I did not plan to work in data.

When I started my Business International bachelor at Inseec, I genuinely had no idea what direction I wanted to go. So I tried everything the school offered — a specialisation in luxury marketing, then entrepreneurship, then international trade. Three different tracks in three years. Not because I was indecisive, but because I was curious about everything and committed to nothing.

That is probably a familiar feeling if you went to a business school.

The internship that changed things

In my third year I landed a Product Manager Junior internship at iExec, a Web3 startup based in Lyon. I did not know much about blockchain at the time — I had joined Kryptosphère, the largest student crypto association in Europe, mostly out of curiosity. But the internship was the real turning point.

My job was to do market research on the Web3 ecosystem — DeFi, DePIN, RWA. To do that properly I started using Dune Analytics, a tool that lets you query on-chain data directly. I was writing SQL queries, analysing market cap trends, tracking protocol metrics. And somewhere in the middle of that, something clicked.

I genuinely enjoyed making sense of messy, complex data. Finding patterns. Drawing conclusions from numbers that most people would just scroll past.

I also ran more than 15 developer interviews in French and English, contributed to a product workshop with the CEO, and came third in an internal hackathon — building a Chrome extension prototype in two days with a team I had just met. It was the most energised I had felt in three years of studying.

Choosing data over the obvious path

When I came back from iExec, Inseec had just launched a Master’s in Data Science. I applied immediately.

It was not the obvious move for someone coming out of a business bachelor. Most of my peers were going into marketing, sales or consulting roles. The Master’s was more technical, more uncertain, and required building skills I did not have yet.

But the iExec internship had shown me something important: I did not want to just use data to make PowerPoint slides. I wanted to understand how it worked, how to collect it properly, how to make it mean something.

Two years of learning what data actually looks like in practice

While doing my Master’s I started a two-year alternance at Cheil France as a Digital Data Analyst, working on the Samsung account.

That experience taught me things no course could have. I managed three strategic dashboards in complete autonomy, built a real-time product launch tracking system used by the team on launch days, and audited more than 100 tags via Adobe Launch. I also presented our tools and methodology at a Cheil x Samsung workshop in front of 50 people — including a live demo of Adobe Analytics.

The biggest lesson was not technical. It was that data quality is an organisational problem, not a tool problem. Getting the right numbers means coordinating with developers, product teams, and clients — and being rigorous enough that people trust what you show them.

What I would tell someone in the same position

If you are in a business school and wondering whether data is for you, here is what I wish someone had told me:

You do not need a computer science background. What you need is genuine curiosity about why things work the way they do, and the patience to dig until you understand them.

The business background is actually an advantage. Most data analysts can build a dashboard. Fewer can explain to a client why the numbers matter and what to do about them.

Start with something real. I did not learn data from a course — I learned it from trying to understand the Web3 market during a real internship, with real stakes. Find a dataset you actually care about and start asking questions.

And if you are not sure what you want to do: try things. Not because switching direction is a sign of failure, but because you cannot know what clicks until you try it.

I am still building. I finish my Master’s in mid-2026 and I am actively looking for a data analyst role in Geneva. But looking back, the non-linear path was not a problem to solve. It was the point.