Episode 11

Galina Vakulina on the Complexities of Hiring: Trusting Intuition and Leveraging AI

TIM: So, Galina, when you think about hiring, what for you are the biggest challenges right now?

GALINA: I think it's a very common problem in every area in the world that it's very hard to find real talents. It's very hard to find because, market-wise, it obviously goes a little bit down, and people are trying to find a job, and if before specialists they were applying for a job that really fed them… Now they're trying to extend the area, and they apply to every job, and for recruiters, it's hell; for hiring managers, it's hell because we just don't understand, and also I think people who go to complete special courses, for example, courses to become data analysts or developers, three, four, or six-month courses I'm not against it. I'm completely for people who are trying to change the sphere and trying to be new professionals, but these companies, they just promise that, okay, you complete human scores, and then you become a senior developer. No, it doesn't work like that, and people have wrong expectations, and I meet a lot of people who come to the interview, and they say, I have completed a six-month course. Now I want to be a senior data analyst, and they don't have this set of skills and don't have the experience, and I think it's a big problem for now too that candidates sometimes have other expectations.

TIM: And have you noticed a pattern that the candidates who've done, let's say, the certificate and then have a sort of inflated expectation of their seniority are reasonably senior people from other professions that have kind of pivoted into software engineering, so they've already got quite a lot of work experience, or are they relatively junior candidates just out of university? yeah

GALINA: Yes, it's relatively junior candidates, and they were not related to this sphere, for example, people who were like salespeople before, or they used to work in the industrial sphere, and I'm not trying to say that it's your own profession; it's an amazing profession, but it's really like a big move, and you can become a senior if you were a senior in another profession in six months.

TIM: Yeah, it's not like that transitioning from I don't know a data analyst scientist, and they've already got some foundational crossover in skills, so it's a lot of people from completely different fields.

GALINA: Yes, yeah.

TIM: Expectations, and so part of that, then, is that's wrapped up in the kind of volume of applicants that you're receiving and the wide variety of quality, to put it nicely, that you're receiving because people are applying for lots of different roles.

GALINA: Yeah, are there any other big challenges at the moment for you in hiring? I think it's always a challenge to find the real talent because real talents are rare, and they're actually not eager to change jobs, and they know they've been hunted, and they are really demanding, and they live by trying to find something very comfortable for them, and all of us are looking for talents. We are looking for people who can contribute, who can change something in our sphere, so it's always a challenge, and now you just have to be very careful about it because it's very hard to find this person among all these candidates.

TIM: And I guess part of the challenge must be I would have thought especially in the early stages of hiring all you normally have is a CV, maybe some application form data, so it must be very hard to look at hundreds of CVs and go, This is the one we need to speak to, especially as I understand at the moment a lot of people are using ChatGPT to optimize the CV. And so now they're all looking the same. Is that what you're finding

GALINA: Yes, actually I wanted to talk about it because now what I see, and I think I definitely see the future that on one side it will be AI recruiters and on another side it will be AI doubles of the candidates, and they talk to each other. That's it; there is no human, and there are bots for CVs that will adopt the CV for the role that you're looking for. There are bots for writing cover letters. There are bots that help you to do mock interviews, like bots for everything AI for everything, but where is the human? And I think it's now a really big challenge for all the industry because, okay, I can tell everything, like I can go and I can fit my CV exactly to the role. and maybe I will lie a bit; maybe I will exaggerate a bit. Okay, I will go to the interview, and then what? And I think now we are on the brink of something new, a new era, and with this AI technology, it's like an infinite run of recruiters. They develop their AI systems to find talent, to find talents. and candidates, they use AI technologies to be like one step ahead of the systems, and it's always run, and I think we have to make a move towards human communication between two people, and as a hiring manager, I think all hiring managers know when you talk to a person. It's okay if this person doesn't have all the skills, but you almost always know if this person is good or bad for the company. Is this person fit for the company? It's a feeling, and I don't think we can teach AI to feel this connection between people right now.

TIM: What about this? So you know any AI system is built upon the quality of its underlying data, and now we're in a situation where a candidate is using AI presumably to generate their CV on the basis of the job description; therefore, a company is looking at all these CVs. They all look good, like they have a high score against the JD.

