The Rise of Hybrid Jobs & the Future of Data Skills

ation was realizing that we actually could understand, we could collect data on both job postings around the economy, we could collect data at scale about people in their careers, create effective ontologies to understand what people are expressing, what signals are coming to the market, and provide critical insights that help companies, that help policymakers, that help educators and help workers understand the job market, plan for the job market and make more effective connections within the job market.

Adel Nehme: I want to set the stage for today's conversation. When preparing for this interview, I was in awe of the level of depth and care you and the Emsi Burning Glass team practice when you're thinking about, and speaking about the labor market. You mentioned here, the data science methodology, underpinning it, do you mind expanding into that methodology and how you were able to model the job market just so effectively?

Matt Sigelman: Yeah, absolutely. It's interesting, because we started not as a data company, we started as an LP company. We had developed a really good engine for recruitment that uses advanced NLP to be able to structure people's CVs and upload all the information and make more effective matches on that basis. And in fact, that's technology that's used even today by the great majority of large recruitment companies and HR management systems and the like. But after a time we came to realize the constraints of this, because on the one hand we created this better mouse trap that's able to structure the unstructured coin of the realm of the job market, CVs and job postings and the like, and use that to help individual clients. But ultimately the job market still mostly works on cosmic coincidences, which is say you go walk into a cafe and you see somebody as a server who's, she's fabulous. And you say like, "This person could be working anywhere. Why is she here?"

Matt Sigelman: And so we've constructed this job market that only works on the spot market, that whoever happens to be looking for a job any given day, and whoever happens to be looking for talent in a given day. And so what we realized was that part of the reason why that's always been the case is that there's no market map. It's hard for an individual job seeker to know what are all the opportunities that are out there, employers, likewise. And so you can't plan for a market you can't connect effectively within a market, but you don't understand. So what we did is we said, "Hey, look, actually the world has evolved in most industrial economies, to where most hiring is happening online, most job postings are online." And so instead of waiting for just processing the data that our clients receive, what if we go out and actually scrape all the job posting you can find.

Matt Sigelman: A lot of, by the way, labor economists around the world, use our data set and find that, general consensus estimate is about 85% of all job postings in the 55 countries that we cover are actually in our database, in the [inaudible] database. And so what we then do is we bring those job postings back and we've constructed a really robust set of ontologies that help us define roles, that help us define skills, that help us define experience levels, credentials, and so forth, across dozens of dimensions. And then we use our NLP expertise to then be able to create, to attach those metadata tags to each experience within each CV, to each job posting. And that's really important, because it allows us to aggregate up the information.

Matt Sigelman: An analogy for what we've done is some work that's being done by an economist at MIT, by the name of Alberto Cavallo. Several years ago, he was a young economist at a university in Buenos Aires, and he was trying to understand inflation in Argentina and government statistics, he felt, weren't giving him an accurate read. And so he said, "Look, I can find prices, I don't need the government statistics, the prices are all online." And it totally transformed them. We actually were doing this before Alberto did his work, but I always find that it's a ... he's now running this thing called The Billion Prices Project. Emsi Burning Glass is, you could call it the many billions of jobs project, and it's been transformative in how we understand the economy.

Adel Nehme: And what I love about this approach is that instead of looking at job evolution, it looks at skill evolution within the labor market. And that gives you a real time view of the job market, the skills are evolving within it. So given this, what makes this approach different from the traditional labor economics approach and what type of value does it offer that wasn't previously attainable?

Matt Sigelman: So a traditional labor economist is usually transacting in survey based data. So all the things, we see the monthly job numbers or whatever, they're based actually on a survey, it's a very 1950s, '60s notion that, hey, we can't possibly be able to analyze the job market at scale, and so we'll take this very narrow slice and we'll use a survey based methodology and we'll use that to be able to understand what's [inaudible]. And by the way, those surveys are valuable in being able to provide a benchmark, in being able to validate what we're seeing from a broad trend perspective. But the weakness of that kind of approach is that, as you can imagine, when you're doing a relatively narrow sample, it means two things. One, you have to keep your categories, three things, actually, one, you have to keep your categories really broad.

