Episode 7

Marta Turek-Olearczyk on the Future of Hiring in Data Roles and Balancing Talent & Technology

TIM: When it comes to hiring data roles, in your view, what are the biggest challenges in the market right now?

MARTA: I think there are two things that stand out to me. And they're connected and that is the conformity of applications and distinguishing man from machine in the application. That's the biggest problem because in my view given data is already a very technical field with a high degree of jargon. When you are receiving a large number of applications that all sound the same, then none of them appear to be viable candidates. even though they might be.

TIM: And they all sound the same because you're implying that everyone's using the same tool to generate them. And maybe the CVs aren't created by humans anymore or.

MARTA: I'm, so I'm implying that for data roles, and particularly when you start going to data science roles there, there is already a high degree of conformity and jargon that is used in the application. And then when you, because you're mentioning different Python libraries and scripts and things that you're applying And then when you start to use AI to write the application for you as well, then it just increases that level of conformity and that's, it makes it really difficult to identify who's a strong candidate and who isn't who isn't. Silence.

TIM: through all this university or this university, this weighted average mark or this weighted average mark. So I'm interested in like, when you've been faced with this challenge at the moment, how have you dealt with that? Like, how have you ultimately selected who would get an interview versus not looking at this sea of kind of similar looking CVs?

MARTA: So something that we've included in the interview process is a number of filtering questions. So they are free form questions that the applicant has to answer and the manner in which those are answered. I actually find myself now focusing more on the answers to the questions there. than I do to the resume itself. I will look at the resume to see if it fits, and then I will look at the answers to these questions to determine whether the answers are thoughtful if they have if they're also, there's sometimes I see very similar answers to these questions, which indicates that. a human didn't write the answers. And sometimes they leave them blank, which also tells me how interested they are.

TIM: And with these questions be like a mixture of tell me your experience in X or how did you solve why problem or are they like almost quiz questions where you're assessing a bit of knowledge? What kind of questions do you include? Silence. Silence.

MARTA: of how they've used those. So we're looking for contextual answers that indicate that this person actually has experience with the range of tools or applications that they've listed.

TIM: So it's almost they have to provide a a specific enough answer that it's believable in the same way that if you're interviewing them and you're asking these questions, the more detail they can give, the more obvious it is that they do have the experience as opposed to a superficial vague answer.

MARTA: exactly. The more specificity, the better.

TIM: and I guess Chat GPT would be especially good at giving a vague non committal type of response that might be easy to spot as a Chat GPT generated answer.

MARTA: Yeah, and what's really interesting is that you start to see this commonality in the answers as well. And then you start to realize, okay, this is how, this is probably written by a machine. And also, ChatGPT has a very particular style of writing which is quite easy to pick out.

TIM: one pack that I picked up on last time we were doing hiring was let's say there's some kind of open ended questions like that. And the candidate has to write something. The not particularly savvy ones won't think that as part of this process, we have some evidence of bias. What is definitely their real writing. And so then if they then copy and paste in some chat, GPT content, it's very obvious to compare those two and go some, someone different has written this content and so they'd have to be quite clever and doing it in a kind of consistent way without making it super obvious. Like for example, even things like grammar, spacing, organization, versus how a normal person has maybe not perfect in English might write it. It's becomes, I noticed like a very stark contrast. Silence.

MARTA: humanity in the applications, whereas. In the past, I used to, before, before AI, before ChatGPT, I would put a lot of weight on the use of formal English in the application because it's a formalized process. And now that is actually something that, that the formal English is a very particular style in which ChatGPT writes in. And so that has become a hindrance to the application because it's now so ubiquitous.

TIM: Yeah. And so you mentioned two challenges, so one was this, yeah, differentiating the human or the machine behind the application. And what was the other one, so

MARTA: they're connected. I said, basically the conformity of the application. And I think that it's the machine because the machine is contributing so much to the process. It is increasing that conformity. And also because we're in such a difficult job market for, it's an employer's market right now, there isn't much out there for job seekers. And so because there's such a high volume of applications it seems, it feels like applicants are putting more weight on quantity versus quality. And so they might be spending, they might be sending a large number of applications, but not spending enough time in each one.

