Data Science is Dying

If you studied data science expecting it to open doors to automatic 6-figure jobs, but instead you feel like the job market slammed shut, pay attention, because in this video I'm going to walk you through the shift most people still have not caught up to yet, what it means for your career, and how to think about your skills in a way that gives you a real advantage gain. Now, if you're a subscriber, welcome back, but if this is your first time here, I'm Kedeisha, and 6 years ago I was delivering pizzas and Amazon packages for $8 an hour. Today I am running a community of over 80,000 current and aspiring data professionals, and I've helped thousands of other career changers break into data analytics without going back to school and wasting money on courses that just don't work. And if you ever want to work with me to land your first job in data analytics, click the link in the description. All right, so let me take you back to where this whole thing started, because the origin story explains everything that's happening right now. In 2012, the Harvard Business Review published an article calling data science the sexiest job of the 21st century, and at the time it was true. The term had only been coined a few years earlier by DJ Patil and Jeff Hamburger, the guys who built the data teams for LinkedIn and Facebook. These were elite roles at a handful of companies with actual data infrastructure. Glassdoor ranked data scientists the number one job in America for 4 years straight. Salaries were climbing past $150K, and the demand was real because there were maybe a few thousand people in the world who could do this work. Then universities saw the dollar signs, and within a few years, over 500 schools launched dedicated data science programs. And the message to students were the same: learn machine learning, learn Python, learn neural networks, and companies will fight for you. Tuition for these programs? $40,000 to $80,000 Some of them cranking out graduates in 12 months with zero industry partnerships, zero career coaching, and zero connections to what companies were actually hiring for. But here's what was happening underneath the surface that nobody wanted to talk about. A lot of these companies weren't hiring data scientists because they had the data infrastructure to actually support them. They were hiring because they read the same Harvard Business Review article and thought, we need one of those. So you'd have a data scientist with a PhD getting hired at a mid-sized company that didn't even have clean data pipelines, no data warehouses, no culture of data-driven decision-making, just a title and a salary and no way to actually create impact. That disconnect between hype and reality, that's the crack in the foundation that everything else collapsed from. So starting around 2022, 3 things happened at the same time that cracked this whole thing open. First, AI tools got genuinely good. The kind of work that used to require PhD-level data scientists building recommendation systems, training models, running complex analysis, suddenly you can do a significant chunk of it with AutoML platforms, pre-trained models, and eventually tools powered by large language models. I'm not saying AI replaced data scientists, but it absolutely replaced the kind of rote entry-level model building that a lot of junior DS roles were based on. Second, the post-pandemic tech correction hit. Over 400,000 tech workers were laid off in 2023 and 2024 alone. Amazon cut 27,000, Meta cut 21,000, Intel 15,000. And the people who got hit hardest weren't the senior-level folks who had built relationships and proven their value over the years. It was the junior hires, the recent grads, the folks who were brought on during the hiring boom and hadn't yet demonstrated business impact yet. And third, and this one really changed the game, companies started unbundling the data scientist role. What I mean by that is they'd look at the $200,000 data science manager position and realize that person was probably spending 20% of their time on dashboards or reporting or stakeholder meetings and think, why are we paying this huge chunk of salary for this one role? So they'd split that one $200K role into a $120K senior analyst role for business-facing work and use AI tools to probably handle the modeling. More efficiency, lower cost, and same output. FANG data science openings dropped 89% year over year by mid-2023, and Amazon's own layoff data showed that only 2.7% of people let go even held the data scientist title, which tells you something. The title was already shrinking before the layoffs even started. So if you got a data science degree and you're struggling right now, it's not because you're not smart enough. It's not because the market is broken. It's because the market evolved and you're still applying to the version of the role that existed in 2018. The job that you were trained for, the job that's available now are two different The question isn't, will I get hired? It's, am I even looking in the right place? Quick pause. Now, if you want to land your first job in data analytics and avoid the lengthy mistakes that I did, click the link in the pinned comment where you can find the exact steps that my students use to land their job in data analytics without going back to school. All right, now let's get back to it. So here's a part that's gonna change how you think about your job search entirely. Everyone's out here searching for data scientists on LinkedIn and panicking when they see fewer postings, but what they're not seeing is that the money didn't leave data, it just shifted. What I mean by that is companies stopped creating standalone data science departments and started embedding data people directly inside their revenue teams, their