Matt Sigelman: So the US Bureau of Labor Statistics continues to track a job called computer programmer. What's a computer programmer? Number two, it means that you have to assume that every job within one of those categories looks the same. So, okay, there is a whatever computer programmer it is, every kind of computer programmer does the same work, same skills, to your point, you don't get that granularity. And then third, whatever you do, because you're trying to create a macroeconomic trend line, don't mess with the buckets. So I think only about three or four years ago, I maybe slightly off that, did the Bureau of Labor Statistics actually start to recognize that there was a job called the data scientists, they just don't want to recognize it because it messes with their sample.

Matt Sigelman: Whereas to your point, what really was breakthrough about this is not even just that it gave a more real time view of the economy, because obviously [inaudible] that survey work takes a lot of time too. So not only is it literally what's going on today in the job market, but what it's doing is it's allowing you to get to all that granularity. So jobs that are Ruby developers, Ruby developers in one industry have actually a slightly different skill mix from Ruby developers in another industry. And so you can see skills emerging within jobs, skills that are transforming jobs, and you can see the birth of new jobs altogether.

Adel Nehme: And I don't want to spoil the rest of the conversation, but the skill based approach sheds light on how traditional jobs like marketing, are becoming differentiated or more technical. And you get a view on how roles are evolving and this helps governments, organizations, educational institutions prepare better for the future skills folks need. And so I'd love to unpack how you see the job market evolving and the type of skills that will define the future of work. Now of course, digital skills of all types are growing in demand at various rates, but given the theme of today's podcast, I'd like us to focus on the demand for data skills. Can you walk us through just how much data skills are in demand today and to a certain extent, where you think they're headed? And of course I'm using the umbrella term data skills here, since there are many sub skills that we can further talk about and break down.

Matt Sigelman: So data skills are very much transforming the market and very much transforming, not just because there's jobs like data scientists, which are in huge demand, but because those skills, exactly to your point, are in demand across the economy, now a whole range of jobs are becoming what I would call data enabled jobs. So one thing that we're seeing is that that pace continues to quicken, so if you were looking in US data, for example, you would see that the number of jobs that require data science skills, and again, I'm using that skill metric, not the job metric, the jobs have also grown, but data science skills showed up in about 450,000 jobs last year, up from about 350,000 just before the pandemic. So literally in the space of two years, you see about a 40% growth in demand for these skills, think about how fast or not labor markets tend to move, which is a couple percent a year, and think about a 40% growth in the space of two years, a tremendous growth in that demand.

Matt Sigelman: Now, when you think about how we characterize that demand, I think there's a couple things I would point out here. One, that the buckets of data skills that show up are going to vary fairly differently, fairly significantly rather, based on the kind of job that we're talking about. If you were to look at the skills of a data scientist versus a data engineer versus data analyst, they have very different skill sets. So common across all of them is Python, not all data analysts need Python, but even a growing number of data analysts need them. Python seems to be winning out a little bit over our both prevalence and also in versatility. I think if you were looking at a set of skills that you would see across all three of, for example, those buckets, but here's where you start to see the differences emerge.

Matt Sigelman: So a data engineer needs ZTL skills, a data scientist generally does not. A data scientist, a growing percentage of data scientists now need ML skills. Five years ago that wasn't true, it's true today, but you don't see ML skills showing up in data analyst jobs, and in a relatively small percentage of data engineering jobs. So those kinds of differences are increasingly coming to bear, but I think the key here is really looking beyond the world of data science from a job perspective. And again, not because there isn't huge growth in data science jobs, but in fact, here's a little fun fact for you, in 2010, there are only 150 job postings in the United States for data scientists. Last year, there were about 50,000 and there were another 50,000 jobs for data engineers. And data engineers, I think probably, I don't know where when you started to see, but I bet even in 2014 or 2015, there were probably pretty close to zero. So you're seeing huge rates of growth.

Adel Nehme: And what I also love about breaking down skills, methodologies, that it also shows you a proxy of where organizations are on the data maturity curve. So for example, you mentioned machine learning skills are much more in demand today than they were, let's say five years ago. It's because five years ago, I don't think a lot of organizations knew how to operationalize machine learning like they do today. So it shows you where the organizations lie on the data the maturity spectrum today.

