Jonas Christensen 3:28
Tim Freestone, welcome to Leaders of Analytics. It is wonderful to have you on the show.
Tim Freestone 3:36
Thank you for having me, Jonas.
Jonas Christensen 3:38
Yeah, this is going to be a really interesting episode and something that's very close to my heart, we're going to be talking about hiring the right data analytics candidates. And I think this show is equally important to those hiring, as it is for those applying for jobs. Both categories of listeners will have lots to learn from this episode. And Tim, for me, this is something that I find to be the most important part of my job, which is hiring the right people for the team because without a good team, we will not succeed. So this is what we're going to talk about today. But before we get to the detail of that, we would like to know a little bit more about you. So, in your own words, could you tell us a bit about yourself, your career background and what you do?
Tim Freestone 4:22
Yeah, for sure. I studied business down at Wollongong. I grew up in Wollongong. I'm a Gong boy. And did finance and economics and my path was set to do the classic finance graduate thing. Get into investment banking, you know, that kind of thing. For various reasons, including the GFC, that didn't quite work out. And I fell into my first analytics role as a financial analyst working for a business on a joint venture. So it's doing your balance sheet, PnL, cash flow projections, all those kind of classic financial analysis things. And progressively kind of got closer and closer to analytics. So moved into almost like sales analysis, then commercial analytics. And then by the time I joined the tech company, just by virtue of the type of data that they had and the types of problems they're solving, I found myself basically in data analytics and bordering on data science. And yeah, joining that tech company about five or six years ago, I learned a lot about how to do things at scale. Moving out of the world of spreadsheets. You know, where you got millions of data points a day, Excel doesn't really cut it. And yeah, from there, I learned so much more about analytics and a lot about hiring as well.
Jonas Christensen 5:31
So what do you think attracted you to this world of data science and analytics in the first place?
Tim Freestone 5:36
Good question. So, I definitely didn't have designs on it. It was just me falling into it. I think it's the rational side of it. Like, the search for truth, and objectivity. I think that's definitely something that's always resonated with me. Like, I hate bullshit. I hate people are obviously just talking nonsense. And I'm always drawn towards what is the actual truth of the matter. And I think analytics can certainly help you arrive at the truth.
Jonas Christensen 6:03
That is definitely what we're trying to do in the world of data and analytics. And actually, I should add one thing, because we have listeners from all over the world on this podcast. Wollongong is a university town, just south of Sydney in Australia. You probably haven't heard that name unless you're from Australia, but it's a lovely place.
Tim Freestone 6:20
It is.
Jonas Christensen 6:20
It is a good place, good wholesome place to grow up, Tim.
Tim Freestone 6:24
Yeah, for sure.
Jonas Christensen 6:26
Now, Tim, let's dig a little bit deeper into today's topic, which is about hiring the right data analysts candidates. The reason I have you on the show, because you're a bit of an expert in this topic, and you actually founded a company called Alooba, in 2019, which tries to optimise this whole process. So could you tell us what Alooba does and why you started the company?
Tim Freestone 6:48
Yeah, so Alooba does a few different things now. So we've got a few different products. We have a product, which we call ''Alooba Assess'', which companies use as part of their hiring process to assess the skills of their candidates. These are typically in analytics roles, like data analysis, data science, data engineering, product analytics, these kinds of areas. And customers use that product in two different places normally, so one is as a top of the funnel initial screening quiz they will share out with every candidate who applies. And this is basically, you can imagine it's like a substitute for manual CV screening. So instead of having to pull through hundreds of resumes, send everyone a short quiz link. And on the basis of their performance in that customised skills quiz, decide who to interview. The other position in the hiring funnel, where it's used as a replacement for the take home assignment that you might give candidates. So a lot of companies have their own data, set their own problems, they'll give candidates to take home and do some kind of unstructured problem. Might take them a week. And some businesses use Alooba as like a lighter weight version of that, that takes less time for the candidate and kind of speeds up the hiring process. So that's what we call ''Alooba Assess''. We also have a product called, 'Alooba Growth'', which companies use to understand the capabilities of their teams and people. And more often than not, this is in the kind of broader basic data literacy skills that you could argue almost anyone in any role now needs in business. So not necessarily just data specialists, but you know, marketers and product managers and accountants and general managers. And we've really found a nice sweet spot there in helping businesses basically baseline their data literacy. And then they connect the outcome of these tests to their learning and development plan to then put in place almost like a data informed, L&D plan for data literacy.
Jonas Christensen 8:33
That last part is really interesting, because that is something that a lot of organisations struggle to understand. How do they even get started on that approach? Because they might know that they need to upgrade the literacy of data and analytics in their organisation, just like we did with computers, maybe 30 or 40 years ago. But the starting point is so hard. Could you tell us a little bit about how you've worked with organisations to do that and typically what they get out of it and what initiatives they take subsequent to your assessments?