GALINA: And alike, and they look alike.

TIM: Right, they look alike; they sound alike; they've all got that chachapiti language, probably, but doesn't this almost make a mockery of the CV itself in a sense that, like, we've been relying on documents for so long that it was always a terrible representation of the candidate's skills and had so much bias? I feel like this is almost bringing to the fore and ridiculing how silly it is okay.

GALINA: to step out of the crowd to be like someone unique, and I think what we can do now at this point is like maybe to collect Like very honest data transfer data I don't know if candidates are willing to talk about themselves, but then it might be that we can develop something like an application that matches the right role to the right candidate, and it's not this hundreds of candidates to one role and candidates sending hundreds of resumes every day because they're desperate. They don't get answers from recruiters because recruiters are not under a thousand resumes, and I think it might be a decision when there is like one match, and it has to be based on real data on an honest resume, not all this chat GPT stuff. yes

TIM: I feel like with traditional hiring, a lot of the underlying problem is that companies are almost like taking the word of the candidate a lot of the time, like the CV; you basically have to take it at face value and believe at that point that it's true and say, Okay, well, this is really perfect. I interviewed this candidate, and then I also feel like a lot of kind of behavioral style interviews are like, Oh, tell me about a time you did this, or Have you demonstrated leadership? or whatever, and if the candidate is well prepared, they know their set of stories and how to tell their narrative, and so if you never really validate or measure anything, it's like almost doomed to fail, and I feel like that's being highlighted at the moment, and what we really need is some kind of actual measurement of the truth, whatever that is, whatever you're trying to hire. I feel like that should be the unlock in the process. What do you think? Okay.

GALINA: For every step, you can just make notes or be able to answer all the behavioral questions. I think it's not about this, and it has to be about the connection between people. How do you feel about this person? I think this is the most important thing.

TIM: Just thinking about it now, like all the problems we're talking about for By and large, like inbound applications, putting up job ads, and getting inundated with applicants of people you don't know, but maybe then referrals, connections, and networking would become more relevant, at least for the candidate side of things. Like if you're just fighting it out with 5,000 other chat GPT bots, you've got no hope; you probably feel, but now if you had a strong network, maybe the value of that's even higher than it used to be. What do you reckon? Okay, one of the things that hiring a hero should have is a network, a huge network of people in the sphere that he or she is hiring for. Basically, posting a job on LinkedIn now, I think it stopped working. It stops working because you just get thousands of applicants, and you don't know how to handle it. But when you do networking when you know people personally and When you get a new role, for example, to fill, you can Just recall, okay? I know this person. I know this person. I know this person. They will be perfect for this role. I will ask them if they're not ready; maybe they will refer me to someone else, and yes, this is how It works between people and I think Hiring hero should Understand this fear; you should understand. If The person is hiring, for example, by data science; they have to know the basics; they have to know what's going on in this sphere. They have to know people who contribute to this sphere, and then they have to know these people. Actually, I think it's better they have, like, their own databases of people they know. And they own maybe channels like news channels or something like that, and they can post there because it's Not as it's more like Professional thing not only LinkedIn or other job postings where everybody can apply to any role Yep, and so if you were hypothetically advising a CEO or founder of a company that was looking to hire their first data director or head of data, then you'd almost be recommending that this person come with a network; that should be part of the value that they bring to the table.

GALINA: Yes, I think for all professionals, especially for high-level managers, it's crucial to have a network because you'll get some people you can hire from this network; also you can go to this network to ask for advice. For example, I'm going to special chats to ask for advice about some issues that I'm facing during my everyday work. and some people, they already faced it, and they have answers, and this is how it works, and this is how you maintain this network, and then I can go to them and ask, Okay, guys, I have this person, and this person applies for a role in my company. What can you say about him or her? and it's very, actually, it's a very small world, and somebody knows them, and they can give you honest feedback about this person. it's very useful to

TIM: Thinking about it now, I wonder if there will be a product that comes along because the fundamental problem at the moment is that there's no good data available; every candidate's writing is there and see, but using ChatGPT is like a real issue. Maybe there'll be a product that comes along that somehow gets insights from the person's actual work, like their references, internally or directly into a task management system, and somehow aggregates or measures that, and that could somehow be used as a more real metric or something like that. I don't know

GALINA: honest data, and I actually have no idea now how we can get this data, maybe really from these network systems like chat groups.