Matt Sigelman: That's a really great point because I think one of the things that it provides is a way for companies to assess their future readiness. Hey, where are we, and we look at a given role, what are we looking for in that role, and how is that different from what our best practice peers are looking for in this same role? If you were look at, for example, product managers at City or JP Morgan Chase, then look at product managers at Amazon, Amazon product managers are data enabled product managers and a lot of mainline companies are not. So you can benchmark yourself around like, hey, what are the skills that we're going to need to be able to manage transformation in our industry?

Hybrid Jobs

Adel Nehme: Harping on something you mentioned here is that we don't want to constrain this conversation just to roles, we want to focus on skills. So I don't think many people today define this as brilliantly as you do, and here are the concept of hybrid jobs. Can you walk us through how you define hybrid jobs, especially when it comes to data skills?

Matt Sigelman: Yeah, absolutely. And first, I'll define terms around hybrid jobs because the reality is that today in the post pandemic world, I think hybrid jobs are starting to come to refer to something different, which is this notion of you work in the office some of the time and work remotely some of the rest of the time. But when we first started tracking what we call hybrid jobs, we were tracking a really fascinating and I think actually a really disruptive trend that we've been increasingly seeing in the job market, and that's the tendency of jobs to blend skills from across domain. Overall jobs are changing and changing very fast.

Matt Sigelman: So from some forthcoming research that we've done with our colleagues at the Boston Consulting Group, we found that the average job, across all jobs, we're not even just talking about tech jobs or data jobs, across al, the average job has seen about a third of its skills replaced over the last decade, and that pace has been quickening through the pandemic. By the way, huge implications for that, because think about your traditional university based learning structures, have they changed a third of their curriculum in the last 10 years? I don't think so.

Matt Sigelman: And by the way, think about employers as well, okay, have the skills of our workforce, has our workforce changed a third of its skill base? Probably not. And so raises real questions, what we were just talking about a minute ago about obsolescence. So that provides an overall frame of the pace of skill change, but it's easy to just say, "Hey, most of that change is probably just people need to learn new tech stacks." Okay, fair enough, but you can do that on the fly. But what we're really seeing much more profoundly is that jobs are absorbing skills from across domains, essentially jobs are having sex.

Matt Sigelman: I'll give you an example, I think actually, you pointed this example before, think about a marketing manager. We all know plenty of people who are marketing, nice folks, they tend to be pretty right brain, they go to marketing because they understand people and their psychology and they can communicate to them and whatever. Increasingly we want marketing people to be able to manipulate customer data. Well guess what, you need some data skills to do it. And so a marketing manager who has SQL skills, I'm not even talking about hardcore data skills, but a marketing manager who has SQL skills and you just manipulate a customer database, gets paid about 40% more than one who does not have those same sets of skills. So just speaks to how we're creating these intersections of skills that were never seen before, and that challenges the job market because apparently, think about that example, right brain person, left brain skills, you're creating an immediate shortage right there. And so for people managing their careers, the ability to blend skills and get ahead of that is the ability to put yourself in the cat bird seat in the economy.

Adel Nehme: That's really awesome. So I want to focus in on the data skills of hybrid jobs. Of course it's entirely job dependent, industry dependent, specifically what type of data skills are needed within these hybrid jobs? But my hunch is that a lot of these data skills needed within these types of roles are not necessarily hardcore technical skills like data engineering or machine learning. So what do you think are the main data sub skills that are rising within these hybrid job categories?

Matt Sigelman: So first of all, just to reinforce your point, part of the reason why jobs can hybridize in various directions, by the way, it's not just that data skills are invading other jobs, but in fact, if you look at a job as a data scientist and compare it to the quant jobs that preceded a data scientist actually is very strong, in fact, it's a great example of hybrid job because data scientists needs significant programming skills, which are related to, but not the same as data skills, but also significant business skills because data scientists need to solve business problems and they need to understand those problems.