Tim Freestone 9:02
Yeah, absolutely. So, basically, in an ideal scenario for the businesses that have engaged with us on like Day Zero, they're just contemplating, ''What are we going to do with data literacy?''. We know is this thing that we probably need more of. We know we need to do something and haven't quite figured out what or how or why. So that's the ideal scenario, because then we can help them basically put together a whole series of different assessments and help them map out like what an expected level of data literacy would be for different types of roles. They can then go and assess their entire business, their entire function, whatever it is, collect that data, identify those skills gaps in either an individual or a unit or a geography or whatever. And then from those insights, they can then put together the learning and development plan to actually fix those gaps. So they might identify for example, you know, in general, their teams already have great visualisation skills. Like, they've been using, I don't know, Power BI for years and they're just great at putting together those visualisations, but maybe they lacked some foundational statistics/knowledge, so they're likely to make maybe poor decisions when they're dealing with insights. So by collecting these strengths and weaknesses, they can put in place a learning and development programme that's been customised based on those insights, as opposed to just, ''Hey, let's train everyone in X,Y,Z tool''.
Jonas Christensen 10:16
I can really see that taking off more generally, in the next 5 - 10 years, as organisations start to realise just how important it is to just have an analytics team that can produce output for the business, but also this ability for the business to consume that output, because you really need both for that output to be a success in the organisation. Now, Tim, back to hiring data and analytics candidates. I'm interested, in your view, perhaps with a bit of research behind it if you have so. What are the biggest challenges for hiring managers in this data analytics industry and how do we solve them?
Tim Freestone 10:53
Yeah, so for hiring managers, I think right now they'd probably be complaining about a few things and these are their biggest pain points. So, for some companies, they're struggling to attract candidates. So, the ''War For Talent'', the this, the that. You've heard it called so many different things. Businesses are struggling to find the right applicants for the roles, they say. And then the other piece is really around having had a set of applicants figuring out easily who the right people are to interview and to hire. That then causes a slowdown in hiring, so longer time to hire and a more expensive hiring process. Now, for me, I think, actually, a root cause of this problem is the way that some more traditional businesses do sourcing and recruiting. If you think of the world before COVID, it would be normal that you would restrict yourself to hire someone who's within 20 kilometres of your office and everyone comes into an office every day and that's your talent pool. The world has changed and I've done some research on LinkedIn 99.998% of all data analytics candidates in the world don't live within 20 kilometres of the city of Sydney. So the world is really your oyster and if you still think it's 2019, when it comes to sourcing, it isn't. So I'd say firstly, it's like defining the geography of who you're willing to recruit and from where, can make a dramatic impact on that key pain point of, ''Hey, I can't get any applicants applying for my jobs''. So I'd say you're rethinking that a little bit. And then also rethinking the traditional screening or filtering process for analytics and data science is a big issue as well. Almost every company we've ever spoken to, starts hiring with a manual CV screen. So someone's gonna go through, look at a CV for maybe 10 or 20 seconds and try to figure out on the basis of that: Is this the right person for the job? And we've got some pretty compelling evidence to show that that's an incredibly inaccurate way to screen candidates. It's quite likely that traditional hiring methods are just filtering out really good candidates who would otherwise be a great person for the role.
Jonas Christensen 12:55
Talk to us about that. That's really interesting. So, you have seen that that is actually an inefficient way, yet it is the predominant way. When I've hired people, I typically hear from my internal talent acquisition team in various organisations, so this is the standard. Typically that we have 100 applicants, ''Here are the 20 best people that you can look at'' and of those, I might pick five that I want to interview. So there's 95 that don't get a chance.
Tim Freestone 13:20
Yep.
Jonas Christensen 13:20
Talk to us about that scenario, and what we're missing out on.
Tim Freestone 13:23
Yeah, so I think the CV itself, it's just a very weak data set for a few reasons. So one is it's someone's own perception of their own skills. Interestingly, for Alooba, every time someone takes a test on Alooba we collect not only obviously their results, but also their self rating. So, they rate themselves on a scale of 1 - 10. We compare this to their actual performance. We come up with what we call a ''Self Awareness Index''. And we know from like tens and thousands of data points now, people are not the best judges of their own abilities and skills. We don't all have perfect self awareness. So, let's say I write my CV and I say I have advanced SQL skills, what does that mean? Like, what is advanced? How have I defined that? How have I compare myself to others? So, Yeah, just the whole self perception element. If you think about it also, all the endless articles online about how to improve your CV, how to make it more attractive, how to make it stand out, that almost implies the process is wrong anyway. Because if you can just literally change your CV and then get a higher clickthrough rate or higher callback rate, there must be something wrong. Because you haven't changed, your skills haven't changed, your abilities haven't changed. So that's an issue. And then there's the whole area of bias or if we were being more harsh, let's just say straight up discrimination. CV basically has all the data points you would want to collect on one piece of paper if you're going to discriminate against someone. You've got their first name, their last name, which gives you their gender and their ethnicity. You've got where they went to school and it tells you their age, potentially their religion. In lots of countries, people put a photo on their resumes. So then it gives you that bias as well. In a lot of countries, people put their religion, their marital status. Like, it's just a treasure trove of all the things that are irrelevant to figuring out ''Is this the best person for the job?'', yet could also be used for discrimination. So I think it's quite a flawed system that we have the starting point of the hiring process being someone looking at a document containing all these irrelevant pieces of noise for 10 seconds and figuring out, ''Is this the best person for the job?''. We shouldn't be surprised that that system is flawed and doesn't work very well.
