In this YouTube monologue, Kedeisha Bryan argues that the traditional focus on technical data‑analytics skills—coding, advanced libraries, and AI tools—is becoming a commodity as AI accelerates development speed (e.g., 55% faster coding). She cites research (Microsoft, GitHub) and industry data (87% of data‑science projects never reach production) to show that the real hiring differentiator is the "last mile": the ability to translate data insights into clear business decisions. Bryan introduces a practical "ROI" storytelling framework—Relevance, Observation, Impact—to help analysts craft concise, action‑oriented visualizations and presentations. She emphasizes three questions for every project (business question, key insight, actionable recommendation), advises trimming visual clutter, and insists on titles that state business value. The talk differentiates learning‑oriented projects from showcase projects that demonstrate real‑world problem solving. Bryan shares a personal narrative of moving from low‑wage delivery work to six‑figure data‑analytics roles by mastering communication, stakeholder empathy, and personal branding. She concludes by teasing a follow‑up video on building a recruiter‑friendly personal brand.
Chapters
You're obsessing over the wrong skills. LLMs, AI agents, complex Python libraries. You're grinding through tutorials thinking this is the way to get hired in data analytics. But here's the reality. Research from Microsoft and GitHub shows AI can now help us code 55% faster. So if coding is becoming faster and cheaper, our technical skills are now becoming a commodity and the market is flooded with people who can write code but starving for people who can actually explain it. I made the same mistake when I first broke into data analytics, spending months focusing the wrong things, and I see the same pattern every day in people that I coach. In this video, I'm going to show you why 87% of data projects fail and why it has nothing to do with code. And then I'm going to give you a simple framework that's going to make you more valuable than 90% of candidates applying for the same roles today. If you're a career changer, you cannot afford to spend more months learning skills that companies are not even testing for. And if you stick around to the end, I'll show you the one shift in how you present your work that will instantly separate you from everyone else. So according to the data, 87% of data science projects never make it to production. Read that again, 87%. These projects do not fail because the math is wrong. They don't fail because someone wrote bad Python. They fail because of what Gartner calls the last mile problem, the gap between insights and decisions. Most people are optimizing for what I call the first mile. That's all the code, that's all the queries, the technical execution. And yes, you need that foundation, but here's what no one else is telling you in these YouTube tutorials. The first mile may get you in the Think about it from a hiring manager's perspective. They're not thinking, I really hope this candidate knows how to write a window function. They're thinking, can this person help me make better decisions with data? Can they take our messy data and turn it into something my VP can actually act on? That is the last mile, and almost nobody is good at it. Most aspiring analysts never learn this, and they stay stuck in that first mile and wondering why they cannot get hired. So if technical skills alone won't get you hired, what will? First, you need to stop thinking that technical skills alone will get you hired. Thinking like a student and start thinking like a consultant. A student will ask, what tools should I learn next? But a consultant will ask, what problem am I solving and how do I communicate the solution? This shift separates people who get hired from those who stay stuck in the tutorial loop. When you're working on a project, whether it's your portfolio or on the job, you need to think of 3 things. First, what is the business question? Not the data question, but the business question. What decision is someone trying to make? Second, what is the insight? Once you've done your analysis, what is the one thing that matters the most? Not everything that you found, but the one thing. Third, what is the action? Based on your insight, what should someone actually do? If your analysis doesn't lead to a clear recommendation, it's just noise. This is the last mile. It's not about the code. It is all about the communication. It is about taking complex information and making it so clear that a busy executive can understand it in seconds and make a clear decision. This skill is rare, and most people in data can pull numbers, but but very few can explain why those numbers matter. And that's your competitive advantage. So I'm gonna give you something practical, a framework that you can start using today. I call it the ROI framework, a method for data storytelling. You can probably think of ROI meaning return on investment, but I'm gonna rebrand it today. R meaning relevance, O meaning observation, and I for impact. So relevance. Before you create a single chart, ask yourself, does this answer the exact business question that was asked? Not a related question, not an interesting tangent that you discovered, but the exact question. I see this mistake constantly. Someone gets a dataset, they start exploring, they find something interesting and build a whole dashboard around that. But that interesting thing has nothing to do with the stakeholder actually needed in the first place. So here's my rule. If a chart doesn't directly answer a business question, delete it. No interesting side quests, just stay focused. Now write the business question at the top of your notebook before you touch any data. Every chart, every insight, every recommendation should trace back to that question. If it doesn't connect, it just doesn't belong. Okay, O for observation. Here's a test for every visualization that you create. Can a stranger find the key insight in less than 5 seconds? If they can't, you failed. Most dashboards I see are cluttered disasters. 