Breaking Bias: Skills-First Hiring for Inclusion with Kathryn Marie | SocialTalent

One of the things that often gets overlooked is the other people in the team. So when we're hiring for someone, we need to have a think about the diversity of the people that are in that team or the diversity in that company. And then we're able to select people based on what their needs are for that role. So a lot of the, a lot of the skills that we hire for can be, can be trained, but those innate abilities are something that we're naturally going to be either really strong with or not.

Hello and welcome to the show. I'm Johnny Campbell, CEO and co founder of Social Talent. And this week on the podcast we're talking to Catherine Marie from Melbourne, Australia about Skills First Hiring. Katherine, you're very welcome to the podcast. Could you give our audience a little introduction to yourself?

Absolutely. And thank you for having me, Johnny. I'm very excited to be here. Yeah, I'm Catherine Marie and I founded a company called Brokeries which is basically a skills based hiring platform to, to simplify the, the ability to do skills based hiring.

And so Catherine, you know, there's a lot of talk about Skills Forest, skills based hiring. Why you, why this particular area? What brought you to where we are today, where you've put all of your love and attention into a business, into an application, into a, into a, a solution for skills first hiring. You must really love this. Tell us why.

Absolutely. I'm very passionate about it. And the reason being is that a lot of my working life I've worked in male dominated industries and I've wanted to, to get more women into these industries. Now the issue was always how do we do that? And the answer is skill, skills based hiring. So I've put a lot of work into building this model of all this framework of skills based hiring and it's built on the, the multiple intelligence model that, that will encapsulate all different industries and all different people and all different roles.

But, but tell me, so, so you know, from, from where are you coming from? Because obviously you've mentioned different models that our audience may or may not understand. It sounds very academic. It sounds like it's got science and research. You said you've worked in very male dominated industries. Tell me about the industries you worked in. You know, you worked in it. You weren't originally from a recruiting background. You came to the space of HR recruiting later on. So walk me through how a journey of your career and how those unique careers you've had have beautifully combined to, to, to position you as the person who should take this kind of scientific model and use it for skills first hiring. Yeah.

And so I was working in tech and dealing a lot with a lot with people who were trying to get women into the, into the tech industry. But the issue tended to be that women just aren't applying for the roles. And how do we do that? So how do we solve the issue of women not applying for these roles? So I had to think about my previous training as a teacher and how, you know, how our brains are all different and how we operate in different ways. And I thought what I need to do is come up with a way that I can match women to roles in tech rather than them applying for roles. Okay. So that meant I would have to build a framework that someone could fill out a profile and say, these are all of the things that I am really good at. Now one of the groups of people I had in mind were women returning to the workforce. So it may be that they have had a family and they've got a gap in their work history. It could be two years, it could be 10 years. And how do we get those women to re enter the workforce and hopefully get them into tech roles? So by getting women to fill out a profile of all the things they could do, they didn't need that job history or that work history. All they needed to do was to tell us a bit about themselves. Okay, so how do they organize their day? How do they communicate with other people and have these open ended questions that they could respond to and feel confident that they could respond to them. So women often will use their will, often downplay what they're good at. They'll, they'll say, I wouldn't be, you know, I wouldn't be able to do this role because I don't have the skill set that is required. I haven't been working for the past 10 years. But what I'm trying to do is get people to think outside the square a little bit. And to do that, I want people to start reflecting on what they do in their day to day life and what that shows about their abilities.

Katherine, can you give me an example of one of these types of questions just to kind of get a sense of what you're trying to ask somebody and then what kind of information you might uncover?

Yeah, so for example, one of the questions would be how do you, how do you know that you've understood what someone has said to you correctly? So it's, it's about the types of questions they might ask that person. Now every person is going to answer that Differently, which is a really powerful tool because it gives us an insight into the, the type of person that, that they are. And we can then use that information to match them to, to a role that they would be perfect in.

And talk to me about the science behind this, because this is founded upon an established principle, our principles around trying to match, you know, certain question types to skills. So, so what's the history of the science behind this?

