Decoding Data-Driven Product Development: All about my Thesis

"So what made you choose this topic?" my thesis advisor asked during our first meeting. 🤔

I smiled, thinking back to my days at AI Research Centre and Analyco. "Because I've lived the challenge of turning data into actual product decisions." 💡

That's how my journey into exploring data-driven product development began at Montpellier Business School. After experiencing firsthand how companies struggle to effectively use data in their product development processes, I knew I had to dig deeper into this puzzle. 🧩

The Quest 🔍

My research focuses on how leading companies like BackMarket, Alan, Dataiku, Stripe, and Qonto leverage data to build better products. But this isn't just another academic study – it's about understanding the real stories behind successful product development.

You know what's fascinating? "These companies all have mountains of data, but the real challenge isn't collecting it – it's knowing which signals actually matter for product decisions." 📊

The Investigation 🕵️‍♂️

The interviews revealed fascinating insights I never expected. At BackMarket, they discovered that their most valuable data point wasn't customer complaints, but the time customers spent reading product descriptions. "When description-reading time dropped," their PM shared, "returns went up almost immediately. That single metric revolutionized how we approach product listings." 📱

Stripe's approach particularly surprised me. Their PM explained how they use what they call "developer friction signals" - tracking where developers pause or backtrack in their documentation. "Each hesitation is a potential product improvement," they explained. That completely changed my understanding of user behavior data. 💻

The Breakthrough Moments 💥

I had surrounded by interview transcripts, everything clicked. I noticed a pattern: the most successful companies weren't just collecting more data - they were collecting different data. Alan, for instance, tracks not just what features users click, but what features they almost click and then don't. These "hesitation moments," as they call them, often reveal more than actual usage data. 🤯

The Cultural Dimension 🇫🇷

There's something uniquely French about how these companies approach data. It's not just about efficiency - there's an emphasis on "le pourquoi" (the why) behind the numbers. As one French PM told me, with a characteristic shrug, "Les données sont importantes, mais l'intuition aussi" (Data is important, but so is intuition). 🥐

The Learning Curve 📈

Working on this thesis has transformed how I think about product development. It's shown me that successful products aren't just built on good data – they're built on knowing how to ask the right questions of that data.

Between analyzing interviews, poring over research papers, and connecting patterns across companies, I've gained insights that I wish I'd had when starting my product journey. The most valuable lesson? Data should inform decisions, not make them. 🧠

The Real Discovery 🔬

Perhaps the most surprising finding came from Dataiku. Their team showed me how they once completely reversed a major product decision despite positive usage data. Why? Because they noticed something subtle: while users were engaging more with the new feature, they were spending less time on core workflows. "Sometimes the data whispers what users are shouting," their PM noted. That insight fundamentally changed how I view product analytics. 🎯

Looking Ahead 🚀

As I work towards completing my thesis, I'm excited about its potential impact. The frameworks and insights we're developing could help companies, especially smaller ones, better navigate their product development journeys.

For fellow product enthusiasts and researchers: The future of product development lies not in having more data, but in being smarter about how we use it. As one interviewee put it: "The best product decisions happen when data meets intuition in just the right way." 🌟

P.S. A special thanks to my advisor, Yi-Ting CHEN, whose guidance has helped shape this research journey, and to all the product managers who've shared their invaluable insights and experiences. 🙏