just going to put that through AI now and ChatGPT to give me some things to call out from their CVs or match it to some prompts. You get this kind of shadow activity and they don't know that actually what they're doing is potentially very, very bad. It's breaking GDPR for starters. But it's, that's again part of where the training comes in is that yes, we want everybody to be using AI, but this is good use and bad use from a hiring manager's perspective as much as it is from TA and candidates.
Jessica: Yeah, mean rule one and one is do not put any personal data through the, especially through the kind of like free tools. You know, you can get your kind of, if you're on sort of enterprise packages, it tends to be, it's secure because it's just your environment. But yeah, but people don't know that. Why would you know that? You know, and so that's what I mean about the gray area of compliance.
I think there's a lot of stuff that people will be doing, but they don't know they're doing wrong. And the unintended consequences we won't know for a while. That's right. And then, you know, the legislation will come down hard on us and suddenly people will realise they've been making mistakes. But yeah, so I mean, my team's role really is to try and get ahead of that on behalf of the company and say, right, these are things we definitely cannot do and train everyone.
Robert:Yes. So final question is having sort of gone through all of that, you obviously run a few workshops and you've seen some case studies where are the one or two areas of AI and automation that you have seen that have given, you talked about 50 % improvement in time to hire, so where have you seen this being most effective?
Jessica: I think it's that top funnel part of process really. I think it's, you know, uh maximising automation where you can in the kind of like sourcing part of the funnel. Okay. And being able to like nurture people and get back to people and make it personalised and relevant to them through automated sequencing. But then I think it's the screening part, getting it right and tailoring it. You know, you can really narrow down the time spent on reviewing applications, speaking to people who weren't ever the right people anyway, and then putting forward the right people for the right roles.
But also keeping maybe the make this sort of the silver medalist type candidates warm doing all that as an individual human, sometimes it's just not possible. AI can help you do that. And so that I think is a real benefit because you can do more with one person and the output is stronger. Then, know, things like just the manual tasks, scheduling. We should not be in a world where people are manually scheduling invites. You know, there's many tools out there. There's nothing wrong with that compliance-wise.
Robert: I think that's where everybody should start really. It's such an easy one to then start getting familiar with how these tools work.
Jessica: And they all connect with your ATSs these days. So that's for me probably where I would look to start and be very intentional about what you choose and make sure it's going to help you in your process. And you're not just going over the shiny thing that's got good marketing.
Robert: Yes, absolutely. Well, you've given some really helpful thoughts to people about how to identify what is more than just a shiny tool and thinking about how that plays into the process. So Jess, as always, it's been fascinating talking to you and thank you for taking the time to come on the podcast and share such great insights.
Jessica: Thanks for having me, I've really enjoyed it.
Robert: Thank you.