Yeah, so it's, it's based on the theory of multiple intelligences. And the reason why I use this framework as a basis for what I've done is because it covers all the different types of ways we operate and how we respond to things. So it means that if someone is filling out this profile, there's going to be things that they can fill out. If they're, you know, if they haven't been working for the past 10 years, they've still got skills that they can use. So when they start filling out this profile based on multiple intelligences, it doesn't mean, means that if they're creative, they've got some really great stuff that they can put in. If they're analytical, they've got some great things that they can enter as part of, as part of what they can do. So that then gives everyone the ability to, to give responses. And it also means that it's a more positive experience. So it's not I can do this and I can't do that, it's how do you do this? Like how do you go about it, how do you communicate in this situation? And so it's more inclusive of everybody and all their different skill sets and all their different types of skills.

I want to come back to the model you're using and how it works in practice in just a minute, Catherine, but maybe can we pause for a second and would you mind explaining to me your definition of skills first or skills based hiring and you know, perhaps what people think it is, what it can be applied to, what it's useful for and then maybe we can dig into some of the challenges of it.

Yeah, I'm glad you asked that actually, because when it comes to skills based hiring, there, there are a number of different things that, that we can talk about. So we can talk about whether we need to test whether someone can use a certain bit of software or whether we give them tests on writing a bit of writing something out. And so, so we can test that way. When I talk about skills based hiring, I'm actually talking about our innate, innate personal attributes. So what I'm Talking about is how we, how we think and how we operate. Now those, those sorts of attributes are really important because when we're looking at getting someone into a role, it's often that they would be able to learn the skills to do that role if they have those correct innate attributes. And those things with us from birth, they're things that we're naturally good at and obviously develop over the years. But those innate things, we don't need a work history to prove that. We just need to respond to a question that is, well, how would you go about doing this? So when I talk about skills based.

Sorry, no, no, keep going.

When I talk about skills based recruiting, I'm talking about what people are naturally good at and how they operate as a person.

So it's not you, you, your model of this is not about, you know, you're proposing, for example, that we use this model to hire doctors, to hire lawyers, to hire an experienced architect. It's more that perhaps we, we use this model to consider candidates to perhaps come in near the entry level, entry to middle level of a career and then they will get the support, experience and additional training to, to build on those natural strengths that they have rather than experience. Is that a correct understanding of what you're saying?

We can actually do it in both ways. So if it's the entry level position and the person needs to be upskilled into the role and to develop their career, we can get people to fill out their profile and we match them to roles that they're matched with. Now if we're talking about a more technical role, so something that they have to have qualifications for because we don't want to, to use that methodology to get someone to do a heart operation on us, what we do is get a short list of the candidates for that role that are based on their qualifications and then put them through these attribute tests. So what that would mean is that a person who is applying for that heart surgeon role, they would then need to demonstrate that they're the right person to fit in with that team and to fit in with that company. So depending on the hospital that they work at and the team that they're working in, they're going to need all different sorts of attributes. Okay. So if they're working in a team where it needs to be, everything needs to be clear cut and there's no room for messing around, then they're going to have to be, you know, that analytical type. This is how things need to operate. But if they're working in an environment where They've got nurses around that, you know, like to have a chat with their doctor. And they, they're the type of people that need someone who's maybe better at communicating than in a different team. So all of these teams function differently. And if we need the qualifications to be vetted first, then we do that first and then put them through the broadcrees process.

So it's interesting because it's not just perhaps the what which might be the skills that they have. It's also potentially a way of assessing the how. So you can use a different model to assess the what, the skills that could be technical, that could be educational, and then how would they put those skills into practice In a team or workplace? There might be particular needs you have there on the how, which might be the behaviors, values, people give them different names. But you're looking at those innate abilities to. It could be, you know, how they make decisions, how they communicate, how they work well within a team, et cetera. So it actually has quite a broad applicability and isn't a, isn't necessarily a competing methodology to educational requirements or other testing. It can be very complimentary from what you're saying.

Yeah, absolutely. And I like to think of it like if you have two different companies, Company A and company B, and they're both hiring for finance managers. Now, company A might have a team of really, really strong people that that person will manage. And what they're looking for is not a person who is in finance at all. They're looking for someone who is really great with leadership and inspiring the team. Yet company B, they've got, they might necessarily not have as strong a financial team. And so they need someone who really knows what they're talking about when it comes to finance. So one of the things that, that often gets overlooked is the, the other people in the team. So when we're hiring for someone, we need to have a think about the diversity of the people that are in that team or the diversity in that company. And then we're able to select people based on what their needs are for that role. So a lot of the skills that we hire for can be trained, but those innate abilities are something that we're naturally going to be either really strong with or not.