TIM: yeah, that's something we've heard so consistently recently. Yeah, it could be, we'd never have this metric, but if you could almost measure like applications per applicant, like how many jobs is the average person applying to through time, some of the job platforms would be able to track this, not sure they've ever released it, but that would be an interesting metric. And yes, a sign of the times of the desperation in some sense or whatever. I feel like also because there's like increasing distrust in hiring in general on both sides of the market, a candidate is probably Oh, this system's wrought, I have to apply to as many jobs as possible because I know I'm not going to get a fair chance. That then creates a bigger problem for them because there's more applicants and then they can't stand out from the crowd. It's like this never ending cycle. You could imagine.

MARTA: Yeah, exactly. We recently posted a data analyst role and we received 250 applications in a period of 48, 72 hours. And and that and the And this is an opportunity for the tools that we use for hiring. There isn't a clean way to filter out a large percentage of those applications. It requires someone from the people and culture team to manually review them.

TIM: Yes.

MARTA: So it has really increased the workload to hire someone to go through that, such a large number of applications.

TIM: Yeah. And I think not only is it a lot of effort and incredibly tedious, but just like rife with bias. When you have someone's document that it reveals like their gender, their ethnicity, sometimes their religion. In Europe, I know people like to use photos, which would be weird in Australia, but common there. I know from recruiting in bits of Asia, like it's common to put Roman Catholic in the Philippines, in India, it's common to put your caste. Just like all this stuff that I just, I don't want to know and shouldn't be part of the hiring process, obviously, but it's just front and center when the hiring process starts with a human and a CV. So I, I think the opportunity to use AI to in principle, at least remove some of that bias is like a huge opportunity.

MARTA: I agree with that. I guess though, at the end of the day, it is cultural, like you say, in Europe, it is common to include a photo with an application. Whereas in other countries, it's completely unheard of to do that.

TIM: what about if you think to your hiring experience, either where you've been the hiring manager or you've been, privy to a process is there any glaring error or like a, a bad don't like to say bad high, let's say a regretted hire or something about the process that really went badly that taught you something or that you learned something from?

MARTA: my biggest learning and I'm lucky that this happened many years ago, and it's it's a Holy Grail for me now. Is that in the past, I would. I hired without doing a technical assessment. And this is when I was still hiring for technical skills in Excel. We're not talking about, Python or data science. But even then with a tool such as Excel it became really clear that some people are just really good at talking. And they can, they sell themselves really well and it appears that they know what they will need to do on the job. And then once they get to the job, their self assessment of expert level in Excel turns out to be an equivalent of your beginner assessment. And so I I learned very quickly that doing an assessment Really early in the hiring stage, just right after you've got your people and culture interview, which is the first kind of level filtering, and then there might be a conversation with the hiring manager, and so either before that conversation with the hiring manager or after that conversation, that's when you introduce the assessment. And I've in the past, I've done. Like live assessments where it's in office, but that was. before the world of remote working. And now the assessments are take home assessments. So the live assessments are obviously much simpler. People are incredibly nervous doing the live assessment. And so you also need to give them a bit of grace when you are assessing them when no, and I understand I always, you make the worst mistakes when someone's watching over your shoulder and looking at what you're doing. But yes assessments I think are critical in any hiring, in any interview whether it is technical or for data or otherwise. But for data in particular, I think it's absolutely critical.