TIM: Yep.

GALINA: applications whatever

TIM: And you've built pretty substantial teams in your time. Did you find that hiring people you knew tapped out at a certain point? Like, did that get you the first five people, but it's like there's only so many people you know; you couldn't fill a team of a hundred, probably, I'm assuming, with your networks, or was it then getting, like, your first team to then get who they knew and sort of a compounding effect? yeah

GALINA: Yes, usually it works like that, and when you manage to hire a person who is well known in this sphere, it's great luck because people will try to go and work with this person. It's very important, I think, in the IT sphere. It's very important because people like working with these famous people in the sphere because they can learn something new; they can improve themselves. And I think when you manage to hire the core team of really cool professionals, then it must be very easy.

TIM: Yep, you were mentioning before about hiring Hero. What about hiring anti-heroes? What kind of methods would they use for hiring, do you think? The first thing that hiring an antihero has is Not answering is ghosting, and I think nowadays It's It's an issue and I read it on LinkedIn every day that People They suffer. From this And it frustrates a lot. And I think It's not even polite not to answer after the interview. and also Okay.

GALINA: You will be able to reuse this connection for the next rule, but you lose this connection because the person will never talk to you again after you ghosted this person. I think this is the first thing also for me. A hiring antihero is the person who makes five, six, or seven steps of an interview. When in company, there is a process of many steps of the interview. I think it's a nightmare for the company and for candidates too, and actually I read a story of a guy just recently. a genius guy, a very good developer, but he had to pass six different interviews, and they were asking almost the same on each step of the interview. and by the sixth step the guy said, No, guys, you know what? I just refuse to go to your company. I just—I don't want to. I'm done. I'm exhausted. I will go; maybe I will get a lower salary, but I'm not ready to work for your company. This is what we get because really talented candidates, they're not ready to go through this. and people who are really desperate to find the job maybe they will be ready to go through five or six, but who do we want to hire?

TIM: Yeah, exactly. I think in terms of the blown-out hiring processes, I wonder whether part of the root cause is again a lack of data being collected because if you're a big enough company, you should have enough information on interviews to know the value add of each marginal interview, and by the time you're getting to the sixth interview asking the same questions, what is the point? like almost surely no value in how accurate your decision is. I guess for smaller companies they might not have enough data around that, but they could probably look to, like, I'm sure Google has published research that said having more than four interviews adds nothing to their hiring decision, looking back over, like, a million hires or something, some ridiculous data sets. So even smaller companies could maybe lean on what the big tech firms have done through time.

GALINA: Yes, and also I think this takes a lot of time from the company from people who make decisions on every step, and it's like just feeding their self-confidence; people just feel important because they're on the interview, and okay, but you can go and do something real; you can go and develop a new product, or you can go and, I don't know, build another team. but you are being on this interview day by day making decisions, and I don't think it's a very efficient way of hiring people.

TIM: It's not, and in my experience working in recruitment, the longer the hiring process, the less likely any candidate will be selected because inevitably someone will find some reason why they don't like them, and most companies have this almost like one strike and you're out policy, and it will normally be something that's very subjective. Or some kind of criteria that the interviewer is looking for that is irrelevant Like, oh, I didn't like their whatever tone of voice or something bullshit like that, so it went downhill quickly. I don't know if you found something similar.

GALINA: I don't know because we did a very quick hiring process, and we established only two steps of the hiring process, so it's much quicker, and people who make decisions sync very well between each other, so they don't need to even discuss for a long time what is like or dislike in the candidate. Then they know what they're looking for.

TIM: That's a great segue to the hiring process, and so only two steps—that is probably the shortest I've ever heard. So what are those two steps, and why do you have those two steps?

