Crush Data Coding Interviews: A Step-by-Step Guide with the ADDA Framework

And 3 proven strategies when you're stuck

Let me guess—you’re feeling the heat of coding interviews.

You’ve probably taken a quick 2-week course online, practiced SQL on sqlzoo, and can code comfortably when it’s just you, the screen, and silence. But when it’s time for the interview, everything changes.

Your mind blanks out on functions you use daily. Your brain freezes. Your fingers feel icy, and the sweat starts dripping onto the keyboard.

Sound familiar? You’re not alone. This happens to the best of us, whether you’re working with SQL, Python, or R.

Here’s the thing: coding interviews are a different beast. It’s not just about knowing how to code; it’s about performing under pressure, with a ticking clock and someone watching your every move.

Interviews have their own set of rules.

And if you want to win, you’ve got to learn how to play by those rules.

That’s where I come in. In this guide, I’ll walk you through:

What the heck do tech screens really look for

The mindset you need to crush your next interview

The ADDA framework—my secret sauce I developed over 5 years. It helped me pass 10+ tech screens, including Meta, Spotify, and Nextdoor

3 foolproof strategies for when you get stuck

3 common traps to avoid

4 ways to speed up your coding

By the time you finish this post, following the ADDA framework and my strategies will make success almost inevitable.

The unwritten rules of the game you can’t ignore

What the heck do tech screens look for?

Coding skills, of course.

But there’s more to it than just writing code. It’s also about commercial acumen and communication

I’d break it down like this: 50% coding, 25% commercial acumen, and 25% communication skills.

Let me unpack that for you.

1. Coding: Can you write clean, functional code with minimal errors? This part is pretty straightforward.

2. Commercial acumen: Can you take a business problem and turn it into code? Are you able to think through edge cases and ensure your code reflects them?

3. Communication: Can you clearly explain your thought process? Can you help others understand your code, seek guidance when needed, and take feedback like a champ?

While most people focus hard on coding (and that’s important), they often overlook the other two. And that’s where things go sideways in interviews.

Imagine this—you’ve put in the hours, completed intense courses, practiced coding till your fingers hurt, and Googled “window function” more times than you can count. But then you stumble in the interview because you didn’t nail the “non-coding” bits. Frustrating, right?

But don’t worry—we’re going to fix that.

In the rest of this article, we’ll dive into how to sharpen your commercial acumen and communication skills, giving you practical tips you can use in your next coding interview.

Now, this article isn’t about improving your coding skills. If you need help with that, check out 7 Essential Skills for Landing Your First Data Science Job in 2024 - it’s loaded with key concepts and learning resources.

Still, I’ll touch on some common coding mistakes later in this post in "common pitfalls". 😉

To master commercial acumen and communication, I’ll walk you through a framework I developed that helped me pass over 10 coding interviews, including at places like Meta, Spotify, and Nextdoor.

Plus, as someone who’s tech-screened more than 10 candidates, I’ve seen firsthand how this framework can make you shine—or how a lack of structure can trip you up.

Ready to dive in? Let’s get to it.

---

First, we need to set your mindset straight

My manager once told me something that stuck:

Companies don't need order-takers; they need opinion-shapers and result-drivers.

That’s the essence of a “thought partner” mindset.

So, what does this mean for you as a data scientist? It’s about more than just crunching numbers. Your job is to create visibility and drive business decisions using data.

“Creating visibility” means you need sharp technical skills to break down complex problems and visualize them clearly. But “driving business decisions” takes it a step further—it’s about being proactive, influencing strategy, and guiding teams toward the best outcomes.

Great data scientists don't just analyze—they partner. They articulate ideas, embrace ambiguity, tackle problems with structured frameworks, and most importantly, they get things done. 

Imagine working with someone who doesn’t just follow instructions but instead, lights up the room with fresh ideas and delivers impactful results.

That’s the person you want to be in your interviews.

So, take charge. Lead the conversation, seek feedback, and steer the process.

I’ll show you exactly how to do that.

---

The ADDA framework: Position yourself as a thought partner, not just a code monkey

Let me introduce you to the ADDA framework—your secret weapon for acing coding interviews and proving you’re more than just a coder.

Here’s how it breaks down:

1. Ask: Start by asking questions about the data, the business context, and any terms that need clarification. Don’t hold back—this is your chance to dig deep. Interviewers appreciate and reward candidates who ask insightful questions.

2. Define: Next, define the business question. This might mean repeating or rephrasing what the interviewer asked. It’s not just about parroting back their words—it shows you’ve listened, understood, and are on the same page. Plus, people like hearing their words reflected back at them. It’s a subtle way to get inside the interviewer’s head and build rapport.

3. Down: Before jumping into your response, jot down your answer structure. I’m a fan of pen and paper—it helps shift your focus and create a mental break. But if that’s not your style, take a pause, type it out on a notepad, or at the very least, mentally map out your approach. As an interviewer, I love when candidates pause to organize their thoughts. It signals that they prioritize clear, structured communication.

4. Announce: Finally, announce your structure and dive into your answer. Let the interviewer know what’s coming next. This not only frames the conversation but also drives the problem-solving process. You won’t leave them guessing or asking, “What else?”—you’ve already laid it all out.

---

Let’s see how this plays out with a real example:

Color code:

Interviewer: Here is a table with below columns. On average, how many sessions did a person have yesterday?