product teams, their revenue teams, and so forth. And those roles, they may not always say data scientist anywhere in the title. Let me give you a concrete example. A company like HubSpot may not post a job that says data scientist marketing department. They may post growth analytics lead. The job description: build predictive models for customer acquisition, design experiments to optimize conversion, translate data into revenue strategy. That's really a data science job wearing a business suit and pays $160K plus. Same thing in finance. Goldman Sachs doesn't need another PhD running models in isolation. They may need a quantitative strategy associate who can sit in a meeting with portfolio managers and explain what the model actually means for their trading decisions. That role may not have existed 5 years ago, but now it's one of their fastest growing titles. This is what I like to call the invisible data economy, and the Bureau of Labor Statistics can't even track it properly because these roles can be classified under management analysts or operations research or marketing manager, but when you read the actual job descriptions, 80% of the work is data science or data analytics. McKinsey even estimated that the demand for analytical talent would outstrip supply by 50 to 60%. The demand did not evaporate. It just stopped wearing the same lab coat and started showing up in P&L statements. So if you're only looking at jobs that specifically say data scientist in the title, you're fishing in a pond when there's a huge ocean right next to you. All right. So if you've got a data science background, whether that's a degree, a bootcamp, or self-taught, and you're struggling right now, I need you to hear this. You don't have a skills problem. You just have a packaging problem. Think about it like this. If you spent years learning statistics, Python, machine learning, experimental design, those are premium skills. Companies pay well for that. But if you're walking in job interviews leading with, I can build a random forest classifier, and the hiring manager's eyes glaze over because they don't care about your model, they care about Q3 revenue targets. That is where the money is hiding. And closing that gap is what I call skills arbitrage. And here's how it works in 3 moves. Move number 1, flip your resume upside down. Most data science resumes lead with tools and techniques. It's like a chef walking into a restaurant interview and listing every ingredient that they know how to use. Newsflash, nobody cares about your ingredients. They just wanna know that you can cook. Flip it. Lead with business outcomes, then mention the method as context. Instead of built a churn prediction model using logistic regression with 89% accuracy, say identified $2.3 million in at-risk revenue and designed an intervention strategy that reduced customer churn by 14%. Same project, but completely different signal to a hiring manager. Move number two: learn the language of the buyer. Here's something that they don't teach you in any data science program. The people who decide whether you get hired—the VP of marketing, the director of operations, the CFO—they don't speak your language. They don't think in p-values and feature importance. They think in revenue, cost, margin, retention, and risk. So the fastest way to make yourself more hireable isn't another algorithm; it's learning how a business actually makes money. Pick one domain: healthcare, fintech. E-commerce, SaaS, logistics, supply chain, whatever it is, then learn the 5 to 10 metrics that domain lives and dies from. In SaaS, that can be monthly recurring revenue, churn rate, customer acquisition costs, lifetime value, and net revenue retention. In e-commerce, it could be average order value, conversion rate, cart abandonment rate, and return rate. When you can walk into an interview and say, I noticed your company's net revenue retention dropped 3 points last quarter based on our earnings call, here's how I'd use cohort analysis to diagnose why, you're no longer a data scientist anymore. You are a strategic hire, and strategic hires don't compete on salary, they negotiate. All right, move number 3, target the role, not the title. Stop typing in data scientist in every search bar and start searching for the work that you actually want to do. If you go to LinkedIn or Indeed and search by skill keywords instead of just job titles, try Python SQL business strategy, or statistical analysis, or stakeholder reporting, or analytical forecasting revenue. You'll be shocked at what comes up. You'll see growth analyst, strategy associate, business intelligence lead, revenue operations manager, Decision scientists. These roles use 90% of your existing toolkit, but because they don't have data scientist stamped on them, most people with your background will just scroll right past them. And that's exactly why the competition for these roles is a fraction of what a traditional DS posting is. And think about what that means. You spend all this time building rare, valuable skills. It's that you've been selling steak to vegetarians. Find the buyer who's hungry for exactly what you've got. All right, so I'll leave you with this: data science isn't necessarily dying, but the version that was sold to you, that is dead and long gone. And that's actually good news if you ask me, because the people trying to figure out how to stop competing for scraps and start capturing the roles that are actually paying more— now you know where the money went, and now you know how to repackage what you already have. The only question is, what's your actual step-by-step plan to get there? That's exactly what I cover in this next video, a complete roadmap to go from where you are now to a high-paying role in data analytics. See you there.