Jonas Christensen 15:36
You know what? It's very obvious when you put it like that. But at the same time, this is the document that we all cherish and we all put a lot of effort into putting together and we add all these seemingly irrelevant items on there to stand up. You know, your sporting affiliations or extracurricular activities and so on. '
Tim Freestone 15:55
Yeah.
Jonas Christensen 15:56
So do you think that the process should be flipped around in a sense that we should do something upfront before we even see the CV? Is that kind of where you're going?
Tim Freestone 16:04
Yeah, I should mention actually, as a candidate, I would be optimising my process for the understanding that my CV will be looked at for 10 to 20 seconds. And it is going to be looked at by a human being. A human being who has all these biases, has all these preconceptions. So if, for example, you thought, an interesting hobby is going to stick out in the eye of someone and it's going to give you a better chance to get interview, I would include that. That's unfortunately, the rules of the game we have at the moment. But there's surely got to be a better way that doesn't involve this data set being used at such a critical point in the hiring process. So what we do for some of our customers, particularly those who run a SAS or tech company, often they attract a very high volume of candidates, they will send every candidate a short quiz. A short skills assessment quiz customised for the role, and then on the basis of the performance in that quiz, decide who to interview. I think that's definitely a more objective and more accurate way. It does come at the cost of candidates to be fair. Like, that's a bigger commitment upfront for candidates than just sending off a CV. So, that is a pain point that we're gonna have to solve eventually for candidates as well. But yeah, any way you can eliminate those potential for biases early on in the process, the less biased your process is going to be, which to me is actually another way of saying the more accurate it's going to be. Like, our biases are the things that are preventing us from doing the most rational, best thing possible. So we can remove those and just have this anonymized objective measure of someone's skills, then we are one step ahead. I think.
Jonas Christensen 17:35
Nice. A few thoughts come to my head there. Well first of all, I should say that if anyone's ever thinking of applying for a job at Alooba, make sure you put in the first sentence that you're a big fan of Glasgow Rangers, because that will definitely bias your application in a positive manner.
Tim Freestone 17:54
I'm a human like everyone else.
Jonas Christensen 17:58
Second to that, I've used the Louis product before and I thought it was a really helpful data point that I could not get from any of the traditional sources that I otherwise use, including resumes. The thing I also found challenging is exactly what you're pointing to there, which is I had 50 - 100 candidates. How can I fairly put them through? It's fairly tough assessment before they even know whether I want to speak to them in the first place. And surely there needs to be some pre-vetting of that before we get to an assessment. So that's an interesting challenge. Have you got thoughts on how you solve for that?
Tim Freestone 18:34
Yeah, for sure. So what we're trying to do is really help our customers be a lot more transparent and give a lot more information upfront to candidates to de-risk it for candidates. So I think a big part of the issue is that when you apply for a role as a candidate, basically all you have is a job ad with some fairly generic information around the requirements, some information about the company. You could go and do some research, look at their blog, those kinds of things. But otherwise, as a candidate, you almost never know who's in the team. What does the team do? What are the objectives? What are my KPIs going to be? What's my manager like? So all those kinds of internal things. You also don't know what's the trajectory of the company. If it's a private business. You don't know this kind of wider things, like what are the core values? You might not know, that haven't been published on the website. You don't know about the hiring process. Like, what are the different steps? How long are they going to take? Why are these steps being included? Okay, you're going to have six interviews. Fine. It seems a little bit excessive, but fine. Who are they going to be with? So, I think the more information you can give up front to candidates before they've even spoken to you, the more you de-risk it for the candidate, the more likely they are willing to commit to that process. If let's say it's a 30 minute skills quiz, the much fairer it is. So the last thing we want to do is, for example, have a candidate take a quiz and they get into the first phone screen with someone in HR and be like, ''Oh, you're not in the right country. Therefore, you can apply for this job'' or you don't physically, you don't have that. Like, we want to just be as open and transparent as possible to candidates. So, then if they make the commitment, then they at least know what they're getting themselves into. And so that's where our product is going. It's helping customers like earlier on in the process, helping our clients early on in the process be more transparent with the candidates.
Jonas Christensen 20:17
Yeah and I think a new job is a little bit like a marriage. We spend so much time together, typically eight hours plus in the day, and we have to produce an output. We have to run a business together but we also have to like each other and feel comfortable in each other's company and build a lot of trust. But when we propose for someone to marry us, typically, we've spent a long time with them, but here, it might be two or three hours to assess each other. So it's a fairly short time. There is a lot of risk in that, that we somehow are willing to take on both sides of the fence. That I think you're right. You can de risk that enormously by just providing more upfront information and being more transparent. And that probably, as I reflect on it, is what has not been done that well in the past. It's sort of like, '' You can't know everything, because then you might know the answers to our questions beforehand''. It has to have a veil of secrecy to some extent in the hiring process. And the more we can remove that, the easier it becomes for all parties to figure out whether they want to be engaged and take each other's hand, in marriage almost, in an employment sense.
