Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Insert AI Here: The Talent Acquisition Addition
3 Plays4 months ago
Just a couple of HR execs who’ve done the hard yards, tested the tools, and figured out what actually helps the hiring process when you add AI . No spin ,just what works, what doesn’t, and how to make your life easier.
Recommended
Transcript
Adapting Businesses in Changing Economy
00:00:12
Speaker
people trends, and the reality of building up businesses in an economy that keeps rewriting the rules.
Introducing Amy Crook & Erin Todis
00:00:18
Speaker
I'm Amy Crook, founder of Strativus.
00:00:21
Speaker
And I'm Erin Todis, managing director of talent delivery, which basically means I live in the universe of finding the right people to make an impact for our clients fast.
00:00:30
Speaker
Around here, we talk about the real side of HR and talent, what works, what absolutely doesn't, and how to build teams that can carry a business, not drag it down.
Podcast Theme: Real HR Insights
00:00:40
Speaker
So break down trends, share the behind the scenes of scaling a consultancy from zero, and probably overshare a little because that's where the good lessons live.
00:00:48
Speaker
So whether you're building, hiring, leading, or just trying to keep your company profitable, you're in the right place.
00:00:55
Speaker
This is Fractional Frequency.
00:00:56
Speaker
Let's get to it.
How is AI Reshaping Talent Acquisition?
00:01:07
Speaker
and specifically to talk about the integration of AI into talent acquisition functions.
00:01:12
Speaker
This is a topic that matters so much right now because organizations are grappling with this enormous opportunity in front of them and also some problems that come with it.
00:01:22
Speaker
And to sort of set the stage, I think Amy and I both agree and feel strongly that AI is not here to replace recruiters or TA teams.
00:01:30
Speaker
Rather, it's reshaping how we must work in order to be successful in this evolving world.
00:01:36
Speaker
And so the goals for our session and conversation today are to clarify, simplify, and give you some practical takeaways that you can use right away.
Challenges in Talent Acquisition
00:01:45
Speaker
So Amy, where should we start?
00:01:48
Speaker
Yeah, I think starting on current state, I think would be a good place.
00:01:53
Speaker
So, you know, you mentioned that people are grappling with it.
00:01:55
Speaker
Like, where are you observing that they're having the most challenges?
00:01:59
Speaker
So I think about talent acquisition, the function has measured itself on three core components for a long time.
00:02:07
Speaker
And those are speed, quality, and cost.
00:02:10
Speaker
And organizations always want all three.
00:02:13
Speaker
And it's always a challenge to achieve those.
00:02:16
Speaker
Typically, you can get one or two, but all three is difficult.
00:02:19
Speaker
And the introduction of AI allows us to be more efficient when used properly.
AI Applications in Recruitment
00:02:26
Speaker
I think there are some really fascinating sort of first step applications.
00:02:30
Speaker
When I think about when I was recruiting in my early career, scribbling away furiously in a notebook as I'm on the phone with a candidate and kind of progressed to clicking away on a keyboard, we don't have to do that anymore because the AI assistant will transcribe our conversation and even summarize it for us.
00:02:52
Speaker
or I think about the task of coordinating an interview, particularly imagine a panel interview with multiple stakeholders involved in playing calendar Tetris, trying to compare visually these blocks of time against one another, that the computer can do that for you now.
00:03:10
Speaker
And so I think those are some really wonderful evolutions and the opportunity that we have
00:03:18
Speaker
To take that saved time and apply it to things that are more meaningful, like connecting with candidates on a deeper level, for example, is incredible.
00:03:27
Speaker
But there are some challenges.
Complex AI Tool Implementation Challenges
00:03:30
Speaker
And I think those are coming in the application of more complex mechanisms that are being applied to tools, tech stacks, processes, teams that are perhaps not ready to receive them for one or more reasons.
00:03:46
Speaker
Yeah.
00:03:47
Speaker
No, I think that is such a good point.
00:03:50
Speaker
I've definitely observed in the market that requisitions are taking longer to fill, candidate experience is less than ideal.
00:04:01
Speaker
So let's flip to solution.
00:04:06
Speaker
Let's talk a little bit more about like practical AI use cases.
Implementing AI at Glassdoor
00:04:12
Speaker
So I'm going to start off by talking a little bit about how I implemented that at Glassdoor.
00:04:18
Speaker
So the first step was to really take a look at the function tasks and where it would be most impactful on our time.
00:04:27
Speaker
And so as Erin already mentioned, your sourcing and screening.