And Catherine, you've talked about how you've used AI in a rather novel way in your solution. But we hear a lot about AI in everything, of course, but particularly in skills first hiring. We've seen the last couple of years a whole lot of applications that propose to use AI to fix this, to match and do this really well. And we've talked about how a lot of those approaches are flawed, in fact, and people don't necessarily stop to think about it. And I hadn't stopped to think about it until we had a chat about it. But I wonder, can you walk me through before we talk about exactly how your system works and the novel way you've used AI? Because I think it is really novel. I'd love for you to share your opinion on some of the flawed ways that AI is being used to perhaps seemingly tackle this challenge where it might actually create more problems than it.

Yeah, absolutely. So one of the issues with using AI in this, in the recruiting process is that AI is biased. It has biases because it's trained on the, the information that humans input, and we all have biases. That means that the results we're going to get from an AI will be biased as well. So let me put it this way. The recruiting software is using AI and it's trying to tell you the sorts of skills that you're going to need for position A. Now, what it's doing is looking at the massive amount of data it's gotten from job boards, creating like a median of the average things that, that are required by those jobs based on everything it's learned. But what the AI can't do is figure out what your job needs. Okay? And so it's going to give you a list that says this job requires skill X, Y and Z. So it needs creativity, it needs great communication, and it needs someone who is really organized. Now, your position may not need those things at all. And I've actually seen this live in action where someone has shown me that the AIs are coming up with matches that a real person who knows the business and knows the team and knows the role wouldn't come up with. They're just the wrong. They're the wrong matches. And they're the wrong matches because it's basing it on all this generic information that it's gotten from all this massive amount of data. And of course, because the massive amount of data that it's using has biases in it, it's going to reinforce the biases and make the situation worse when it comes to trying to hire for diversity.

And so if I look at the promise of AI, which was that we would be able to flick a switch and match candidates and perhaps look at these extra skills, what you're saying is actually it regresses to the mean. It, it does very average matching. So if you are an app. You are looking for the average and your role is somewhere in the middle and the same as everybody else is. And it's generic. Maybe it's fine, maybe this works for you. But your point is that there's a richness to human experience and background and the difference between companies, cultures, teams that is just missed by this. So, so we've talked about, you know, the, your methodology and how the methodology you're using to assess a candidate and using these questions. And one thing that struck me, which was really interesting and unexpected for me, was that the questions aren't multiple choice questions. They're free text, open answers. And you hinted at this earlier on when you're talking about how everybody will answer that completely differently and you really mean that it's not that you'll take a different multiple choice option, which is what we kind of expect from these surveys. It's, it's oral or written answers that are just free, free text or free speec for someone to describe how they think or how they work or how they approach a problem. And it's on average, how many questions would you typically ask a potential candidate to try and get a broad enough profile they can.

We start with about 10 questions. And so the list, there's a long list of questions that people do need to answer or they don't need to answer all of them, but there's a long list of questions and how I approach it is I want people to look through each of the questions and answer the ones that resonate with them mostly. So the ones that they're going to be able to easily answer and they're the ones that they do first. So scroll through, say, yep, I can easily answer that. I, you know, don't really know how I'd answer that, so skip that one and go to the ones that are easy for them to answer because they're obviously the, the ones where their strengths lie.

Gotcha.

The other thing, the other thing with the this list of skills is I haven't. Or attributes is I haven't labeled them. So I haven't labeled them, you know, communication styles or I haven't labeled them analytical. And I did that for a very specific reason. And it is because we're all biased, but we're also biased against ourselves and what our capabilities are. So what I've tried to do with this, with this list that people or this profile that people complete is to make it so they can start to think about the types of things that they are good at without, without checking themselves out before they even read the question. So they look at it, they, they see a prompt that says, oh, you know, it might be when you've done this, how have you done it? And they go, oh, yeah, of course I've done that. You know, and if they'd seen what the skill was at the top, they might not have even read that question. So it's, it's, it's supposed to be a, a powerful experience for people to, to start to delve in and dig into, you know, what am I really good at? As I said, we often don't apply for jobs or don't even read a position description based on how biased we are against ourselves or what we think about the assumptions that we make against ourselves. And I want to take that out of the mix because if we're trying for true diversity, we need to take the bias from the companies, but we also need to take it from ourselves. And we need to answer honestly about, I can do this and I can do that, and it's supposed to be empowering.