10 charts crammed into one page, rainbow colors everywhere, no clear message, and just data dumped onto a screen. That's not analysis, that is just pure noise. My rule is to remove 80% of the ink. Get rid of the gridlines, get rid of the unnecessary labels, get rid of the legend that's taking up half the screen. The goal is clarity, not complexity. You want want someone to glance at your work and immediately know what matters. I had a student that came to me with a dashboard of like 15 different charts. She was proud of the information that was so packed in, but I asked her, what's the one thing you want someone to take away from this? And she couldn't answer. So we rebuilt it all with 3 charts, clear titles, strategic color, and one message per visual. She used that dashboard on her portfolio and it landed her interviews within the next couple of months. Less is more, always. Okay, I for impact. So this is where most people completely drop the ball. Your chart title should always state the financial or business value, not just describe the data. A bad title is sales by region, but a good title could be the West region is down about 50%. The first title tells me what I'm looking at, but the second title tells me what I need to do. You're not just about presenting data, you're driving decisions, and your title should be action-oriented. They should tell a story before someone even looks at the numbers. When you do this, stakeholders stop asking, what does this mean? And they start asking, what should we do about this? And that's when you know you've nailed the last mile. Okay, so now you've got the framework, the ROI, relevance, observation, impact. So how do you actually develop this skill? Stop building projects for learning and start building projects for showcasing. Projects for learning are where you can follow tutorials, use clean datasets, and practice some syntax. That's fine when you're starting out. But projects for showcasing need to demonstrate that you can think like a business person, that you can take messy data, real-world data, and turn it into something that drives decisions. So here's a practical exercise. Take any dashboard that you've built and apply the ROI method. Relevance. What is a business question that this answers? Write this down, and if you can't articulate it clearly, it's a problem. Observation. Can someone look at my dashboard and find the insights in under 5 seconds? Show it to a friend who knows nothing about data. Time them. If they're confused, simplify from there. Impact. Does your title state the value? Rewrite every title so it's action-oriented. Sales by region will then become West region underperforming by 15%. Do this for every project in your portfolio, especially your older ones. It will transform how hiring managers perceive your work. And don't just build dashboards, build slide decks too. As data analysts, we don't just build dashboards anymore. We make presentations. The best presentations follow the same ROI principles. One insight per slide, action titles at the top, and clear recommendations. This will separate someone who actually gets hired in data analytics versus those who get ghosted. It's not about knowing more tools, it's about communicating better. All right, so let's bring it home. The job market has changed. AI is making technical skills more accessible every day, and the barrier to writing code is dropping fast. But you know what AI can't do? It cannot understand the company's specific context. It can't sit in a meeting and read the room, and it cannot take a confusing dataset and figure out the question that actually needs to be answered. And it can't build relationships with stakeholders who trust your judgment. This is a human element, and that's part of the last mile, and that is what companies are desperate for in 2026. When I was delivering pizzas and Amazon packages for $8 an hour, I didn't have connections, I didn't have that fancy degree, and the odds were all stacked against me. But I focused on all the right things. I learned how to communicate insights clearly, I built projects that solved real problems, and I positioned myself as someone who could deliver value from day one. And within a few years, I went from making $8 an hour to making several 6 figures in data analytics. The path is there, but you need to stop chasing those shiny technical skills and start developing the boring ones that actually matter. Data storytelling isn't exciting. I know it doesn't get many views on YouTube. Nobody's making viral content on how to write better chart titles. But this is a skill that's going to get you hired. It is a skill that gets you promoted. And this is a skill that will make you irreplaceable in the age of AI. So now that you understand why communication matters more than code, there's another piece that people get wrong. They build these skills in isolation, but they never know how to package themselves well for the job market. Your LinkedIn, your resume, your portfolio— they all need to tell the same story. So in my next video, I break down exactly how to build a personal brand that attracts recruiters instead of repelling them, and I show you the exact 3 brand assets that you need to have in this new job market. I'll see you there.