Matt Sigelman: But look, when I think about the world of data skills or of hybridization and hybridization of data skills and others, one thing that's important in framing our understanding of why this is happening is that skills, even technical skills have gotten a lot more accessible. This is where I get to do, I'm an old guy, so this is where I get to do my when I was a boy, I used to walk up hill both ways to school and barefoot in the snow moments, but my original work and data was in Fortran and in SPSS and ...

Adel Nehme: Yeah, that must have been challenging.

Matt Sigelman: It was a much less accessible language. By the way, I'll give you, I know this is off on the side, but it makes me laugh to remember it. There was a command when you're working with big data sets, was like call tape whatever the tape number was. And it was only, and I'd realized that half the time you'd call that it was very slow, I figured that was just a processing issue, and some half the time it would hang. And I only later discovered that when you type call tape whatever, it rang a bell at somebody's desk and someone had to go and fetch a tape and load it. So they're away from their desk or whatever they were doing, that's why it would hang, because they just didn't [inaudible] across.

Adel Nehme: Oh, wow, that's hilarious.

Matt Sigelman: But so look, when we look at the state of a lot of tech stacks today, they are more accessible, I might say in some cases easier to use, they're also more powerful. But what it means is that people in a broader range of backgrounds can actually leverage those skills because you don't need a deep specialist in order to be able to use data skills. You used the example of marketing we were talking about before, 10%, almost 10% of the jobs that ask for data science skills, not just data skills, but data science skills are in marketing. So that speaks to that accessibility has facilitated this hybridization, hybridization isn't just data skills, it's also design skills are being used across, various programming skills are in demand across occupations, business skills, even a nurse, for example, needs project management, because she's managing care across providers, and also regulatory skills showing up in a broad array of jobs.

Adel Nehme: Yeah. So this is especially useful for finance jobs where it's heavily regulated industries, and to harp on that point around modern data tools being accessible, outside of accessibility, they're also free, Python and R are open source, you no longer need to pay to license, like similar that you had with SPSS or SaaS, to start doing interesting data work. And following up on some of the points here, outside of marketing and some of the jobs you mentioned, do you think only a set of these jobs will become more hybridized, such as finance or marketing, or do you think this is a secular trend that's really going to impact most jobs?

Matt Sigelman: This is broad based trend. And by the way, it's not even just a across professional jobs. We recently looked at the digitalization of what are sometimes called middle skilled jobs, that means north of secondary school and south of university. And what we found is that about eight in 10 of them today are digitally intensive and digitally intensive, middle skilled jobs are twice as likely to pay a living wage, they're growing twice as fast. The two in 10 middle skill jobs that are not digitally intensive are increasingly just in construction and transportation. So this is really broad based. I think one of the things that we've been looking at recently in our data, is trying to identify what are the foundational skills of the new economy and what are the sets of skills that are broadly in demand. And I think traditionally, if you had looked at foundational skills, you would say, "Okay, it's the human skills. It's what people like to call the soft skills." But actually we're seeing at least as much as broad based demand for data skills, for digital skills, for business skills.

Soft Skills

Adel Nehme: So I want to riff on that a bit, you mentioned here soft skills and I think when people often think of hybrid jobs, they immediately assume that these jobs will become inherently more technical and I'm going to have to become a hardcore programmer. However, one thing I've seen you cover within the concept of hybrid jobs is that hybrid jobs are increasingly become technical, but the more they become technical, the more valuable soft skills are when it comes to succeeding in this hybridized economy. Do you mind expanding on that notion?

Matt Sigelman: Yeah. This is a fascinating conclusion when you look at these core foundations. Normally when you use, in fact, even just using the vocabulary of a foundational skill, you're expecting that foundational skills are the stuff that's on the bottom. Think about the food pyramid, where you've got carbohydrates down here and then the important stuff is protein, that's the technical skills. Actually, your career works exactly the opposite, that the further north you go in your career, the more you advance, the more relative value employers place on core foundational skills. It's also, and I think this is to your point, one of the other things that we found is that the more hybridized the job is, and just think hybridization in a lot of ways proxies for the jobs that are most tech enabled, that are most data driven, the more tech enabled the job is, the more data driven a job is, the more intensive it's demand for foundational skills, for soft skills.