Tim Freestone 21:21
Yeah, as a quick pick up for one of our clients, actually. So we work with a business, GetYourGuide, based in in Europe, and they're really great at this. So before a candidate has even spoken to them, they will give them a document, which is like a thorough Q&A with the hiring manager for the role. So all the most common questions you're likely to ask as a candidate. So again, like what are the objectives? What are the exact problems we're working on? We're trying to solve this problem on the website, blah, blah. Obviously, everything around salary and remuneration, like straight up in the job ad, the first thing you'll see is exactly the remuneration, there's. Nothing hidden at all, you know exactly how much you're gonna be paid. They give a document around relocation. So if you're coming to Germany from another country, they show you exactly how the process is going to work, how they're funding it, who's going to take care of it. And they have a document around show options, like the exact quantification of how much show options you're gonna get and why. So I haven't seen any other businesses do that. And when we presented this to our candidates who we've worked with this business on, they've been really, really happy with that and saying, ''Well, this is this is amazing''. And that's not hard. Like that's just compiling information and making it available. So I don't see any reason why companies can't do this more often.
Jonas Christensen 22:28
Yeah, you've certainly inspired me to think a lot more about this for the next time I am going to hire new candidates, Tim. So, thank you for that. Now, you've touched a bit on it, I think, but are there any other sort of typical mistakes that hiring managers and candidates make when they recruit and apply for roles respectively, that you want to mention here?
Tim Freestone 22:47
Yeah. So for hiring managers and businesses, our thesis is basically that hiring is done in too much of an ad hoc, unstructured manner and it's too subjective. Our global vision for the next five years is basically trying to create a world where every candidate can get the job they deserve. And to do that, I think we have to make hiring as rational a decision as possible. And to do that, we need to have it as structured as possible and measure the right things in the right time. So I'll give you a quick example of business we're working with recently, where maybe this wasn't the case. So we started working with them a few months ago. We came in a little bit late to the piece and this company had already interviewed three candidates, a week and a half ago. These candidates hadn't got any feedback yet. They're kind of left on the backburner. I spoke to this business like, ''Yeah, we're kind of thinking through what the next stage is going to be. We're thinking. We're going to have six interviews stages''. I'm like, ''Okay. What are these interviews going to cover off? What are you trying to assess? Who are the interviews going to be with? When are they going to happen?''. ''Oh, we thought maybe this and maybe that''. Like, just such a vague process that obviously is going to be very subjective. Because if you don't define what you're looking for, how can you measure it? So, I think in general, businesses need to have like a really structured process and say, ''Okay, from the get go, here's exactly what we're looking for. Here are the three technical skills we need. We want a candidate with an advanced level of SQL and basic statistics, some visualisation skills. Alright, here's our four core business values. Like, we value having a data driven mindset, making it happen'', and whatever. You know, like, whatever they are for your business. Like, just defining exactly what you're looking for and then setting up each stage of the hiring process to measure those things and only those things. So, no moving the goalposts, no adding in 'Oh, let's just have an extra interview here''. Suddenly, the candidate speaking to some random person in the business who turned out was actually the key decision maker, and they hadn't been involved at all in figuring out the structural process. So, I think this is, to me, the big issue. Just like the lack of structured thinking. So, I'd say sort that out before you get going. Just define exactly the stage itself. And then as a general rule, I think most hiring processes are too long. Like, do you really need five interviews for an individual contributor? Come on, seriously. Like, can you not figure it out in a couple of interviews? And then the other big piece is the feedback. So there's nothing that's going to burn your reputation as an employer more quickly than ghosting candidates or giving them no feedback or crap feedback. Like incredibly vague, ''Oh they weren't a good cultural fit''. Like, what does that mean? You know, how did you determine that? How did you measure that? So yeah, try to be as structured as possible. Try to just focus on measuring the things that matter, and then doing that, and really, really thinking hard about your biases and avoiding those. So, a big one we get is hiring managers who just have these preconceived notions. I can think of a few in the last few weeks. Like, ''Oh, I don't hire people who come from consulting backgrounds''. So if you were applying for a role and you'd come from Deloitte or PwC, you had 0% chance of getting this particular job, irrespective of anything else. ''Oh, we've interviewed candidates in this business before and they haven't done very well. Therefore, everyone from this business doesn't have the right skills''. Like, hiring managers have a lot of things that kind of feel like experience or rules of thumb. But actually, if you dig deep, it's just a bias. It's just a discrimination. And again, sitting down thinking really hard about those and removing those. As a quick tip, I'd say anything that's placing someone in a bucket or a group, like ''Oh, someone from this country or this business or this industry, or has used this tech stack, I don't want to hire them''. Like, that's a bias and you really need to focus on assessing someone as an individual,
Jonas Christensen 26:31
Nice. And we are biologically wired to be biassed, because it's a very easy way to make decisions very quickly. But it also is a challenge when you have to make decisions on the large uncertainty, which is what hiring decisions often are and you're trying to take that uncertainty away. And if you think you're not biassed, then you're already biased by default. So we're all biassed, and Tim you're making me reflect on my own interview processes that I have run hundreds of times. I've interviewed many hundreds of candidates over the years. And I put a lot of energy into noticing my own biases. When I am listening to the candidate, I'm also listening to my own interpretation, which is tricky. But I do notice a lot of bias. And I often notice how I typically form a reasonably strong opinion on a personal candidate, about 10 minutes into the conversation. And if I had followed a traditional hiring process, then by this time I have read their resume, I may have watched the video often if that is part of the hiring process, but other than that this is me and this candidate having spent about 10 minutes together, which is not a very big part of someone's personality and their life and their skills and all that that's been presented at that time. So we are very biassed. How do we help hiring managers sort of step back from that and be more objective? You've described a lot of things already. And you have a product that tries to solve that. But what are some of the essential elements of that?