TIM: Yeah, I feel like so many of the problems in hiring, if I think about it now stem from taking the candidate's word for it. Instead of just saying this is what your CV says, therefore, I'm going to interview you've done an interview talking about these things you've done in a sort of behavioral interview style of thing, I'll accept that they were your experiences. And so it's just set up for failure in that sense, because it's not a level of validation of what does this candidate actually know? Are we measuring anything like that? And I feel like, yeah, that's where a lot of hiring goes awry. And it sounds like you, you learned this early on. Is that because of a person you hired in particular before you decided to do assessments that didn't work out so well? Did you get burned by an especially bad experience or you just saw a better way of doing it?

MARTA: to be perfectly honest with you, I don't even recall if I don't think it was

TIM: ahead

MARTA: a poor hiring experience. More so it was related to

TIM: No, it's not.

MARTA: to Objectively assess the strength of candidates without an assessment. It, so being a data driven person I missed that level of objectivity. Can they do the job or can't they do the job? And then As you said, as the whole hiring process seems to be set up a little bit for failure because it's based on subjectivity and rapport and ability to speak with someone. But particularly as the role gets more senior you're speaking with people who have a lot more experience and who are able to have any type of conversation that you want them to have, and they can read as well as you're reading them. And so without that objective I think that it leads to many more poor hiring decisions.

TIM: candidates have done, to be honest, I couldn't even remember the first candidate of the day, that I interviewed and everything just becomes this blur of a feeling and a sentiment and an intuition. It's just, that's where our processes break down very quickly. I think if there's no tangible, like this is what we're evaluating. Here's how they scored. Even if it's just like a yes, no, or a good, okay, bad, like some, something has to be structured. Got it. Silence.

MARTA: agree. And especially when they're. These days it seems like hiring rounds are just getting longer and longer where candidates are going through four, five, six interviews that when you're hiring for a technical role, it's so important for the data leader to feel confident in the skills before they Send that person off into the next rounds of interviews because those will be with senior executives. Those will be with non data team leaders. And so you need to be sure that you are pushing through a candidate that is able to fulfill the role. Especially when they're going to be speaking with people who are no longer able to make that technical assessment and are assuming that you already have.

TIM: Yes, so you're rightly and fairly viewed as the gatekeeper and your wall. The hiring manager says they got the right technical skills. I take that as a given. So you'd almost not have done your job in a sense if you didn't have that measure.

MARTA: Exactly. Exactly. And it would be I would be questioned for why did this person make further in the interview process if we deemed that they were not capable of doing the work?

TIM: What about this? Have you seen anyone who's aced, let's say a behavioral interview or even a technical process could be a test or a technical interview, but ultimately got on the job and you realize they just couldn't do it. Either one of those aspects, either they couldn't deal with the people. They just couldn't do the job. If you notice any of those false positives, those regretted hires to any spring to mind at all without naming names, of course, Okay.

MARTA: knows what they're doing and be how much effort they put into the assessment, particularly if it's if it's not A standardized assessment, but one that is free form, and they need to provide some type of a creative solution to a data problem. But I have experienced it actually quite recently as well where there was an individual that did really well from a behavioral perspective and we see this more in senior roles. where someone knows all of the jargon, knows everything like in a let's say in pseudocode terms, they can speak through the problem. They understand how to solve the problem. They can instruct a team how to solve the problem. But this is where if you're hiring for someone who You need someone in a senior role who can still execute and this person may have more recently been in a senior role where they were managing a group of people and delegating the work and not doing the work themselves. And they're out of touch with, their hands off out of touch with the, the actual technical skills they need to implement. That is where I have seen problems where our requirement is someone more senior that can do both that can that understands the technical requirements, but can also execute upon them. So that is a challenge.

TIM: yeah. I can imagine if I just take myself, I could get pretty far in a software engineering application for any company, because I've been so close to engineers and build a product, work with them day to day, fixed hundreds of bugs myself. But I would be. A terrible engineer for any company. So I could talk enough and sound convincing enough to probably get past most levels. But if someone actually gave me a proper technical interview or test, I would rightly bomb out on it completely. And so I think it's, yeah, where you really need that proper measurement in place especially if this person's yet on the tools and leading a team, they need to be able to actually get in there and do stuff.