00:04:32
Speaker
So everything.
00:04:33
Speaker
recording those conversations so that you can have it already transcribed.
00:04:38
Speaker
You're not having to rewrite your notes from a conversation.
00:04:44
Speaker
You can have your interviews scheduled super quickly.
00:04:49
Speaker
And you can gain those like data-driven insights that are really helpful when it comes to tracking your team's performance.
00:04:59
Speaker
The challenge is I think is that there's so many tools for every step of the process and there's not necessarily like a one tool that will overlay on your full recruitment cycle.
00:05:10
Speaker
So that means that, you know, your process can be pretty much completely changed by having more tools layered onto your ATS currently.
00:05:22
Speaker
So, you know, I'm thinking about this, we've both implemented AI into talent acquisition functions.
00:05:30
Speaker
So Erin, I'm curious, like, what do you feel is like really critical when you're starting or restarting to introduce AI into your function and you really want it to be successful?
Understanding AI Beyond Curiosity
00:05:45
Speaker
I heard something really interesting the other day at a summit I was participating in, in that leaders should stop telling teams to get curious about AI.
00:05:54
Speaker
Oh, yes.
00:05:55
Speaker
And I have heard that from a number of different leaders.
00:05:59
Speaker
And what I thought was really curious about this statement is that the speaker then followed with getting curious about AIs like,
00:06:09
Speaker
giving your seven-year-old a can of gasoline and a box of matches and just throwing it on the floor in front of them and telling them to get curious about it.
00:06:17
Speaker
And when you do that, one of two things will happen.
00:06:20
Speaker
Either nothing will happen or they're going to burn themselves.
00:06:23
Speaker
And I thought that was so interesting because when you think about the concept of AI in the way that I think some organizations are thinking about it, they're thinking about it like software.
00:06:34
Speaker
And AI is not software.
00:06:35
Speaker
It's not something that people inherently understand
00:06:38
Speaker
have some baseline understanding of that they can just learn in a quick kind of workshop or learning lecture type presentation.
00:06:47
Speaker
I think starting with teaching people to understand how the AI works, at least at some level, and then teaching them to communicate with it.
00:06:58
Speaker
These are sort of step one and two before you can actually start to wrap your arms around a particular flavor of the technology.
00:07:06
Speaker
And I think
00:07:07
Speaker
The best way to sort of illustrate that is through one of my lessons learned as I was applying a top of funnel sorting mechanism to our applicant tracking
Case Study: Failed AI Sorting Tool
00:07:17
Speaker
system.
00:07:17
Speaker
It was this shiny new tool.
00:07:19
Speaker
My team and I were the first to implement it as a large organization.
00:07:24
Speaker
And as we were going through the implementation phase, we noticed that for a huge subset of our positions, the tool was not working.
00:07:34
Speaker
It was essentially this tool that would sort
00:07:36
Speaker
candidates or applicants by relevancy and the high relevancy sorted candidates weren't actually good fits for the jobs.
00:07:44
Speaker
And so when we asked ourselves why and started to dig into or sort of peel back layers of the onion, we found that the model was largely built off of relevancy to the job description.
00:07:55
Speaker
And we hadn't taken the time to evaluate or audit the quality of the job descriptions.
00:08:02
Speaker
And so we were using in these cases or in many cases,
00:08:05
Speaker
low quality old job descriptions that didn't articulate well what we were looking for in that talent profile.
00:08:11
Speaker
And so it was sort of like this garbage in garbage out concept.
00:08:15
Speaker
The AI cannot read our minds.
00:08:18
Speaker
And we had this whole backpedaling exercise to go through to sort of tear down and do it right and restart.
00:08:26
Speaker
Yeah.
00:08:26
Speaker
Which, um, was just sort of, uh, almost seems really obvious, but when you're in it, um, sometimes you need to learn the hard way, but tell me a little bit more about your experience.
00:08:35
Speaker
What do you think?
Ethical AI Implementation Strategies
00:08:37
Speaker
Yeah, I agree with you.
00:08:38
Speaker
Like you have to know what the capabilities are of the technology and where it will best suit the needs of what you're trying to achieve.
00:08:51
Speaker
So I actually took the time to do a course and it was less about
00:08:58
Speaker
you know, a selling opportunity, which, you know, everyone's been in those demos where they're like, we can do this, we can do this, we can do this.
00:09:04
Speaker
And you're like, okay.