Tim Freestone 28:02
So yeah, this is a great question. The research I've done on this seems to indicate that it doesn't really matter how aware we are of our biases. It doesn't matter. If you've had 500 hours of unconscious bias training. It doesn't matter. You can't unwind 100 million years of evolution that is hardwired our brains to think automatically in a certain way. It's impossible. So to me, it's more about removing the opportunity for bias to happen at all. So, for example, CV screening. Like, you have a CV with all this person's personal information on it. Eliminating that entire process is the only way to do this. Now, how you did that in an interview is a much tougher question. I could imagine maybe a point where we hire people without ever having actually seen them. If we can automate cleverly enough of the process, if we realise actually, humans are more of a hindrance to the hiring process than helping it because we also fundamentally biased, then maybe it just ends up being a completely automated process. Now, we've done some limited trials of this ourselves for certain roles where I find people I've never spoken to before and they've just done a simple task related to their role, and it's worked out really well. So, I wonder whether the eventual solution to this problem will be something that seems quite radical and crazy at the moment. Like, Oh, you hire someone without ever speaking to them? I would not be surprised at that, if that's where we end up in five years. Because all the evidence we see is that there's a few big problems. People are very biassed. People are very overconfident in their own ability to tell someone's skills, abilities, whatever in the space of an hour. Like, talk about small sample size warning. And really interesting experiments that I think people could do to notice this themselves would be to interview a candidate together and then anonymously grade them at the end of that interview. And you'll straightaway see the variants. Like, we do this all the time when we interview together. We'll have a candidate. We'll ask him a bunch of questions and at the end, we'll submit our ratings blindly, for lack of a better term, and then we'll go through the differences. And sometimes I'll rate someone very well on a particular trait and someone will rate them very badly. We're both in the same interview with the same candidate over the same hour. How can we have such different opinions? It's because it's subjective. And so, we have to find a way just to make it as unsubjective and as objective as possible. However, that is, and I feel like that's probably initially adding a lot of structure, focusing on measurability, but then maybe eventually removing humans altogether, who knows.
Jonas Christensen 30:30
You're making me reflect on the way that I hire freelancers, which is something that I use a lot to get things done, including the team I've put together to do this podcast. The way I hire them is I give them the actual task to do and then I see how they perform and then if it works out, we can continue. That is something we never get to do in a job interview. We go through all the stuff and we sign a contract and whatnot, before we even get to have a go at the organisation and the people in it and our colleagues and all the rest in a typical scenario. So ideally, it would be the other way around. You come and work a bit with us and then we see how we go. But cause it's hard to fit in with the structure of the way we also hire people and the way the workforce is constructed today. Do you see that changing in the future if we took a really sort of long term vision to this?
Tim Freestone 31:17
Yeah, well, there's a few big picture trends that are happening. So one is what meant for us, so our entire team, we hire anyone from anywhere in the world. We always have. We always will. So at the moment, I'm just looking at our board here. We have someone in Sudan, someone in Egypt, someone in Brazil, someone in the Philippines, someone in Australia. So, the world has changed, because COVID forced us to realise that you could work reasonably effectively from home. If you can work from home from your bedroom in a suburb of Sydney, what's the difference between a suburb of Sydney and a suburb of the Philippines? Nothing effectively. So I feel like we're at a tipping point of maybe an incredible liberalisation of work and where people live and access to jobs. And it could be one of the most profound trends of our lifetimes. If suddenly, you could be living in a village in Brazil and get access to the same job as someone in Silicon Valley, that is genuinely amazing. You fast forward and think of the economic impacts and social impacts of that. And that's also wrapped up, I think, in just more fairness and objectivity. Like, that saying, ''Well, okay. So they're from overseas. Like, why couldn't they do the same job as someone here? Like, what's the difference?''. If you look at traditional hiring, even from 5 - 10 years ago, it would very common that someone would apply for a job in Australia, and the experience from overseas was just discounted as irrelevant, or somehow different, which is just bullshit. It's just straight up discrimination. So yeah, that and then yeah, the con of outsourcing platforms, Upwork, your Freelancer.com. I was a big advocate of those 10 years ago. I invested in all of them. And I'm surprised actually, but they haven't become massive scale since then. So there's still huge room to grow there. And similar for them, if they find a way to objectively measure people and connect them to the right job at the right time, then that could be transformative. I think this is all wrapped up in the same issue of bias and learning to understand people objectively and accurately.
Jonas Christensen 33:20
Yeah, wonderful. So if anyone's interested, the sound of this podcast is done in Newcastle in England. Hi, Tom. And the graphic design is done in Portugal. And the technology is done in the UK as well. So I'm just recording in Australia. So there's a lot of setup and it doesn't matter where people are. It can actually be really handy to have people in different time zones, because I typically do my work on this in our house. And that actually helps me and overnight something's done, and we're ready to rock the next day and things can move on.