MARTA: Yeah agreed. I saw

TIM: cultural fit interviews? So I hear like a wide variety of views on these. Do you think it's yeah, hiring for culture is the most important thing. We need to get that right. It's got to be an essential part of the hiring process. Or do you view it as Oh, slippery slope. Like how can you have cultural fit and diversity? Aren't they oxymoronic what are your thoughts on cultural fit interviews? Silence.

MARTA: and while fitting in with culture is important it's equally important to hire a diverse group of people who will challenge the status quo. And actually, when I am hiring, I I assess how aligned a particular person is with our culture, and I try to select those individuals that Might be a degree or two out of our culture so that we can diversify our culture. I think that the more, the bigger a group of people that you have that conform ideally to your culture, the more it CR creates a clique. And and, nevermind talking about the group think element or the lack of kind of critical assessment or creative approaches to solving problems. But. You I think you end up having you. You risk having a group of mates working together versus a team of individuals that are working towards a common goal.

TIM: Yeah and I think about a lot of more traditional hiring processes. I feel like they It's just set up to go I'm going to hire people I like. It's like a likability contest. Which maybe if I thought about maybe this is fair enough. If you're a consultant and how likable you are directly dictates how much revenue you make for a company. Then I think, okay, fair enough. Or sales or something like that. But if you're an analyst inside a company, yeah, your likability matters a bit. You can't be a complete prick, but I think your skills and experience and what you're going to bring to the table is surely more relevant.

MARTA: Exactly. No I agree with that. And yeah I've actually seen people when I've been in a group interview I have, whereas my recommendation is that a candidate does not move forward, I have seen others recommend that they do move forward on the basis of a strong cultural fit. And so I, I revert then back to the trying to assess on objectivity and technical skills versus the cultural fit. Culture is definitely important but I think that it's a lot more valuable if you can get those unique perspectives.

TIM: Often it's been the sort of talent acquisition versus hiring manager view. Where as someone in talent acquisition, to be fair, you don't know anything about data science or software engineering really. And so the only thing you're able to evaluate in a candidate is like how friendly were they, did they seem to have the right values or what have you? Is that where you've noticed the bigger disparity that it'd be like a technical viewpoint versus more of a HR viewpoint. And that's where there'd be a difference.

MARTA: it wasn't HR, but it was a non tech like non technical team members versus technical team members. Yep. Yes.

TIM: that makes sense. What about pivoting a little bit to AI? Everyone's favorite topic these days for the past couple of years. If you had a crystal ball you'd be using it to bet on the stock market and make a billion dollars. I'm sure. But apart from that, if you can imagine what a data analyst is going to do in several years time what do you imagine? How do you think that role will evolve? Will there still be a data analyst or will what they'd be doing just be drastically different? Will they automate away the low hanging, annoying data cleansing, data wrangling fruit. And then it's just going to be this world of insights or yeah. How do you imagine that evolving? Silence.

MARTA: There are, businesses won't necessarily want to reveal. all of their business contacts to an A. I. And so it's going to be really important for the data analyst to understand the business, understand the context, understand the background and the problems and then be able to apply that analysis. Now that's incredibly important today, but in addition, you still need to be able to do the analysis, generate the charts, et cetera, et cetera. Whereas I think at the baseline, the baseline analysis is going to become more standardized more cookie cutter out of the box. Data analysts are going to, the ability to write basic code or produce basic graphs and charts is going to be less important. But what's going to become more important is to know how to do the analysis. And that's going to, that becomes I think the more technical the work, the more important that becomes. For example I can get ChapGPT to build me a bunch of predictive models or run a range of statistical tests. But if I don't know what model works best for this particular data set or what statistical test is the correct one to use, then I have, then I'm in this danger zone because I am potentially producing an analysis or an output that is theoretically completely incorrect. But if I do not have the theoretical baseline in order to make that assessment and, even worse if I'm handing the work off to someone who is non technical and cannot make that assessment, then you have a problem. And I think that is on the flip side, a massive challenge that is ahead of us where. We are at risk of just producing garbage analysis because the code works, the, it produces the model, but unless you know how to assess, it could produce a model that is completely incorrect. And that doesn't actually work, but but if you don't have the skills to assess that, then then you have a big problem.