00:09:06
Speaker
It was more about understanding the mechanisms behind how it works and why it's important to consider the full process, the full implications from a quality perspective, from an ethical perspective and, you know,
00:09:23
Speaker
Also, while it is incredibly intelligent, incredibly smart, there are some places that we need to make sure that we're covering our bases to make sure that we're not opening ourselves up to risk by this machine coming up with its own answers.
00:09:47
Speaker
they're not always correct and we have to be prepared for that.
00:09:51
Speaker
So that was a really interesting course that I did.
00:09:55
Speaker
And so I was able to kind of take that learning with me through the implementations that I did at Glassdoor.
00:10:05
Speaker
So the key things that I think people need to get their head around, and you touched on it here with the job description, is like you have to be very confident in the materials and the data that you have when you start.
00:10:21
Speaker
Your job descriptions, your questioning, I've forgotten the word of it now, the application form.
00:10:30
Speaker
There we go, my goodness.
00:10:32
Speaker
Yeah, you've got to be mindful of all of those steps, the criteria that you're setting, what type of technology you want to use.
00:10:41
Speaker
You know, I was very clear when we put in a tool similar to yours that would go through the resumes and match to the job description that I didn't want something that was going to learn behaviors, because I think that's really dangerous from like an ethical perspective.
00:10:56
Speaker
And so...
00:10:58
Speaker
you have to know and be very confident that all of your steps are crisp before you overlay a different technology over it because it's going to soak all that information in and then that's where you're going to get out as you mentioned.
00:11:16
Speaker
Start small, like implement.
00:11:18
Speaker
I did not implement it throughout the entire candidate lifecycle in one step.
Integrating AI Without Compromising Quality
00:11:23
Speaker
We started small with a tool that would transcribe our conversation so that we could submit candidates quicker.
00:11:32
Speaker
That was something that I observed that recruiter phone screens already take 30 minutes and then they're taking another 20 minutes afterwards to present their notes in a way that's going to be really impactful to the hiring manager.
00:11:45
Speaker
Didn't need to do that.
00:11:45
Speaker
Found a tool that would summarize it in a way that you could just send it straight away.
00:11:49
Speaker
So that saves 20 minutes per phone screen a day.
00:11:52
Speaker
Great.
00:11:52
Speaker
That's working.
00:11:54
Speaker
Second thing that we implemented was scheduling.
00:11:58
Speaker
We already had a really great tool that was scheduling.
00:12:00
Speaker
They created another component to it, which was like an AI assistant that would not only take care of the original
00:12:10
Speaker
um scheduling of the interview like the hiring manager one-on-one but became much better at uh doing panels and rescheduling because i think we all know when we've implemented a scheduling tool um from an ai perspective that the rescheduling thing was the was the thing that was like oh this is our this is our plan
00:12:30
Speaker
point where you need to get a human involved because it's just like, oh no, this is not what we had planned and now it's all gone.
00:12:36
Speaker
So you have to like fix it.
00:12:37
Speaker
But the AI assistant was really good in the rescheduling component.
00:12:42
Speaker
So like, great, we saved a ton of time scheduling interviews.
00:12:46
Speaker
And then the final thing that we did was resume review because as we know, at the moment, application volumes are huge.
00:12:54
Speaker
Recruiters are overwhelmed.
00:12:56
Speaker
And so I wanted to implement something that was going to shave down some of that time so that we could spend more time with the candidates that we were really trying to build a relationship with.
00:13:06
Speaker
The key thing is I did not want the experience for hiring managers or candidates to be impacted.
00:13:13
Speaker
So that means
00:13:14
Speaker
communications are still personalized um we take the time to pick up the phone to candidates that we're trying to um build a relationship with as a potential employee um we give feedback um we consult with the hiring manager you know it's not just that okay
00:13:36
Speaker
AI has done a wonderful job of presenting this profile, but we had the conversation.
00:13:42
Speaker
So we want to share those nuances that you pick up from human interaction that do not come up on a transcribed note.
00:13:51
Speaker
So I wanted to be able to keep that quality experience that we are so well known for and not compromise our days to fill, even though volumes were incredibly high.
00:14:04
Speaker
So my motivations were to save the time to be scalable so that we could take on more work without having to increase headcount, but still not compromising the experience.
Importance of Metrics in AI Implementation
00:14:21
Speaker
I feel like I was just talking for a really long time, but I do want to quickly touch on metrics because I think another thing is you've got, you know, leaders that are very much like
00:14:35
Speaker
Get curious, you know, rather than like having some very clear goals because it's so new for a lot of companies.