Matt Sigelman: So the most highly hybridized jobs are about three and a half times more likely to value creativity skills, about twice as likely to value collaboration skills, about 50% more likely to require writing skills, problem solving skills, research skills. So I think you're exactly right. There's an economist at the Harvard Kennedy School, who has been doing a lot of work inside the Emsi Burning Glass data, and has found that the jobs that are growing the fastest and they're highest value, are the ones that combine deep technical skills with human skills, with judgment skills and the like. And so when I think about how people are managing their careers, a key thing to remember is that you want to make sure that you're not just acquiring an individual tech skillset, as important as they are, but that that training is baking in the soft skills that actuate that skillset, that technical-

Adel Nehme: And this is what gives me hope, to a certain extent, a major part of the economy are right brain type roles, whether marketing roles or even traditional roles have already embedded within them solid communication skills, the ability to collaborate with others. And so once you supplement these roles with technical skills, you start seeing a lot more powerful, more effective people in these roles.

Matt Sigelman: That's exactly right. I think you're seeing two things happening at once, one, a range of careers that are increasingly being enabled by tech and data skills, that are being rendered more valuable through tech and data skills. And by the way, they're becoming more future proof and robot proof because they have those tech and data skills. But at the same time, you're seeing a growing number of people in tech careers who are realizing they need to acquire management skills, that they need to acquire leadership skills and other human skills to be able to make themselves effective.

How are technical skills and data skills are quickly evolving over time?

Adel Nehme: So one thing that I've also seen you discuss is how technical skills and data skills are quickly evolving over time. And that even within the professions that are highly sought after like data science and data analysis, you have a certain degree of skill obsolescence that wasn't necessarily true in other professions before. Do you mind expanding to that notion as well?

Matt Sigelman: So first of all, I think this is a really important point because the skills of roles change much faster than the roles themselves. We're used to hearing these hyperbolized statistics about, well, hey, by 2030, 70% of us aren't going to work in jobs that have been born yet, it's total nonsense, but the skills themselves of roles actually change much faster than that. So as an example, if you were to look at the skills of a data engineer or sorry, of a data scientist, if you were looking just over the last five years, you would see a huge increase in demand, literally about a tenfold increase in demand inside data scientist jobs, which themselves are growing. But for data visualization skills, for deep learning skills, for NLP skills, about a 500% increase in demand in big data skills. And then at the same time, things like Pearl scripting, like MATLAB and C++ significantly declining in demand within those roles. So those are pretty significant transformations that you see.

Adel Nehme: That's great. And what is a strategy that data scientists or folks can adopt, to keep their skills competitive? If you were a data scientist, how would you go about your career growth and planning?

Matt Sigelman: So I think data scientists are no strangers to data, and my advice here would be to data driven in how you manage your own career. So you need to be able to use data to understand what's the landscape of opportunity, you need to be able to use data to be able to understand what skills are emerging within your field and across fields, and you need to be able to use data to figure out what skills you need to acquire to stay ahead of the game.

Adel Nehme: In some sense, the hybridization of jobs and the increased velocity of skills transformation, and the rise of data and digital skills, have ushered in this paradigm shift within the labor market and how we think about skills and jobs. I'd love to segue here into how you think organizations, specifically educational institutions as well, can adjust to managing such a change. How do you see education evolving to address the skills transformation we're witnessing today in the labor market?

Matt Sigelman: So I think there's a bunch of things that are going to need to change. So first of all, in the education system as it exists today, even there, I think institutions, universities, others, need to become dramatically more agile in terms of how they track the landscape of opportunity for their graduates and build skills into their curricula, evaluate their curricula, make sure that they continue to be aligned, make sure that they are building differentiation for their graduates. But I think more broadly to your point, we're going to see a significant transformation in how education happens in the format and structure of education, because right now, education is, for the most part, in most countries, a once and done phenomenon. You go to school, you slog through it, you get your degree and you never look back.