TIM: Yeah, I feel like that is generally one of the main issues with the LLMs is to the untrained or unskilled eye, a first pass of anything that it's produced. It's Oh wow, that sounds pretty good. But if you're actually really expert or knowledgeable, You go that's wrong. That's misleading. They've missed this bit. That's hallucination. And so yeah, the thought of all these analysts running around pumping out and less models to change business decisions without really knowing anything about how they work or being able to validate it is scary. One thought I had, as you were describing this change, I was thinking maybe a data analyst might become marginally less technical. Then maybe the domain expertise would become even more relatively valuable. And so you might see more of those transition people who are like a business analyst or I don't know, like a marketer or something in their specific domain and they can almost transition across because that technical skill gap has been closed by AI. I wonder if that will happen a bit more.

MARTA: I hope it does because in hiring that, that is also a big challenge for us is we do data science in the field of B2B marketing and there are. It's a field that not necessarily every data scientist wants to be applying their skills to the business world. And so it's difficult to find candidates who have, as you said, that contextual knowledge. And and I agree with you that I think the most valuable data scientist is one that can apply these skills to a domain. Because data science is it's an applied science. So you need to, if you don't know how to apply it, then you're effectively completing university assignments. And then that's, in, in the business world, you're trying to solve real business problems and not just build something cool.

TIM: Yeah, I think having that being able to look at the data and know exactly how it originates, what it relates to, how it's being used. The customer generated or whatever, and being able to think through the almost business acumen side of it is often quite a big gap and you see a lot of candidates who are just viewing the data almost abstractly, they're looking at a table of users and bookings, but I don't think about, hang on, this is a user, this is a human. They went to this website, they logged in, they click this button, they booked a hotel. They went to the, they don't quite make the connection between the data and the reality. So maybe having people who come from like the opposite perspective, where maybe they don't understand the intricacies of the raw data initially, but they understand the business that might end up being a more valuable profile potentially the end of the day. I'm not sure.

MARTA: I think so. I think that there is a a gap in there's this hard core pure data science like data people. And then you've got business people. And so having more people that can bridge the two is a I think is incredibly valuable because you even you can do the world's most sophisticated and amazing data science related kind of analysis or project. But if you cannot one, explain it in non technical terms to your non data stakeholders, And if you cannot translate what you're seeing in the data to applicable business decisions, Then you haven't produced anything worthwhile because the recommendations need to be actionable. And oftentimes you cannot just take the raw output from the model or from the analysis and apply it. You need to apply business context. And that's where I think that those people that play that bridging or liaison role are incredibly valuable and I do think we need more of them.

TIM: Is that a role you yourself play, do you think? Thank you. Silence.

MARTA: And explaining, how it's relevant to a client and producing a deliverable that the client can use.

TIM: What about this? If you think back to all the candidates you've interviewed over the years, which they're probably quite a lot are there common patterns in where they failed where they've fallen down and why and, yeah, can you think of any ways then that candidates might improve on those typical Points of failure.

MARTA: the typical point of failure, particularly when it comes to assessments is the cookie cutter, bare minimum output. So what, one of the things that I really like about I'll describe it as a free form assessment where, everyone's asked to do the same thing, but they have the creative freedom to produce, to create. whatever output they feel is most applicable to solving the data and business problem it shows you how how much they've applied themselves to the application. And you just see a large number of candidates who will produce, with the basic libraries, produce the most basic charts without any form of formatting, without any attention to detail. And for me, a job application is. An opportunity to show a potential employer what you can do, what you're capable of. And so if I'm reviewing hundreds of applications and I'm just, I'm seeing just a lack of interest and the lack of effort that I would say is a big failure point where You know, they clearly showing that they're not really that interested. And if they put in that effort, it, it would make a massive difference. And it goes back to what we were speaking about the beginning about just this volume of applications that candidates are producing and potentially Employers are asking candidates to jump through more and more hoops and to do more and more work for every application and candidates are also getting tired of that because every application requires hours of input.

