“No More Flying Blind Across Channels”—with Julia Schrobenhauser and Sabrina Oehme (Kloster Kitchen)

Recorded in German—English translation of the original conversation.

Intro

Maria-Lena Matysik: My guests today are Julia Schrobenhauser and Sabrina Oehme from Kloster Kitchen.

You probably know Kloster Kitchen from the supermarket shelf or the online shop—they’re the pioneers who, around ten years ago, invented the ginger shot category in the first place and made it big. Today they’re a true multichannel brand in the world of functional beverages—and, as of recently, part of the Rauch Group.

But back to my guests: At Kloster Kitchen, Julia is responsible for marketing strategies and activations across online and offline. As Head of Brand Design, E-Commerce and Online Growth, Sabrina takes care of the entire digital presence—and also likes to share her hands-on knowledge of omnichannel on industry stages such as CMCX.

When the two of them took over team leadership together almost a year ago, they faced a challenge that many of you probably know: a very mixed team, an extremely large number of channels—from Meta and Google to Amazon and trade marketing—but absolutely no consistent numbers. Every channel tool made its own successes look better than they were, and in the end, marketing was flying blind.

In this episode, I talk with the two of them about how they put Kloster Kitchen’s marketing on a completely new, data-driven foundation. We talk about

• how Kloster Kitchen is getting marketing out of silo thinking,
• how they built a data culture through transparent communication and internal workshops,
• how the team agreed on a new KPI structure,
• how they cleaned up their tracking and even built their own reporting add-on with AI
• and, very concretely: why a clean data foundation is a lifesaver when a company goes through an insolvency and an acquisition and the marketing channels have to be completely rebuilt.

Here we go…

Interview

Maria-Lena Matysik: Julia, Sabrina, it’s a great pleasure to have you on Analytics Sprechstunde today. Thank you so much for being here. I’m really looking forward to this.

Julia Schrobenhauser: Yes, we’re excited too. Hi!

Sabrina Oehme: Yes, we’re excited too.

Maria-Lena Matysik: Very cool. Would you like to start by briefly introducing yourselves and describing what you do and what your role looks like?

Julia Schrobenhauser: Right, as you already said: We’re Julia and Sabrina from Kloster Kitchen. I’m Julia, and I’m responsible for marketing strategies and activations here, both online and offline. Kloster Kitchen became known for ginger shots—we founded this market about ten years ago. Today we operate in the world of functional beverages, and since early July we’ve been part of the Rauch Group.

Sabrina Oehme: I’m responsible for brand design, e-commerce and online growth. What makes Kloster Kitchen exciting: We’re a D2C brand that has been around for over ten years, and we come from a multichannel background. That means we’re active both online and offline. And as part of the Rauch Group, we’re now part of a very large FMCG corporation.

Maria-Lena Matysik: Cool, thank you both. Then I’d say let’s jump right in. If you look back a bit over the last, I don’t know, year, year and a half: How would you describe the way marketing at Kloster Kitchen worked with data back then? What did the situation look like, and where, from your perspective, were the biggest problems and challenges?

Julia Schrobenhauser: Sabrina and I took over the leadership almost a year ago. The first point was: How can we restructure the teams and capture the status quo—where do we stand right now? We quickly noticed that everyone was working with different numbers in their own area. Right, Sabrina?

Sabrina Oehme: Exactly. Back then there were eight of us on the team, really a completely mixed bunch. We covered everything online and offline: from online sales channels like Shopify and Amazon, through performance channels like Meta and Google, to CRM, influencers and trade marketing—there really was everything. So a very, very broad range of channels, all covered by a small team.

For us, it was always the case that I came more from the brand performance side, and Julia’s strength lies in marketing and activation. And when we took over the leadership, we immediately noticed that every area had different numbers in mind. There were no consistent KPIs and no consistent data foundation where you would have looked at two or three main KPIs. Nobody could compare their channels with one another. We couldn’t analyze at all which channel was actually performing well and which recommended actions we could give next—to see what fits the strategy and how we move forward.

Maria-Lena Matysik: So that means, if… sorry, did you want to add something?

Julia Schrobenhauser: Yes, we were kind of flying blind in that area.