00:14:41
Speaker
They're not really sure.
00:14:42
Speaker
They know they want to create the efficiency.
00:14:44
Speaker
They know they want to implement their technology because they're being told from higher up that like,
00:14:50
Speaker
you got to do it.
00:14:52
Speaker
So they're passing that message down.
00:14:54
Speaker
And so there's almost like a sigh of relief when they find a tool that they think will work and is in budget and they've implemented it.
00:15:02
Speaker
It's like, oh, okay, we're done.
00:15:03
Speaker
We've hit it.
00:15:04
Speaker
We've implemented AI.
00:15:05
Speaker
We're winning.
00:15:07
Speaker
But that's not...
00:15:09
Speaker
how it works with like all new technology there's going to be kinks there's going to be challenges um something that i like to say when i'm consulting with people about ai integration is be careful putting new technology on old processes like you have to have an open mind
00:15:27
Speaker
to map through the process to make sure, okay, we're layering these tools on.
00:15:31
Speaker
Are we still keeping the experience in mind?
00:15:34
Speaker
Are we still getting the information that we need to make good decisions?
00:15:38
Speaker
Are we still consulting with the business?
00:15:41
Speaker
Are we just now,
00:15:44
Speaker
in a factory of people you want to avoid that um it's not good for either side it's not good for the company when they're trying to attract talent um that's going to take the take people to the next level in their organization and it's not a good candidate experience either so quality is really important
00:16:05
Speaker
Experience is really important.
00:16:06
Speaker
And is it even moving the needle on your efficiency?
00:16:09
Speaker
If it isn't, don't keep going.
00:16:11
Speaker
Like make changes that you need to make it
Consulting Advice for AI Integration
00:16:15
Speaker
work.
00:16:15
Speaker
Similarly with the scheduling tool.
00:16:19
Speaker
You know, we didn't just set it and forget it.
00:16:22
Speaker
We were constantly maintained probably for six months to see like, okay, are the interviews still getting scheduled within the SLA?
00:16:28
Speaker
Is there any problems?
00:16:29
Speaker
Is there any no-shows?
00:16:30
Speaker
Is there any...
00:16:32
Speaker
any challenge, all the things.
00:16:33
Speaker
So yeah, I, I, again, I've talked for a long time, but I do feel like, um, it's not as overwhelming as people might feel.
00:16:42
Speaker
Um, but there are some like very key components that you need to take with you as you move forward into your integration to make sure it's successful.
Free AI Integration Consultation & Insights Report
00:16:50
Speaker
I think you made some great points about not being afraid to get back to basics, both with,
00:16:56
Speaker
knowing your process, knowing how recruiters are spending their time, do a time study, ask everyone to log how they spend every 15, 30, 60 minutes and understand where time is being spent today before you plug in the tool so that you can actually compare and contrast with some certainty and confidence about where the tool is actually giving you lift.
00:17:17
Speaker
I think those are incredible points.
00:17:20
Speaker
And so to summarize, maybe we should wrap this up.
00:17:25
Speaker
Yeah.
00:17:25
Speaker
done a lot of talking.
00:17:27
Speaker
Yeah, we're good at that.
00:17:29
Speaker
And I just, I think really that if this is a challenge for you right now, which is totally understandable, you're not alone, but I just want to offer out there, if you want a free consultation with me, I'm more than happy to do that.
00:17:45
Speaker
As I said, Strativist, myself, I've implemented this in TA functions.
00:17:49
Speaker
I've kind of done the legwork so I can definitely help talk you through what works, what's challenging and just give you that support that you need.
00:17:59
Speaker
Because I know how it feels when the pressure is coming from above to create the efficiencies, be able to demonstrate that you your production value goes up, be able to talk about it intelligently, how you've improved.
00:18:14
Speaker
implemented it.
00:18:15
Speaker
So you're getting that direction, but not necessarily the what or the how.
00:18:20
Speaker
So I'm definitely here to help.
00:18:21
Speaker
If you just reach out to me, I'd be more than happy to set up that call and we can talk through some of the challenges that are affecting you today.
00:18:30
Speaker
And quick plug, follow Strativus on LinkedIn.
00:18:32
Speaker
We'll also be releasing an insights report before the end of this year.
00:18:36
Speaker
And we plan to have that released quarterly.
00:18:39
Speaker
And so check that out.
00:18:40
Speaker
Awesome.
00:18:41
Speaker
Thanks, everybody.
00:18:42
Speaker
Have a great day.
00:18:42
Speaker
Cheers.