GALINA: Okay, we are hiring data analysts, right? So the first step is a quick screening of HR, of course, to briefly understand if the person is good or bad for the role. They ask them some questions, and if they think that yes, it's all right, they send a short homework Actually, for now it's a bit tricky because now we see that our dear friend Chad GPT is doing all the jobs, and it's very obvious when you see it, and usually it's like similar answers, so we're trying to make it more sophisticated, but the candidate doesn't have to spend days to accomplish this; maybe one or two hours. I think it's enough, and then if you like the homework, there is a team of senior-level data analysts who are also in charge of the community, so they're in charge of the processes of hiring, grading, and onboarding people, so they do all this job, and for now I think they do it brilliantly, so they have an interview with the candidate, and they check hard skills and soft skills as well. They ask questions not about SQL and other instruments; they ask questions about logic about different situations in previous work and so on, and then that's it. They make a decision, but the thing is, why it's so easy is because we hire to the company; we hire to the community; we don't hire to the particular team. So they make a decision that this candidate fits all our requirements for data analysts to work in the company, and then we do a short onboarding, like two weeks, when the candidate works hand in hand, like shoulder to shoulder, with these senior data analysts, and we introduce this person to our database, to our BI instruments, to all the infrastructure. And then there is the most interesting part, so we always have, like, several teams that are waiting for data analysts, so they spend some time in two or three different teams, one by one; usually it takes two weeks for each team, so they go, and they already are able to do, like, real tasks to contribute to the business, and then they decide which team they like more. And if the team likes this person too, it's a match; they work together. If no, then no, and it happens sometimes that none of the teams—they didn't like the candidate—it's okay, so we say goodbye, and it's easier to say goodbye in this stage, and during this period, our people, like our senior data analysts, keep an eye on this person. They check if all the skills are right because sometimes a person is able to pass all the interview steps, and then during this period of time of onboarding and spending some time in the teams, we were surprised to find out that this person has no idea what SQL is. I don't know how it happened, but actually it's another step in how to check candidate skills, so this is how it works, and actually I think it works pretty well, better than a classical process.

TIM: So there are a couple of interesting aspects there. One is the pooling, just hiring a general set of analysts, and then eventually they end up embedded in a team, which is quite similar to the way, well, I say there's that aspect, then there's the rotation among the teams, which is quite similar to the way graduate and internship programs work in some kind of way. Would you do this for any seniority? Like, would a senior data analyst also go through this rotation?

GALINA: We are doing it up to the middle class for seniors. For now, we're not able to do this because for seniors there are a lot of different requirements from different teams, and they have their own specific—one of them should have, should know a bit about testing, like on a brilliant level; another should know something about, let's say, compliance on a brilliant level. So, there's a more special thing that these people should know: for seniors, we don't do it right now; maybe later, let's see, but for now, up to middle class, yeah.

TIM: Okay, so that makes sense. There are more junior roles across the domains that are more similar. there's enough of a crossover in the skills that it's easy to hire them all through one process as opposed to the more senior ones, which might have less crossover, and if you tried to hire them all into one pool, it wouldn't really work because you'd have to expect to have some unicorn basically you could somehow do any of the jobs

GALINA: Yeah, but usually we have the same requirements for data analysts who work in the company, and the candidates should have almost everything: SQL and BI instruments, A/B testing, product management, simple logic, mathematics, and an introduction to data science. Like, everything should be like a soldier that can do everything. and then we can rotate them between teams and find the best match.

TIM: That's a really clever way of doing it because then you kind of build in the standardization and the skill level right across the organization. You don't have one pocket of analysts who are amazing and the other one that's maybe not as strong; you kind of keep a really high bar. What about if you mentioned that after they do this kind of rotation, then they might get matched with the team if you know both sides like it? So what if no team likes them after two weeks? What happens? Is that like their probation period or like a trial period? How does that work?

GALINA: They are waiting for a data analyst. Their struggle is our data analyst, so if they didn't like the person, if they said no, we are not ready to work with the person. There are very strong reasons, and if none of the teams, if all teams said no, we're not ready to work with the person. It's a strong reason to say goodbye to this person and to let this person go, so yes, you might just… It's like it's a probation period; it's less than two months usually, so yes, it's better to let this person go because we tried, and we didn't maintain the connection, and it's better for us to go different ways.

TIM: Yeah, better for the candidate, I mean, surely.

GALINA: Yeah, yes, yes, yes.

TIM: What about if you imagine or think about all the different candidates you've interviewed over the years? Are there any recurring patterns that differentiate the successful ones from the unsuccessful ones?