Matt Sigelman: But think about a world where a third of the skills of an average job changed in the space of 10 years, where if you look at a job like a data scientist, by the way, data scientists and data engineers were the two jobs that had the greatest pace of skill replacement across thousands of jobs, over the last decade. And so when you think about that imperative, it says that the structure we need is not once and done, that increasingly people need to have to shorter form programs, to programs that are adapted to learning on the fly, that enable people to acquire skills from across domains quickly. It also says by the way, in a world where the job landscape is changing quickly, where lot of jobs are automating away, other jobs are getting born, that one of the things we also need is to develop structures of learning, which are titrated specific to specific transitions.

Matt Sigelman: Hey, I want to get ahead in my career as a data scientist, what are the sets of skills I need to acquire, very specific sets of skills, or I'm currently working as a financial quantitative analyst. And by the way, as the skill landscape changes, some of the transitions that are available change. So used to be a financial quantitative analyst, can't even say that, would go on to become a computer scientist because they had C++ and MATLAB skills and other things like that. Increasingly the skills that they have, they need to have, position them to be a data scientist. Okay, I'm a financial quantitative analyst, I said it right this time, I want to transition to be a data scientist, what are the skills I need to acquire to make that transition? And so we're going to start to see learning be much more personalized to the kind of transitions that enable people and empower people in their careers.

Adel Nehme: And in some sense, this is a bit of a controversial question, but do you think that the business model of universities today is geared towards this transformation?

Matt Sigelman: I think it's going to challenge higher education, but I wouldn't count it out. I think you're right, that current university business models are structured on once and done, and I think you're going to see them be very resistant to change. Now, I think you'll see more future for traditional higher education players in countries where higher education revenues are tied to student enrollment, as opposed to just government grants. In a lot of countries, a lot of continental European countries, university just gets a budget from the state, from the nation, from the national of government every year, and not a lot of impetus to try to drive enrollment. In a lot of places like the UK and the US, the Netherlands, there's more incentive to drive around enrollment. and I'll give you an example.

Matt Sigelman: In the US today, there's 17 million people enrolled in higher education programs, at traditional colleges and universities. I would argue that that number needs to be more like 30 million. The growth of that market, and especially given shrinking demographics, your only prospects of growing are to grow to be able to serve people in the middle of their careers. And there's a lot of incentive to do that, but a lot of organizational resistance, a lot of faculty resistance, a lot of business model resistance that would need [inaudible].

Addressing the Skill Gap

Adel Nehme: Something we experienced at DataCamp, we work with a lot of organizations trying to fill their talent gap, with up skilling. There's tons of discussion today on the role of organizations and learning and development teams and reskilling and up skilling their workforce to accommodate these hybrid jobs and the increasing demand for data roles. What do you think of how organizations are addressing the skill gap and what are your recommendations here?

Matt Sigelman: So I think we're right at the precipice of seeing a significant transformation in terms of how companies manage talent and how they invest in learning. Right now most companies haven't the faintest clue who works for them. I mean, look, they know your name, they know your tax ID number, they know how much they pay you, before the pandemic they knew where you sit, they don't anymore, but what they don't actually know is what skills you have. And in fact, most companies don't even necessarily even know what skills they need. And so one of the first steps that companies are starting to go down the road of today is to define a role architecture. What are the skills that I need role by role in my company? Because if you don't know that it's hard to figure out how do I build up the pipeline of talent? Where is my talent today? Where is its obsolescence risk? Where do I need it to go?

Matt Sigelman: And so they're starting to do that initial mapping, once they do that I think that's going to change a lot because today, a lot of the way that companies think about learning is either systems and compliance training, [inaudible] and I need to be able to make sure that my customer service representatives can use the new reservation system or whatever. Or it's these learning as a benefit platforms, notion that, okay, well we'll let people learn stuff and maybe they'll be more engaged and maybe they'll be more likely to stay.

Matt Sigelman: And there's nothing wrong with engagement and retention as metrics, but think about how much that transforms in an era of talent shortage when I can say, "Wait a second, I've got the talent that I need, I've got people with a lot of the right skills already in my workforce." And instead of firing them over here and hiring more people over here, I can instead build that pipeline that connects the reservoirs of talent to where I'm having talent droughts and to bring in learning partners, to your point, who probably aren't universities, to be able to say, "Okay, how can we create those very specific, learning those skill pipelines that enable internal mobility?"