Tim Freestone 33:51
Exactly. Isn't that a great world?
Jonas Christensen 33:53
We are living in a different time. It's a wonderful world. And I really liked what you said in the beginning, that 99.9% of the talent that is available is probably not living within 20 kilometres of where you're located and your workplace is located. So really, we need to expand our horizons there. And this is potentially the dawn of a completely different way of structuring our societies as a whole. Now, Tim, there's one bit here that we haven't touched on yet that I'm really interested in. Because I've mentioned already that as a hiring manager, I sometimes have to read through 50 or 100 applications to just find a small group of candidates that I want, because I'm looking at these applications. So there's a cover letter and a resume potentially. And there's of course, two big problems with this process. Firstly, we might not actually be attracting the right candidates. That's why we're getting so many resumes that seemingly aren't relevant. The other problem and this is a lot harder to see is the resume doesn't actually reflect what the person can do. It reflects what they have done to date. So if I look back 20 years ago, my first resume had pretty much nothing on it, other than a supermarket job. But I could do a lot of the analytics that I could do because my brain was functionally very similar to what it is now. I didn't have the leadership experience and all that. So don't hire me to be your next CEO. But for all intents and purposes, I was very ready to do that job and I could probably be hired for cheap if someone wanted to do that. But I did not have a resume that reflected that. How do we find these hidden gems that are everywhere in society and really are the ones that we want on our team?
Tim Freestone 35:30
Good question. So for more like potential ability types of candidates, so those who might have no experience at all, I guess there's a couple of options. I can think of a few people that we've hired who fit into this category, who we hired to quit uni to join us. So they had literally zero work experience. And under any traditional hiring process, they would have been rejected, because there's nothing on their CV. Again, we gave them a skills quiz, customised for the role, for the things we needed this person to do in this role. And they nailed it and beat everyone else. They got an interview. They were really good. We hired them. So, in that sense, it's pretty simple. If that's someone who already has the skills, even though they have no experience, I think that process works quite well. And this is great, because these people who have taught themselves those skills. Like, how amazing is that? Like, they have no experience and somehow in their own time, they're someone who'd like built a mobile app and launched it in his own time. This is exactly who you want. They're amazing, right? Becaus they're self motivated enough to do that. They didn't need to go to university to have someone teach them. They didn't need to get the experience in the workplace. They just did it. They're amazing. The other areas maybe are candidates who don't have the skills or experience. I mean, it's a little bit different. All the research I've looked at on this basically says that IQ is going to be the best predictor of someone's eventual ability to do stuff. So, giving them IQ tests. So the psychometric, verbal, numerical, diagrammatic reasoning types of assessments seem to be the best predictor. That's my incredibly layman's summarization of psychometrics. And so yeah, using that early on in the process, which to be fair is what most large organisations do as part of their graduate intern intake. It's they use those kinds of assessments. I think, I'm not sure there's anything better than that at the moment.
Jonas Christensen 37:14
It is a hard one. And it's also why we probably see advice for newcomers, for young candidates, for early career candidates, I should call it. The advice there is typically this recommendation to do extracurricular activities. In our space of data analytics, it might be Kaggle competitions, or filling your GitHub repository with experiments, joining hackathons, writing articles on Medium or LinkedIn or other places. Do you think these are the right types of activities that someone in that part of their career should be doing and how does that sort of stack up against a more black and white assessment, like what Alooba provides?
Tim Freestone 37:54
Yep. Okay. So, from the candidates perspective, I guess, there's two different parts here. There's them needing to acquire the skills and experience, even if they don't have a job. So, from that perspective of actually just learning the thing and trying to do. Yes, the world is your oyster, in terms of potential analytics things. Like, you could download your Netflix data. You could have weeks of interesting projects there and you could learn end-to-end analytics, just by doing that one thing, cleaning the data, visualising it, building a model to predict when you're going to next watch this show. Like, you know, anything. There's so much data out there. You could do all those types of things to teach yourself the skills needed. Whether or not that helps you in the hiring process, I think is another question. I'd probably split hiring into two different categories. One is you graduate in internship programmes, which attracts like often tens of thousands of candidates for big banks, insurance companies and then let's say junior roles that aren't going through those formal graduate internship programmes. The reason I split those up is just because the hiring process is quite different. So, as a candidate, I probably optimise different things. So, for the internship and graduate programmes, because they have such an enormous volume of candidates, as you can imagine, they use automated techniques as the first screening layer. So they might have 10,000 applicants. Normally they'll filter based on something like your weighted average market university, and/or they'd send you psychometric tests and you do those, and they'd reduce the candidate pool by 95% off the bat. Then they'd be looking at your CV, interviews, assessment centres, all those kinds of things. So in that situation, if your aim is to get a foot in the door, I probably wouldn't worry about GitHub and Kaggle and everything. I just practice psychometric tests. And the reason I say this also is because this is exactly the situation I found myself 10 years ago. And that's the origin of another business I created. I was basically practising these IQ tests to get better at them. Despite what people may think, actually you can improve, just through rote memorization and patterns and those kinds of things. So I would just understand the hiring process you're going into, and, like, the rules of the game and know how to beat those rules of that biassed and silly game that you basically have to play. Now, once you get to an actual interview stage, then I think that's where the extra pieces become really valuable. Then you can suddenly have a really intelligent conversation about this particular project that you did. But in terms of getting a foot in the door, it might not help that much. Now, in terms of, like, the hiring manager and the screening process, comparing, let's say, ''I've got 200 candidates applied. Could I look at their Kaggle and their GitHub vs, let's say a skills quiz?''. I think the difference there would be a few things. One is the scale issue would be one. So manually going through and looking at all these different Kaggle competition results or GitHub repositories will be quite unscalable and difficult to compare, I'd say. Like, how do you say, ''Well, this person's Medium article is better than this person's Kaggle competition result''? Like, it's hard to compare. And third thing would be the only people who could do that comparison are probably the hiring managers with a lot of skill. And often, the initial screening processes are not done by them. They might be done by a recruiter or someone in talent acquisition who probably isn't best place to make those kinds of nuanced decisions, I would say.