GALINA: I was thinking about it recently, and I can tell you about the pattern of unsuccessful ones. It's very funny, actually, and for now, people who are now doing all these interviews, they come in with the same stories as candidates when they're asked questions in the interview. They might say, No, this question is incorrect. This task is incorrect. Why are you asking me this question? I'm not a for my role; I don't need to answer this question, so people probably think they're not confident or they don't know the answer to the question, but they blame interviewers that the question is just incorrect, so you gave me wrong I don't know database you gave me You can't do this in SQL; you can't solve this problem using this. I don't know instruments, so they're trying to teach us how to ask questions and how to give them tasks, and this is a strong pattern of unsuccessful candidates because it's much better when you say, Sorry, I don't know this; I can't answer your question, but I am eager to to learn I'm eager to learn, and if you give me some time, I will be able to answer your question correctly, and if the candidate is honest and is ready to say, Okay, I'm not; it's not my forte; this part is not my forte, I'm sorry, but I'm really good at this. This is a way to success, I think.

TIM: And is that kind of really visceral rejection and quite egotistical answer common? Like, this happens often, you're saying, among unsuccessful candidates, I mean.

GALINA: Among unsuccessful candidates, not at all. Not all of them are doing this, but I think it's a very interesting part of these candidates that are doing this, and I think, I don't know, once in a while, yes, I hear it, like once in a couple of weeks I'm hearing about the candidate who did

TIM: Yeah, that's a bit of an own goal. If you don't answer the question that you've been asked, isn't anything else springing to mind in like Differentiating the unsuccessful from successful candidates

GALINA: Successful candidates, they always can tell you about the results. They can tell you about what exactly they did. Maybe it might be a small project; it might be a small task, but they can tell you, Okay, the goal was this; I did that, and the result was this, and this probably the result was another, and the goal wasn't achieved. Okay, but you can tell the story okay, and for unsuccessful candidates, it's more like, generally, I was in charge of the team and something like that. You can't understand what exactly this person did, and you can't understand what goals this person achieved, and it's also very obvious, and one more thing. It's not about the person; it's not about the hard skills of the person, but usually when you talk to the person, you already feel how he or she fits the values of the company, and it's not good or bad if the person doesn't fit the values, but it's very important too. And, for example, if you're not ready to work in an organization without a hierarchy, let's say without classical leadership and management, and you're going to the company, for example, that declined to work in a classical hierarchy, then I don't think it will be a match, so I think digging a little bit more into the values of the company is very important when you go to and you prepare for the interview.

TIM: Again, a lot of hiring could be improved if the candidate and the company were a bit more transparent and honest and just really direct and upfront, like, This is what we want; this is what we expect; this is what I want; this is what I expect, and if we can kind of sort that out as soon as possible, you can almost disqualify each other, and it's fine. It's not like being rejected; it's the right

GALINA: Yeah, I agree.

TIM: What about AI? So I've talked about that a little bit, how it's already changed hiring in terms of the generation of CVS and maybe automatically applying and then now the automated CV screening and those kinds of things. There are also some AI interview products out there where you talk with a robot rather than a human, which must be an interesting experience. I would have thought for candidates, but how can you see hiring being further changed over the next few years? Like, what do you suspect is going to happen with AI?

GALINA: For now, as I already said, for me it looks like a mess because AI, I think, now it makes things more complicated than before because of the amount of different applicants that really can fit their CV to any role, and there are special technologists that will find a new job posting; they will fit your CV to this job, and they will send your CV to this job. You don't do anything, and for now I think it's a mess, and for me, using data and using AI, I think it will be more useful when we talk not about hard skills but when we also talk about soft skills and the profile of a person, and maybe using AI you can check the profile of a person to understand if he or she fits the values of the company. and also I was thinking that maybe for the big companies and for the hiring agencies it might be good to get more statistics and to understand why they did not succeed with the candidate, why they were successful as a candidate; for example, if they can get some data about how long the candidate worked for the company, what the person's success rate was, successful, or this person was not efficient, and if we gather all this data, maybe we can do some conclusions about that. Maybe we can change the process; maybe we can change the profile of the candidates a little bit, so I'm thinking more in this direction because all these instruments that help for candidates and have recruiters right now, I think it's just an endless challenge and endless run.