Adel Nehme: And to your point here that the biggest challenge is cultural, how do you create a mindset shift to become a learning organization? And that requires a deep appreciation of the subject matter expertise your people have and creating that pipeline for them.

Matt Sigelman: I think it requires another cultural change as well, which is a belief in your workforce and in people's future potential, because it's easy to look at somebody who's doing a job and say, "She's doing this job and that's who she is." But the ability to be able to take a more skill based view of somebody is very liberating, because it says, "Hey, look, there's a broader potential here that each of my workers has, and how do we unleash that potential? How do we invest in it?"

Adel Nehme: It's much more humanizing inclusive in my opinion, because it creates a much stronger and more engaged culture.

Matt Sigelman: Exactly. And I think the good news is I think we've got a golden moment right now, there's two key imperatives that are going to drive companies to rethink their talent and to rethink how they invest in skills. One is again, talent shortage, if I can't find the skills that I need and it brings the idea of talent management from being number six on my top five list, being a number one, two or three issue because I literally am, and you see this all the time today, a lot of companies that are reporting earnings misses because they can't actually produce. And so all of a sudden it's like, where do I find the talent? I can't buy it. How do I build it? How do I build it up from within? And so that's going to be one of the things that changes the culture, the other golden moment we have is around equity, because I think companies around the world are increasingly aware that they need to build more inclusive workforces.

Matt Sigelman: And again, if you take a zero sum game mindset, that is to say that there's a finite amount of diverse talent, and all I can do is just try to compete more aggressively for it, then you're going to wind up feeling pretty stuck. But if instead you say, "Hey, wait a second, most companies are more diverse at the bottom then the top, most companies have more women at the bottom than the top." And you say, "Hey, wait a second, how do we create those skill pathways that unlock the power of that talent?" You can wind up with a not only an organization that can find the talent it needs, but it can build the equity that it wants to display.

Adel Nehme: 100%, I couldn't agree more. And I really believe in the power of learning and giving people the opportunity to realize their full potential, as opposed to just filling a quota, and that's where the culture transformation is in that sense. The final question, we spoke about organizations, educational institutions, but how should individuals looking to find jobs that are meaningful and provide upward mobility, how should they approach their own career growth?

Matt Sigelman: So I think this goes back to this notion of being data driven and managing your own career. So you need to develop essentially a ways map for your career. There is no app like that out there, which means that you need to actually go out and do the underlying data science to be able to say, "Where is opportunity, where is opportunity today, but more importantly, where is opportunity going to be?" Look, data science is in no small part about building predictive models. So build a predictive model about your career, where is the ball rolling on the field and run there. And then the other advice I'd offer is when you start to think in terms of skills, not jobs, then it makes those trends, it makes running that distance much more achievable, because you can start to think about how you can construct a transition or a set of transitions skill by skill, instead of feeling, wow, how do I leap tall buildings in a single bound? Well, you don't need to.

Call to Action

Adel Nehme: That is awesome. Finally, Matt, do you have any final call to action before we wrap up today's episode?

Matt Sigelman: So look, we've talked a lot about this idea that the future is about skills, not job titles, probably not even jobs. And again, I think that's a really liberating idea, because it means you can control, you can take control of your own career and you can build your own destiny. The world is changing very fast and the ability to be able to adapt to it, not only adapt to it, but to get in front of it, is going to put you in the driver's seat in your career. Now none of this is new, if you haven't seen the movie Hidden Figures, I really recommend it. And for those who haven't seen it, if you rewind the tape by 60 years to the age of the Apollo mission and the like, a computer didn't mean a thing that was on your desk, computer was a person who was doing computational math at scale.

Matt Sigelman: And I bring that up here be because it says that skills have always been changing, the pace of that may have increased. But it's an inspiring movie because the women whom the movie portrays, could have found themselves displaced, and instead, they reinvented themselves, they acquired new skills and they kept ahead of the market and wound up having tremendous achievements. And I think I would put that in front of each of us.

Adel Nehme: Thank you so much for this Matt, and thank you for coming on DataFramed.

Matt Sigelman: So enjoyed this.

Adel Nehme: Likewise.