Walk me through the other side. So if we're getting this really powerful view of someone's innate skills, abilities, strengths, it's not based on necessarily work experience. How do we match that to the organizational requirement? Is that just still matched to a typical job spec or what's the process on the hiring side that we need to do to match this data set?

Yeah. So on the hiring side, we work with companies that are open to improving their diversity and for their roles, we work with them on what are the actual tasks that that role needs. So what are the 10 main things that that role needs? So we delve into the task might be answering, answering correspondence. Okay. We look at how that translates into what type of skill a person is going to need. Now that will change depending on whether that correspondence is because they're answering the complaints email or whether they're answering the feedback email. There would be different skills that relate to that task. So the task might be the same, but the skills can be different based on what the, what the actual role is. So we work with the company to try and figure this stuff out and to, to talk to them about, okay, you know what, when you're talking about answering these emails, what are they answering? Like, what are the sorts of qualities they're going to need to do that effectively? And also if they've got strengths in certain areas, then they're going to be able to complete these tasks without that being a stressful experience for them. So if I'm not naturally Good at conflict. And you put me into a role where I'm answering emails that are complaints emails that'll mean that that role is actually going to, to make me stressed and you know, I'm not going to be happy within that position. So you know, answering emails, absolutely I can do. But what is the actual skill or the, the innate ability that you want the person to have that is doing that task?

What's that requirement like? Are you looking to get 5, 10, 15, 20 tasks? Is it an interview forum?

Yeah, yeah, 5 to 10. So picking the, the most important ones, the, the key, the key skills that, that role will, that role will need to, to be able to succeed.

So it's not exhaustive. It's not trying to find every single task and break down every single skill, which probably.

No, no, absolutely not.

That's scary. So it's getting, yeah. The top five plus skills. And you know, is that again through a survey or is that through an interview or what's the process that drives that?

So at this point it's by working with us directly and working through the, their position description and talking about what, what they actually need for their role. So it is an interview at, at this point, the, the goal is to, to step people through this for their first maybe five positions that they list. After that they should be getting pretty good at going, okay, this is because they need to think about the job differently. They need to think about it as you know, I don't need this work history. I need this type of person that will be able to do these tasks. So after, after doing it a couple of times, it should be straightforward for them to do on their own. And they're just picking those, those five, maybe 10 tasks and what are the innate abilities those people need to succeed in those tasks?

And I guess this is where the AI comes in because matching that fuzzy data to two data sets of, you know, free text answers, you know, in a traditional manner will be very difficult because we like to structure data, match the structured data, and that's how most systems work. But that's probably the promise of AI and G generative AI is that you can match these two seemingly unstructured data sets together with a high degree of confidence in terms of their match. So talk to me about how that came about, how you test that, what are you trying to achieve there and how does it work?

Yeah, so what it does, absolutely, like you said, it matches the text that a user has put in, a client has put in, and it matches the text that the, the company is entered based on that role. So it is, I can't tell you a lot about how the, the AI works behind that, but it's giving, it's using a scoring system. So the scoring system will, will score a user based on what they've entered and then it will match it with which, with a, a task that the, the company is entered and the scoring system that they've got. I don't want to tell you too much.

No, I don't want you to give away the secret sauce. But I think we can get, we can understand that you put both data sets into a pot that you have put structure into, powered by AI that then comes out with a match. And so tell me about how, what have you seen so far? Who are you working with? What kind of examples can you give us of, of where you're testing this, the kind of use cases you're seeing this work on or be strong on? What are some of the outcomes that you've begun to see from this? Catherine?

