Bold Brands

Tuani Stelter

Social Media Specialist

Dr. Oetker

Winner 2026

Tuani Stelter

Let’s start simple. Who are you, and what do you do with social data that others might not expect?

My name is Tuani Stelter, and I’m a Brazilian living in Germany. I’ve been working at Dr. Oetker for almost 6 years. One of the things I do with social data is trying to inform and inspire our local teams and support them with insights for content creation. A lot of our researches are related to flavor and ingredients  trends that can inspire recipe content for social media and ultimately product insights. 

What’s something in our industry we pretend to understand, but don’t?

I think the concept of “trend monitoring” or “foresight” is often more complex than we admit. Sometimes, what we call trend prediction feels like a self-fulfilling prophecy. A topic gets claimed as the next big thing, and suddenly the entire industry rushes to embrace it. And then, suddenly, everyone is talking about it so it actually becomes a trending topic, but not necessarily because it was inevitable. But, in the end, I’m not sure if we are really being able to correctly predict what’s coming, or if we’re actually influencing the conversations with our hypothesis and assumptions of what the next big thing is.

What’s a moment this year where social data helped your team do something bolder, faster, or better?

This year, we picked up the noise of the latest Taylor Swift album announcement early on and were able to share it with local teams very quickly. This small signal inspired different ideas, from Mexico to Germany. In just a couple weeks, the team had the idea and developed “The Life Of a Showgirl” inspired sprinkles, which had an outstanding sales performance following the album release. It was definitely a standout case for using social data within the organization, proving that we are able to catch and react to trends quickly regardless of the usual constraints in a corporate environment.

What’s one belief about your audience that social data completely upended for your teams?

We had a product launch that was perceived internally as a total failure. People thought the negative feedback was overwhelming and word spread around the hallways that it was a catastrophe. So I decided to dig into the data and see if that was really the case. And then I realized that the reality was more complex than that: yes, sentiment was slightly more negative than usual, but there was also a significant amount of positive feedback that had gone unnoticed, as we sometimes tend to focus only on the negative. In the end, the teams were positively surprised which helped shift the narrative internally and reminded us once again that perception can be misleading without data to validate it.

If you could build your dream social intelligence team from scratch, with no legacy and no limits, what roles would you include?

I’d start with data analysts who really “get” social, someone with a very strong analytical abilities but that also has a sense of how those insights can be turned into something (content, campaigns, products) that resonates with the audience. A chronically online content strategist that knows about everything that’s going on around every corner of the internet, and can easily spot emerging conversations and turn them into creative opportunities. And finally, an AI/automation expert to make reporting and predictive analysis faster and smarter. 

When do you feel like you’re doing your best work?

When I have the time to dig deep, refine the research, test and learn. I like feeling like I’m delivering something that feels thoughtful and valuable and is not just ticking a box. The best part is seeing those insights come to life, whether it’s a piece of content, a product idea, or simply helping someone make a better decision. 

What’s your browser history giving away about you this week?

A lot of tabs on trend reports and predictions for 2026… and a few shopping carts waiting for checkout. A little work, a little play because balance matters :D 

Get Social with SILab