Jonas Christensen 41:17
Yeah, I think for 99% of the population out there, Kaggle might sound like a cleaning product rather than a really important tool in data science. So you're absolutely right. And unfortunately, you might need a bit above to get through.
Hi there, dear listener. I just want to quickly let you know that I have recently published a book with six other authors, called ''Demystifying AI For The Enterprise: A Playbook For Digital Transformation''. If you'd like to learn more about the book, then head over to www.leadersofanalytics.com/ai. Now back to the show.
Okay, Tim, that was really enlightening. So if we sort of broaden our horizon a bit and look more generally at candidates wanting to stand out in data and analytics, what skill sets would you recommend for them to acquire? You are totally allowed to split this up by different role types and also different levels of seniority and so on.
Tim Freestone 42:20
Yeah. Yeah. Great question. So firstly, I just, if I were a candidate, I'd be thinking, ''Okay, well, what's in demand?''. Like, for whatever role I want to go for, as a product analyst, as a data scientist, as a manager of analytics, like, what is actually required of me? A reasonable proxy for that is just going to be looking at a whole bunch of job ads. So go and look at, I don't know, 20-30 job ads, product analyst roles, types of companies you want to work at, and look at the requirements. The reason I say it's a reasonable proxy is sometimes they end up as laundry lists, with all these extra things that actually, you're not actually going to be evaluated on in the process and they're just nice to have. But I think you'll get a general sense from that. So let's say as an example, a product analyst role, so almost always SQL is going to be a required skill. And generally, I should say, I take like the Pareto approach to this, you know, the 80/20 rule. Probably there's like two or three things, if you got really, really good at them. Then that's going to be 80-90% of the effort and getting through the process. So almost always is going to be some SQL. And that's going to be like your bread and butter skill. Probably gonna have to visualise data in some tool. Doesn't really matter what it is. Power BI Tableau, Looker. I would probably, if I had to choose, I would choose one that's more likely to be used by other companies I'm applying for. So for example, if you are generally targeting, let's say, big banks, corporates, they're more likely to use Power BI, because it's the Microsoft stack. Whereas if you're going for more like an early stage tech company, they're unlikely to use Microsoft, More like a Looker. So you could optimise within that. And then another big one would be just like defining metrics and connecting that to the business. And for any role I've ever seen, There's not a single hiring manager who wouldn't argue that communication skills are really important. Now, what is good communication and what is not good communication is very subjective. But generally, you can think of things like being succinct, being very articulate, explaining what you mean, so that anyone could understand in very easy to understand manner. Using the STAR method in an interview is probably a good rule of thumb. If you don't know what that is, look that up. Basically talk about, ''Oh, I was involved in this particular problem at work. Like, we had a drop in the conversion rate. I was tasked with investigating this drop in conversion rate. I ran a couple of reports and I broke down the results by a few dimensions. I found that the drop in conversion rate was just in the mobile app category. I investigated that and there was a bug and we fixed it. And then the conversion rate went up''. Some kind of easy to understand STAR methodology. So yeah, I'd focus on the things that seem to be most relevant and most important that companies are trying to assess for and just get really good at the three or four of those and forget about all the other fancier fluff stuff because you just waste your time probably.
Jonas Christensen 45:01
Nice. I agree with all those points. My additional comments would be definitely look up the STAR methodology. Situation - Task - Action - Result, I think it stands for this off the top of my head, which is really your structure for describing something. Which is, again, what Tim talks about. If you can structure your communication so that it's easy for someone to interpret, then half your job is done. The other thing is find the cross0section between all these things you need to do and whether you actually enjoy them or not. It's really important. It's very hard to learn a skill if you hate doing the learning part. But I think a lot of people actually end up forcing themselves to do something that they don't intrinsically enjoy, and therefore they will never become really good at it.
Tim Freestone 45:41
Yes, that's I think, common in any hype cycle, isn't it? Like, data science became so cool few years ago, maybe still is. So tracks all these kind of bandwagoners, who maybe view it in a certain light. And then when you get down to the ADF line of annoying SQL query to clean the crap data that you have, maybe your heart's not in it anymore by that point. So you have to really enjoy the nuts and bolts of what you're doing. Otherwise, you're probably not gonna be very good at it, or you're not going to last too long.