Yeah, so I'm working with a large international manufacturing company at the moment. Now they've got a policy of hiring more women because they don't have enough women that work within their organization. So they've got policies around whether or not they're allowed to even interview based on if they have women applicants. And what I've heard from them is that, that women aren't applying. So there's so women self select out of, out of roles based on a lot of different criteria. And it may be the company name that they're just like, well, I don't want to work for that company because I don't think I would be a fit for that company when in reality these are exactly the type of people that that company wants to work with. So we've put in a couple of, a couple of roles that they have that they're wanting women to apply for. And these are roles which are perfect for their testing of the system because they're roles that they need to be people need to be upskilled in when they're on the job anyway. So it wouldn't matter if that person had work experience when they start that role. The the tools and things they're using are in house tools. So unless they've been working for that company, they would never have seen them before. Okay. So yeah, we're now doing some matching on, on those particular roles and we're getting some really great results on, on a list of candidates that will have the ability to be upskilled into those different positions. So you know, we have one role which is to support their software, their internal software and to support the people using it, and another one which is a sales sort of support role. And so, yeah, we're getting some really good results out of the matches that we're getting on that.

I guess if you could fast forward to a future where, if you think about LinkedIn today, where it has maybe a billion people who list their experience and qualifications, imagine if the same database also had everyone complete a similar test or go through the questions and complete that profile. So it had your skills and qualifications, but also had this, this background of your strengths. And you could then as an organization go to that pool of candidates and go find the people you already know are going to be a strong match and go tell them that you've done the analysis and they're a strong match and there's take it to that next step. I think that's kind of what you're proposing, is that it's a different way of approaching recruiting. It's not just advertising and hoping somebody comes and then it goes through an application process. Because your point is that what you have seen in your careers, the people that the right candidate, right being the person who generally doesn't apply or get the job, that they're just not going to go through that process, they will self select out, they'll never consider themselves where they could be the perfect candidate for us for that role. So Catherine, this could be used by any type of organization hiring for any type of skill or any type of background. But your focus in particular is around perhaps underrepresented talent in different areas. You know, you've got your own personal background, family backgrounds, where you see people being excluded from society, from, from the workplace, from opportunities either by the organizations or they're self selecting themselves out for different reasons. So talk to me about your real passion for how you'd love to see this type of a technology or process scale up.

So what I've been doing is partnering with not for Profits. And what that does is give me that, that talent pool of the unrepresented talent that we're talking about. So whether it's whether it's based on gender, whether it's based on culture, whether it's someone who has a disability or it could be someone who's neurodiverse, if I partner with a not for profit that specializes in assisting those people and then that becomes the talent pool that I match against the different roles, it means that we're opening up a whole wide range of New talent that would have normally never been accessible to the businesses that are running their jobs through brokeries.

I love that because I think, you know, we've both discussed how there are so many pools of underrepresented talent who could do the job, who can do the job, who have the skills and abilities, but perhaps who don't apply, aren't great at getting through an interview, don't have the confidence or just don't have the CV for it or the background because no one's ever given them a shot. But to your point, they have skills. We all have these innate skills from just being alive by, by, by engaging with society, by using our brains. And it's about how do you basically allow someone the opportunity, opportunity to use those skills and to show that to another organization, to an organization who need those skills, who could put that to work, who currently have that gap? Catherine, you know, if someone's interested in your this story, they're interested to hear more. How can they get get in touch with you? And also what kind of person or persons or organizations would you like to get in touch with you? Like what kind of folks do you want to hear from?

Yeah, absolutely. Anyone can get in touch by going to my LinkedIn. So Catherine Marie, founder of BroadCareers, you'll find me there. Or I do have the business page of broadcrees that you can also have a look at. And the types of people I really want to talk to at the moment are those people from the not for profit that are trying to find work for these underrepresented groups and they're struggling to find jobs for those people. So this is like a bridge between, between a not for profit who, who wants to find people work and a company who's trying to improve their diversity. Okay, so if you're a not for profit and you're supporting people getting back into work, please reach out. If you're a company that has issues with diversity and you do want to improve those issues, then let's have a chat, we'll put a couple of jobs through. It's completely, of course, it's completely free to, to go through that process. And so it's, it's, it's, it's a no risk way of, of testing for yourself if this works because you don't have to pay to, to, to go through the process. If you happily get a match, then that's, that's a different issue. But, but it's, yeah, it's a no risk way of testing and getting more people to hire. So even if you interview five people that have been shortlisted for your role, you've at least interviewed people from a not more diverse group, even if you haven't hired them. So you know it's a no risk way of, of starting to improve your processes and yeah, so, so anyone can reach out, anyone who, who is not sure how they would get started. So maybe they're not with it, not for profit, but they're that stay at home mum who, who is like, well, how do I.