TIM: Yep. And they probably behind the scenes are almost like measuring the probability of them getting each job. If they feel like, Oh, wow, there's probably 20 other candidates not up against at this round. It's not, I haven't been told it's me versus one other person, which point they say, okay I can commit five hours to this then they're probably playing that numbers game in their head on yes, such a high volume of applicants they've put forward. Again, a situation where it could quickly spiral in a race to the bottom, I think, in terms of this kind of effort on both sides or lack thereof.

MARTA: exactly. I think we need to go back to a quality versus quantity game in the hiring process.

TIM: Yeah I agree, although I think it's about to be the polar opposites very aggressively with a higher volume of applicants applications generated through AI and other means, but we shall see Now, I was going to ask you about your experience on the other side of the table on, on getting roles and ask you about particularly negative experiences you've had as a candidate. But you said actually, I haven't really been through those for years. And you mentioned the last couple of roles that you got were not necessarily through a formal elongated process where you applied through a job ad and those kinds of things. So it'd be good to hear a few insights on how you've landed your last couple of roles and maybe. For more junior people in data, if they want to get to your level or wants to at that level, like what should they be thinking about in their job search? What can they learn from your approach to it?

MARTA: Yeah, so my last so my last couple of rules is the first came through a referral. So I had worked with someone previously who then referred me to the company that I was then working at. And so it was ultimately this relationship that I had. And given the individual who referred me was held in high regard at the company that they were working at. And my application was taken serious. I w it wasn't even an application. I was invited to a few conversations. And what was interesting. In that particular situation was that I wasn't looking. And so it was it was actually a very, comfortable process versus the typical job application process where you are looking, that's a lot more stressful. And so ultimately. What was nice about that process felt very, what felt like a privilege was that I felt equally that I was assessing them as much as they were assessing me. And I was making that assessment as to whether is it worth me leaving my current job where I've had a long tenure and things are going well what's in store on this next role for me that, that would be worth it to leave. And then my second role. The next one after that. It was actually I had left the workforce for a couple of years to go back to university. And so in the second situation, I was actually what you would call a boomerang where after completing my studies I was invited back to rejoin the company, by the co CEO who really wanted me to come back. And to use the skills that I had gained at university. They had at that stage, they had the experience where they understood what I could do before I went. And they were excited about me then applying my new skills to a new role at the company.

TIM: for formula E and what was interesting was he had a similar story to you where he said, Oh, isn't it weird how the more senior you get? Sometimes the simpler the hiring processes are. Even though the risk of the company is so much higher, you're paying the person a lot more. The total impact, positive or negative, on the organization is much higher. Yet, the hiring process is simpler. That's a curiosity. We're putting these grads through a ringer, eight interviews, blah, blah, blah. Yet, the person leading the team is like a coffee chatter and informal and blah, blah, blah. Which I found interesting and, I think slightly cynical, more junior candidates could go, Oh this is just, this is jobs for the boys, jobs for the mates, rub, it's all about who, not what, I think that's a terrible takeaway because I think what it really shows is that people like you have put in a lot of effort through a sustained period of time to do really good work. Such that someone would go out of their way to spend their human capital to refer you, which is like the biggest thing you could do is say, you should hire this person. Like you put your name at stake there, but that person doesn't work out. You're screwed. So I think the real takeaway is do really good work for years, build strong relationships for years. And eventually they will come to the fore, but it's maybe it's like going to the gym or eating healthily, like it's something that accumulates over time. And you might not see the benefit for it for years but it is definitely worth doing.