Maria-Lena Matysik: Okay. Can I picture it like this: The person responsible for the Meta campaigns only looked in Meta, the person responsible for Google Ads only in Google Ads, and so on? So everyone optimized within their own area, but you lacked the overarching view to really steer the big picture?

Sabrina Oehme: Exactly, right. You probably know this yourself from lots of other areas: Meta naturally presents its KPIs in a way that makes them look most attractive for Meta—just like every other tool. And it was just the same for us. We did have a shared data foundation in one tool, but simply nobody worked with it. Instead, everyone still pulled the KPIs from their own channel tools.

Maria-Lena Matysik: So what did you do then? How did you get to the point of knowing: This is what our target state should look like, this is what we need now?

Julia Schrobenhauser: I think at the beginning, it was actually fairly easy for us: The target state has to be simple to start with. Everyone has to work with the same KPIs. We said: three main KPIs per channel—those were ROAS, CAC and, optionally, every now and then contribution margin—that everyone can rely on, where everyone uses the same terms and means the same thing. The target state simply has to be relatively concise, short and simple before we go any deeper.

Sabrina Oehme: Exactly. You can also think of it like this: At the end of the day, you always have targets or directions from company leadership on where the company should be heading. For us, profitable growth was clearly the top priority. And of course, the cost-to-revenue ratio and ROAS in particular are two very, very decisive KPIs there, as Julia just mentioned.

And of course, that also involves a lot of detail—maybe also looking at costs beyond these KPIs: What are you working with? We all know it from e-commerce: Marketing costs are rising enormously these days. So we really just took a look at everything we work with and which agencies we partner with. We also noticed that many people might not really know what to make of the term contribution margin—and I can say that about myself too. I think if you’re not that deep into the numbers and you work in other departments, you need a lot more clarity and explanation at the beginning, for example in workshops where you go through the topic together, so that everyone can participate and contribute.

Maria-Lena Matysik: Absolutely, I can really relate to that. Listening to this, I gather that action was needed in two areas: The first area is the whole topic of the data foundation, so more the technical side and the data itself. And the second area is the team, meaning the people who ultimately work with the data. Can you walk me through both of these directions—what you did there and what your path to a solution looked like?

Julia Schrobenhauser: I’ll start—I can maybe say something about the team. For us, it was actually the case—also as part of the restructuring we did almost a year ago—that we said: Okay, we need to rebuild the structures a bit, and this data mindset and the classic knowledge transfer between the areas need to be fostered more. We said that each team should first present its area internally, so that everyone gains an understanding of: What does one person need, where is the other coming from? We then set up our own workstreams. That was basically free training for everyone, and everyone responded well to it.

But of course you also have to say very clearly: When you change all these things and alter the usual processes, it means more effort at the beginning. And more effort always means change at the same time. At the beginning, of course, it felt like this for everyone: Okay, back to square one and doing things differently.

But what was actually received very well: Sabrina and I said we’d go all in on transparent communication, and we involved everyone right from step one. So we said: Okay, we’re not going to dictate the reports from the top down, but create and build them together, so that everyone can say: Okay, this is necessary for me and this is what I need. And then this data foundation was checked again in depth: Does it work for everyone? Are we all talking about the same thing? People then got on board, and it went in the right direction. But of course that also took energy and time, you have to say that very clearly.

Maria-Lena Matysik: How long did it take until you felt: Okay, maybe not everything in the team is perfect yet, but it’s already much better?

Julia Schrobenhauser: I’d say four to five months, easily. You really have to take the time for it, and the day-to-day work can’t fall by the wayside—that has to be kept up too. So it just takes a while. And even now, after a year, we’re much further along than a year ago, but we’re not done yet either.

Maria-Lena Matysik: There’s always something to optimize.

Sabrina Oehme: Absolutely. On the data side, maybe I can add—because we touched on it briefly at the beginning—what had to be optimized and tackled on top, again and again: namely in our shared tracking tool, Klar—in case you’ve heard of it…

Maria-Lena Matysik: Of course.

Sabrina Oehme: Great. That we cleaned up the data foundation there together and really challenged all the basics that had simply been entered there. Together with the tool and with AI, we built a small add-on that ultimately delivers consistent information to all teams on a monthly and weekly basis.