TIM: Yeah, it's really hard to see how it's going to play out. I'm trying to reframe and think a bit more positively, so candidates are applying en masse with TVs that are AI-supported or optimized or whatever. Companies are now going to screen those with probably the same AI tool. At least I feel like that step is an improvement on the current step because at the moment it's a human reading a CV. If you're lucky, for a second, not at all, to be honest, and then a human brings all that bias a CV has, like a person's name, it says the gender, their ethnicity, maybe it says their school, how old they are—all these kinds of things. In theory, I should be able to do that almost costlessly and in a very consistent way; in theory, it could be free from bias in a way that human CV screening could never be. So maybe that could be at least an improvement.

GALINA: Yes, I totally agree with this. I totally agree, but on the other hand, let's say some candidates use AI to write a cover letter, and some candidates really put themselves into it; they put their souls into writing the cover letter; they maybe make small jokes; maybe they try to connect personally with the hiring manager, and they put in some effort, and if all of this is checked by AI, I wouldn't understand it. It won't distinguish, so I think sometimes it's really important for real people to read the person.

TIM: Yeah, it's interesting. I'm thinking of a very similar scenario, actually, at the moment in sales, like in doing outreach to companies by email via LinkedIn. What we found personally is we basically put 100 percent of our effort into making sure the person, when they read the email, thought, Oh, this is actually written by a human. That was our one goal, and so that's very difficult to do.

GALINA: Yeah. yes

TIM: quickly, but I wonder if candidates would have to employ the same thing, like if they just called the person and had a conversation with them. That's probably a human; you know, if they took, like, a snapshot of their LinkedIn profile with their face next to it and emailed it to them, that's probably real at the moment, but given how quickly AI is developing, who knows before it's indistinguishable from human communication, which is pretty terrifying.

GALINA: Yes, it is, but I hope we will find a nice solution for Refit and Industry 2 so it won't be almost like it is now.

TIM: I have a feeling it might get worse and then get better; that's my sensation at the moment, but we shall see. What about on the other side of the table as a candidate? Do you have any especially memorable experiences trying to get a job, either positive or negative?

GALINA: No, I don't think I can recall any because I used to work for a very long time for one company, so

TIM: What a great answer! I mean, you must be happy then, and you've avoided the misery of so many candidates going through awful hiring processes.

GALINA: Yes, yes, but I think it's like, on one hand, yes, but on the other hand, it's a good experience, and it's a good experience to connect to other people and to talk to other people about your achievements, and it's also a part of your job and a part of your career.

TIM: Yeah, actually now that you frame it that way, I remember like the very first interviews I had when I was a teenager for part-time jobs, and it was very confronting just talking about myself in positive terms to a—like, that was just immediately like I'm not meant to be doing this kind of thing. And so that takes a lot of learning. I think it's definitely a skill, isn't it?

GALINA: Yeah, yeah, I think imposter syndrome is very common these days, and I cannot imagine myself saying all these great things about myself, and usually when people ask me about what I do, I'm like, But I do my job. And sometimes you just don't see the big picture of your achievements. And sometimes I think you have to stop; you have to look back and say, Okay, I did this and this, and it's really great, and it makes you feel like you have more self-esteem.

TIM: Yeah, for sure. I feel like it's when you're in the weeds each day and you just go day to day to day, task to task to task, that you wouldn't notice any growth because it's like a little percent each day maybe, but then you look back and you're like, Wow, that's a big difference, like compare my skills now to 10 years ago. it's almost like a different

GALINA: Yeah. Yes, that's true. That's true, yeah.

TIM: What about hiring? So this podcast is all about objective hiring and trying to think about ways to make it a bit fairer. Do you have any thoughts on how hiring can be fairer? You mentioned having a simple and short hiring process, like not making candidates go through the hiring process, you know, weeks and weeks of bullshit interviews. any other ways you feel like hiring could be made a bit more objective so the best candidate gets the job as often as possible

GALINA: It sounds like a fairy tale to be honest for me, and actually I don't have a good answer to this question to be honest because it's always very personal, and it depends on the personality of a hiring manager, right? And okay, we can use different AI tools. We can use different automated tools to reduce this bias, but at the end of the day, the person will make a decision, and I don't know, maybe this person prefers different sides of characters of other people; we never know, but if we exclude a human being from the process, I don't think it might be better. So it's always a bit biased, and I think it's very important for the hiring manager to understand this responsibility very well and to maintain this ability to think about how fair the decision is.