Jonas Christensen 46:11
That's right. I think data cleaning is coming back in vogue. We're all talking about the data centric modelling and all that stuff now. To me, that just means we're putting focus back on data quality, not just models in an essence. And I think it's very stressful for candidates really to consume the plethora of tools and techniques that are out there. There's stuff coming out all the time. ''Oho, now I need to learn R. Now, I need to learn Python. Now, I need to learn all the three main cloud providers. I need to learn the tools that they provide''. And I grew up doing SQL and SAS, so therefore, it's actually hard for me to transition. I've spent 15 years on that. How do I let go of those things? And it's really overwhelming for people all this stuff. I think it's fair to say that, at least from the statistics I've seen, most people in the industry don't know more than two types of tools or techniques, like the ones I've described really well. So don't stress too much out there.
Tim Freestone 47:06
Yes.
Jonas Christensen 47:06
If can't do that. What's your view on that actually, Tim, because this is part of your back end of your tool? You can see how well people score on various techniques. So you must have a bit of an insight into all this stuff.
Tim Freestone 47:17
Yeah, for sure. I definitely read out what you've said. I mean, I own a business in skills assessment analytics. And I can tell you, probably 70% of the tools in analytics, data engineering, whatever, I've never heard of. So do not freak out for a second, if you don't know these tools. You're not experts and nobody is expert in 100 tools. So again, our customers choose to test candidates in pretty predictable skills. And again, follow kind of like the Pareto rule that I've mentioned is that, let's say most of our product analysts, data analysts roles, they will choose to assess SQL, bit of R or Python, some basic statistics and visualisations. Like, it's fairly predictable and concentrated in that sense. See, maybe that gives people a bit of hope that if you just focus on the bread and butter, and maybe taking a step back for a second, just saying like, ''Well, what is my job as a data analyst? My job is to solve business problems. No, to investigate things. To make more money''. Like, that's it. If you think of those simple terms, you can just do away with all the fluff, all the mayhem all the overcomplication and fancy jargon and words and what have you. And, ''Okay, well, what do I need to do that? Well, I need to get some data from somewhere. Okay. It's in a database. I have to write some SQL code. I gotta clean it up. I got to understand what these metrics means soI have to have some business acumen. And then I've got to demonstrate to people my findings somehow. Okay, so I got to communicate it in writing and graph. Cool''. Like, it's was not hard. If you boil it down to its basic essence, it's not rocket science, as they say.
Jonas Christensen 48:44
First principles. Very good. Now, Tim, we're sort of towards the end of the interview here. So I've got three questions left for you. Before we get to the closing remarks, are there any must have skills that you see hiring managers typically overlooking that you want to call out?
Tim Freestone 49:01
Overlooking in their candidates?
Jonas Christensen 49:03
Yeah, or not focusing on enough.
Tim Freestone 49:06
None spring to mind but then again, we maybe have a different perspective on this in the sense that we say to hiring managers or businesses, ''You figure out what you need and we'll help you assess that'', as opposed to saying, ''Well, actually, I think you need x''. So we're a little bit passive in that regard. So no, nothing springs to mind. If anything, it's probably the opposite, which is trying to evaluate things that don't matter or evaluating things that you can't really evaluate or doing valuations in a subjective way. You know, like, how many times have I heard the word ''cultural fit'' in the last year? Oh my God, I hate that term. You can reject a candidate for cultural fit for any reason. Anyone can say anything for that, I just really think harder about, ''Okay, what are the three things that we can evaluate and measuring those as objectively as we can, and not arbitrarily excluding people for other reasons?''
Jonas Christensen 49:59
Nice. I like that answer. Simple is key sometimes. Now, Tim, my last couple of questions. So the first one is one I always ask guests on the show, and it is to pay it forward. So who would you like to see as the next guest on Leaders of Analytics and why?
Tim Freestone 50:16
I got three great recommendations. So Iman Behzadian is GM of Data Science Products WooliesX here in Sydney. They're doing some really interesting stuff from what I hear. I could recommend Rosyll Xavier from Linktree. Linktree are an Aussie scale-up doing really well. Just raised another 100 and something million in series A or B funding last week. She's heading up data analytics, so she'd be someone really interesting to speak to about how to manage the chaos and how analytics works in such a high growth environment. And that'd be really fascinating. And then Ben Jarvis, who heads up sales analytics for Google in Australia, New Zealand. He's got quite a different bent and example on analytics as well. So definitely have a chat to him.
Jonas Christensen 50:57
Brilliant, all three will be contacted by me very shortly. So thank you for that, Tim. They are great recommendations that I can see will fit into this show. Very well. Now, Tim, lastly, where can people find out more about you and get a hold of your content?
Tim Freestone 51:12
Yeah. So best place would be to connect with me on LinkedIn and to find us at alooba.com. It's alooba.com.
Jonas Christensen 51:20
Great. Tim Freestone. Thank you so much for being on leaders of analytics. It was a very enjoyable conversation and also very insightful. I learned a lot and I really am reflecting on just how I am totally wrong and biassed in the way I do things in the hiring process and how I can get much better at that with my colleagues locally and around the world. So thank you for this enlightening conversation and all the best for you and Alooba in the future.
Tim Freestone 51:46
Thanks for having me, Jonas. It was a great pleasure.