A very big pain point that we had the pleasure of discovering in the course of the switch was that our tracking backend didn’t work, or rather was totally outdated. It couldn’t deliver the right data to us at all. Those are all the little things you have to optimize on top of that. Like Julia says: Alongside day-to-day business, that also costs a lot of time and effort that you have to invest. But I’d say it was really worth it. The combination of teamwork, restructuring, training together and reviewing the data has, I think, moved all of us forward a bit.

Julia Schrobenhauser: Totally, I fully agree with that.

Maria-Lena Matysik: How would you describe the interaction between data and team? I could imagine: Before, everyone was working with the data, but nobody really knew who was looking at what. And then, step by step, there was a new, consistent data foundation for everyone. How did trust in the new data grow—how did that unfold?

Sabrina Oehme: I think at first, the classic thing happened: Change is always uncomfortable at first, and naturally you eye the whole thing a bit skeptically at first—I can totally understand that. We spent a really long time with the team working out what fits us and the company. And I think it happened relatively quickly that many people were very grateful for this consistent way of working together. Because there was much less individual stuff going on; instead, everyone really knew: This is what matters, this is important, this is where I look. And once the first positive results were visible, people recognized the point of it relatively quickly, and it was received positively.

Julia Schrobenhauser: Yes, totally. I actually think that because we were then able to make decisions faster—shifting budgets and things like that—everyone on the team saw: Okay, it was definitely worth sitting down again, looking at it and going over it together, because it’s simply more effective. As a result, in some areas that maybe aren’t going so well, there was no longer this pain along the lines of “Now we’re spending so much money and not seeing any success”—instead, you could clarify it right up front. And I think that simply creates motivation again, too.

Maria-Lena Matysik: Do you maybe have one or two more examples where you can see: This is how our processes and marketing decisions work today—and that wouldn’t have been possible before?

Sabrina Oehme: The example that immediately comes to mind always sounds super sad: We took costs into account much more intensively in our marketing KPIs. What does an agency cost compared to the performance the channel brings us? Meta is a super nice example: You can go incredibly deep there when it comes to the performance of individual ads. Simply putting what we invest in the individual tracking tools and agencies in relation to that—and we really did this for all channels—brought an incredible number of insights. Namely, whether we’re really set up correctly on the channel right now or whether we might need to switch tactics completely, because it was actually a typical money-in, money-out business: Whatever we spent, we earned back one for one. And then, of course, it’s not an efficient or profitable channel for us. That was, I think, one of the topics.

And as we already mentioned earlier: One of the best changes was really defining one dashboard per year together at the start of the year, based on the goals. That means we always looked at five KPIs, compared them with the channels every week and, at the beginning—even if it was a bit more intensive—made decisions together from week to week. Internally, that gave us a much faster and better feel for how things can and need to be managed.

Maria-Lena Matysik: I could also imagine: When the whole team sees, this is the data we’re working with now, and we’re all doing it together—and really takes that in—then it helps a lot, also for trust in the data.

Sabrina Oehme: I also wanted to say that it’s always super exciting, because it also gave us the opportunity to deliberately shift budgets to the channels that really performed well. And of course that also makes every individual on the team who’s responsible for it happy: You see that if I deliver good performance somewhere, then significantly more gets invested in that potential.

Julia Schrobenhauser: Exactly. And what we also did on top of that: Because you no longer had to invest so much time in your own numbers—you knew exactly that you could look it up in Klar and everything was set up—we also reduced meetings. Simply so that, in the end, everyone has more time for strategy and other topics. And then it’s also easier to say: That was the right direction to go in.

And what also comes into play nowadays, you have to say that too—and what we’ve intensified—is the use of AI. We use it to gain more freedom overall again to tackle other things. That also simply takes time at the beginning, but in the end we see that the investment has paid off. And that was also super important for the team.

Maria-Lena Matysik: By AI, do you mean, for example, the add-on you built? Or do individual team members work with AI, or rather with the data with the help of AI?

Julia Schrobenhauser: Both, I’d say—it all goes together. I think nowadays it just doesn’t work without AI anymore. Of course, you still have to make sure that everything is always accurate. But it makes things a lot easier, because you can set up many automations that would normally sometimes take several weeks. Now you have that within a few hours. It’s simply an enormous relief.

Maria-Lena Matysik: Maybe one last question to wrap up: From the whole process of the last few months—or almost years—have you taken away any lessons where you think: Others could learn from this so they don’t make the same mistakes again?