MARTA: percent agree. I think one of the most I think one influential books I read and by, very influential writer in, in the business space, I don't know if I'm going to say his name correctly, but Seth Godin wrote Linchpin, And I have for years, that is the that is my philosophy as I go into work every single day and I ask myself, what am I doing today that is going to make me that little bit more indispensable?

TIM: that's a great mindset. That's fantastic. Okay. I think we've probably got time for one more question. Do you have a preference out of the last two points actually that we have? So the question about the kind of data driven world or the hiring hero, do you have anything more to say

MARTA: Yeah. Yeah. I have very little to say about the last one.

TIM: Okay. All right. That's fine. Okay. So basically we're in this data driven world. Product analytics, marketing analytics, more and more data, AI, blah, blah, blah. I don't need to go on about it. Every data leader is like helping that business become more data driven, more data informed to make decisions using data because it makes it more efficient, makes it more productive, makes it more profitable, etc Yet, ironically, when it comes to hiring, even a lot of these. Very same data leaders would just go to hell with the data. I'm just going to wing it with my gut. Why do you think that is is there any reason for that? Is there any reason why we almost default when it comes to people decisions for our intuition normally? Like most people would do that rather than use the data. Do you have any, views over that?

MARTA: yeah, I was thinking about this and I think I think we overestimate our ability to assess candidates without data. We I think because it's related to people. We are on the side of thinking of well, this is softer skills related and we trust ourselves too much. I think we are, we might have overinflated egos when it comes to our ability to assess somebody for a job. And that's why we are on the side of gut feel and intuition versus using data. I think arrogance stands in the way of data here.

TIM: Yeah, for sure. I wonder whether almost ironically, but think about it now. Data might be the solution to show people how inferior their decisions are relative to using data. And so it's a lack thereof for some things that is telling. So as a quick example I feel like in hiring the kind of false positives are pretty memorable. If you hired someone and they didn't work out and you fired them, that's a pretty horrible experience for everyone. Maybe you don't have that many in your career and they stick out, but you, nobody ever measures the false negatives. If the 500 CVs that applied 490 of them never got a callback, but nobody's measuring off where I did one of these people end up founding a billion dollar company or go to this business. Like that data set exists, but it's not really measured at all anywhere in the world. So I wonder whether just actually measuring more stuff will then surface more insights that hopefully will make people realize, oh shit I was really wrong about this thing. Maybe there'll be like a reversion to rationality if they're just bombarded with the truth in their face. Yeah.

MARTA: Yeah, and I wonder whether when it comes to HR or PNC, however, what everyone's preferences for saying that whether it's the same challenge that we have in marketing. Whether, you're traditionally the cohort of individuals who take on those roles. are not technical and nor, I don't know what percentage of data people are interested in applying their data skills to that department or field. And so I think that there's potentially just not enough impetus or volume driving the data piece. In that department because it is traditionally non technical people that are running

TIM: Yeah, that would be a great LinkedIn study to do like just scrape all the HR talent leaders, look at their profile backgrounds, actually Chachapiti would be good at aggregating that saying which of these came from a software engineering or data science background. Yeah. I've never met or heard of anyone falling into that category in five years of working in the field, which is a great shame because if people are our most important assets. The biggest determinant of the success of the company, especially like an early stage one is hiring the right people yet. It's the thing that is least optimized for data. There must be so much low hanging fruit that could be done. If it just had a different mindset, someone come in and say actually like why aren't we measuring any of these things? Like you've been starting with the basic stuff. Surely would have a pretty quick ROI. I would guess for a typical company.

MARTA: Yeah, and maybe there's a business opportunity there where if appealing to the CFO, I, I don't, I haven't personally looked at any studies around the costs of hiring, but everyone knows how incredibly high they are. And Maybe appeal to use data to appeal to the people that care the most about money, which is going to be your CFO to try to get hiring and that department in general to be more data driven.