TIM: What about reflecting on your hiring career and all the people you've hired? Do any big fails spring to mind, like any particular person you hired, of course not naming names, or anywhere you used to hire, or even a colleague in the way they hired? Like, can you remember any striking moment of, Oh God, we've made a mistake here?

GALINA: I can tell you about the main mistake that I usually made with unsuccessful candidates: I didn't trust my gut, so when, for example, there is a situation, you see the candidate, and this candidate has a brilliant set of skills, very good heart skills, and a proven track of different achievements. and he or she talks very confidently, and this person is really a professional, but something inside you is just saying maybe not, and you're in a hurry, and you have a team that is waiting for the candidate, or you have a product that is struggling without this person, and you go to the compromise, and you say, Okay, let's do it. Let's give it a chance, and it never works. I don't know; I can't recall any situation when it worked, and I don't know; it's just something in between people; you just feel it's not a good It's not the right person. Maybe it's a good person, of course, but it's not the right person for you. It sits in a different shape, and it just can't fit the shape of your company. and usually it doesn't end well. Usually this person is getting fired, or the person just goes to another company in a couple of months, and you have to start this process again, and actually when I see a person who is lacking hard skills, you feel that you feel this connection. It's much better when you see a great professional. But you feel like it won't work, so I think not trusting my gut was my biggest mistake all the time.

TIM: I'm interested in this idea of gut feel and intuition maybe versus data in hiring, and so if you think about some of these candidates who you were like, Oh, probably not, but we hired them anyway, and they worked out, Was it for reasons that weren't part of the evaluation criteria? So let's say you're looking for, I don't know, the candidate needs to have these technical skills and these soft skills. Was it something beyond that, even that just put you off that wasn't related to necessarily a communication or something you're measuring? It was just something else. Or was it actually something that could have been measured in the hiring process if you'd kind of brought it to the forefront if you see what I mean?

GALINA: Yeah, but I think it's what we talked about before: if you prepare for the interview, you can just answer all these questions, even for a behavioral interview, about your fails and successes and about different situations, and I think it's not very hard to prepare for a behavioral interview and then show yourself as a… But sometimes, yes, it's something like in the way of how a person is talking to you and in the way how you like how you feel about a person being honest or not. and I think, like for hiring managers, it's a very common thing when you've been through this process for a long time, this feeling, this feeling when the person is honest. It's super cool. Yes, I want to hire this person after five minutes of the interview. and sometimes you're like, Yes, everything's so good, but I just… It's a feeling, and I don't know how to put this feeling into data; this is the problem I think.

TIM: So that's really interesting. Okay, so every time basically that you didn't trust your gut, it failed. What about—I'm really interested in the opposite—were there any candidates that you can remember who you trusted your gut on and rejected but then you heard Oh wow, so and so started this amazing company, or they've become really successful, or like any sense of the false negatives in your prediction

GALINA: No, I don't recall, but I remember one situation when I trotted out my gut, and there was a junior analyst, and all my many people on my team were against it. They said, No, it's not what we're looking for." I said, No, I like this candidate, and I want to try, and it worked out. and I remember now, and sometimes I talk to people from my previous team, and they said, Do you remember there was a situation? And look now, and they're like, Yes, yes, we remember, and we already felt like a bit sorry that they were against this candidate, who is now an amazing specialist and person who still works for the company.

TIM: That's amazing, so you've got a bit of a sixth sense for really spotting the real hidden talent.

GALINA: I think it's a It's a thing that every hiring manager should have. It's not only about trying to match the job posting and CV; it's not about trying to get the right answers to the questions in the behavioral interview; it's not about this formal thing. It's about understanding what's behind this face of the person, what's their real motive, what are they about, what do they like to do, and yes, it's a little bit about a personal thing, but we will have to work with personalities; we will have to work with real people; we will have to maintain connections with them. and I think it's important to just touch this area a bit.

TIM: Yeah, well, AI certainly won't be touching that anytime soon, I don't think, because that is a very human-based assessment, isn't it?

GALINA: Yeah.