Julia Schrobenhauser: I actually think—and this is something we’ve also said—you should always check whether you’re really talking about a clean data foundation. The good thing is: Then the numbers don’t lie. And then, as we already said, you can set up future actions, strategies and things like that much more sensibly and, above all—which is very, very important these days—more profitably. That’s why I’d recommend that everyone still challenge things every now and then: Are we talking about the same thing? Does everything still fit? Simply to be sure that we have a clean data foundation.

Sabrina Oehme: Yes, absolutely. And maybe to add to that, even if it always sounds so easy-peasy: I think the most important thing is a clean structure and clean organization within the team and for the areas of responsibility. That is and remains the be-all and end-all, and it also leads to fewer misunderstandings in communication later on.

I think Julia and I really invested a lot of time over a year looking together with the team at: What works, how do we work, which processes and reports really fit us—so that later we can also set much clearer targets for the next milestones.

Maria-Lena Matysik: And if you look into the future now: Are there still topics where you say, we still want to tackle that, that can be done better?

Julia Schrobenhauser: To be honest, this year we also had to deal with an insolvency, and that’s how the acquisition by Rauch came about. That really threw a wrench in our plans, because at that point we had to set a lot of channels to zero and now have to rebuild them after a substantial decline. As a result, a lot of the internal workshops and workstreams we talked about earlier simply fell by the wayside—you have to say that very clearly.

And now, of course, there’s also a new direction, a new strategy. Now we have to approach all of that anew and say: Okay, over the next few months we simply have to set everything straight again and make sure we find our footing again. But I’m definitely very, very confident, because the work over the last year, the last year and a half, has meant that we’ll get back on track relatively quickly. So that’s pretty good. Right, Sabrina?

Sabrina Oehme: Yes, absolutely, I can only confirm that.

Maria-Lena Matysik: When you just said that, it was somehow my thought too: Once you’ve put the energy in and built a shared strategy, a shared picture and a shared data foundation, you have a really solid base that you can now build on again as you rebuild or expand the marketing channels.

Sabrina Oehme: Definitely. I think what’s still a big topic—and something we talk about again and again—is attribution across D2C customer behavior, so everything around customer lifetime value: How does a customer shop? I think there are more opportunities to extract knowledge there than we’re currently using. And I think we’ve created the perfect data foundation for that—so that we can really dig even deeper there and set up the entire online business even more strongly and even more on a per-customer basis. Because for us, of course, that’s the absolute USP: customer service and how you engage with the customer in communication and with offers. And I think that’s also a topic on a lot of people’s minds in the market right now, because of course it’s super exciting how overall shopping behavior is changing these days, also with AI.

Maria-Lena Matysik: Absolutely.

Julia Schrobenhauser: I think what’s also exciting for us there: As we said from the start, we’re dealing with multichannel. Accordingly, we have to look at it online, but of course also offline. It’s exciting to see how much that differs.

Maria-Lena Matysik: Thank you both so much! That was really exciting. Thank you for all the insights and for taking me through your last few months.

Outro

Maria-Lena Matysik: I found it super inspiring how honestly the two of them talked about the change process—including in the context of the insolvency and acquisition—and about the ups and downs of the last few months.

But if I had to pick out very concrete takeaways, they would be these:

1. Don’t trust any single platform’s version of the truth: Every tool—whether Meta, Google or TikTok—first calculates its own performance in the most flattering way. If you want to grow profitably, you need a neutral, consistent data foundation and few but clear KPIs. At Kloster Kitchen, those were mainly ROAS, customer acquisition costs and contribution margin.

2. Data culture is 80 percent change management: A new dashboard is useless if the team doesn’t understand the terms or doesn’t use the data. Julia and Sabrina didn’t dictate reports from the top down but built them together with the team. That took several months, but it creates real trust in the numbers.

3. Clean data structures provide stability in difficult times: When unforeseen hurdles come up—like an insolvency, an acquisition or having to temporarily set channels to zero—the whole construct doesn’t immediately collapse. Anyone who has built a reliable foundation of clean data, well-understood customer flows and customer lifetime value can then ramp back up extremely quickly and in a targeted way.

If you enjoyed this episode of Analytics Sprechstunde, subscribe to the podcast and feel free to leave me a rating. Until next time!