“Clear Priorities Through 80/20 and Funnel Thinking”—with Marco Jenzer (Würth)

Recorded in German—English translation of the original conversation.

Intro

Maria-Lena Matysik: Welcome to Analytics Sprechstunde. I’m Maria-Lena Matysik, analytics consultant and managing director of Analyticsfreaks. Here I share with you how marketing teams can really use their data for better decisions and sustainable success—with concrete insights, actionable tips and no buzzwords at all!

Today I have a guest visiting from Switzerland—and you’ll hear that right away. I’m talking to Marco Jenzer. Marco heads digital marketing at Würth Switzerland. He’s now been in this role for a little more than five years and has been part of the Würth family for more than eight years in total.

Marco leads a team of six marketing specialists. His job is to pull everything together and create common ground—especially when it comes to data—in other words, to get the team ready for the constantly new challenges in digital marketing and in analytics, too.

Maybe a few more words on the context Marco works in: Würth is an international trading company with national subsidiaries, and it sells screws! And everything else tradespeople need. That means Würth relies on direct sales, that is, on sales staff who drive directly to trade businesses and take orders on-site. At the same time, though, online sales also play a really big and, of course, ever bigger role.

In the interview, Marco tells me how the group structure affects working with data, how they manage within the Würth family, that is, the Würth Group, to work together internationally, collaboratively and with a real sense of community—and what the concept of “Steal with Pride” is all about.

We’ll talk about how marketing teams really use data in their everyday work—without over-engineering, without huge dashboards and super complicated Excel models. And Marco explains why using data is an ongoing process for him, how to keep the setup lean and understandable, which KPIs really count in day-to-day business—and above all, how to create a shared understanding of data and why a pragmatic 80/20 approach often gets you further than the one perfect solution.

Okay, let’s go!

Interview

Maria-Lena Matysik: Hi Marco, welcome to Analytics Sprechstunde! I’m very happy to be able to welcome you here as a guest today.

You already know this, of course: I always ask my guests here on the podcast the same question, namely: What learning have you brought along for us when it comes to working with data in marketing? Was there something at your company that sort of made things click in terms of working with data and insights from data?

Marco Jenzer: Yes, actually a very simple realization—but then really anchoring it in day-to-day business and actively living it afterward, that’s what sort of made things click, or maybe also took some of the pressure off. Namely, when we really sat down for a while, or rather took an hour or two, and thought: Okay, what do we actually need? And what don’t we do? What do we also want?

Moving away from an over-engineered setup and simply admitting to ourselves: We don’t always need a 360-degree solution. It doesn’t always have to be the Swiss Army knife from day one; sometimes it’s also good to start out with an 80, 85, maybe 90 percent solution. You don’t always need the super high-tech, ultimate premium solution from day one. If it develops over time and you keep improving a bit, after a certain time you get to the point where you have that solution, or rather the great setup. And it’s completely okay if it grows like that over time. Becoming aware of that and admitting it to ourselves—that was it.

Maria-Lena Matysik: Yes, really interesting. I always really like it when the insight turns out to be a simple one but brings a whole lot more value. How would you assess how important this topic is in your day-to-day work?

Marco Jenzer: In our team and in the company in general, very important. Of course, there are areas that aren’t as deep into the numbers or don’t work with the numbers as intensively as we do. For us, it simply comes with the job—everyone who works in digital marketing or in marketing in general knows that. And when it’s even tied to digital sales—for us, that’s the e-business area, meaning all the digital sales channels, in close collaboration with our direct sales, since we’re a direct sales company—then it’s a lot about revenue, a lot about sales, a lot about evaluating data and assessing initiatives: What works, what doesn’t? Then it’s super, super important.

And you can take a lot of complexity out of it if you know: Okay, what do I need, what don’t I need—as I said earlier. And if you can simply get everyone in the team, in the department, on common ground by making the topics more tangible and understandable and taking out the complexity. Because experience has simply shown: If you have a very cumbersome and complex setup—even if it’s just a complex Excel sheet—then the people who work with it avoid it, because the level of complexity simply makes using the data and reports harder.

Maria-Lena Matysik: Yes, I can totally relate to that. Could you say that this was sort of your starting point? That you saw: Okay, we’re somehow trying everything at once, we want the perfect solution—and as a result, we’re actually just making everything even more complex, even more cumbersome to work with and even harder to get the relevant insights from the data?

Marco Jenzer: Yes, to a certain extent you could say that. If I look back a bit, maybe three, four years, then I’d say: definitely. Back then, people were partly working with over-engineered Excel spreadsheets and reports. You knew: Now the start of the month is coming around again, I’m doing my monthly reporting and sitting in front of this Excel analysis for three hours again. For the employees, that was sometimes a nightmare. And you simply collected the data because you had it—so you’d have it, in case you ever needed it.

That’s not a bad thing in principle; I don’t want to condemn it at all. Of course it’s totally awesome when someone from management comes and says: “Show me the development of—I don’t know—our Instagram performance over the last three years,” and I have the file and don’t have to make it first. But let’s be honest: How often does that happen? And then you can still do it, even ad hoc.

And that brings us right to the point of doing things ad hoc: If I keep doing something ad hoc, I’m of course glad if I have a template or an approach that’s always the same or similar, so I don’t keep starting from zero. That was actually the starting situation: It was complex and extensive, and then we really had to look: Okay, how can we approach this better so we don’t start over every time and, above all, don’t have to jump through hoops every time we need numbers from somewhere?

Maria-Lena Matysik: I think you’ve just teased your solution. My next question would of course have been: What did you do to get out of this over-engineering and the manual overload with data?

Marco Jenzer: I’d take another half step back and describe the situation as you already teased it earlier. As soon as the question came up—Oh, how did things go over the past one or two months or weeks?—you’d kind of panic: a moment of shock, maybe a few minutes: Oh, now we need to have the data! And then everyone stood there, and nobody knew: Okay, how do we best approach this now? Do we do a big analysis now? Do we go through the controlling team? Or do I first set up a meeting, bring in seven stakeholders and say: One person does the analysis, that colleague does this, I do that and you do that?

We moved away from that. We simply asked ourselves: Okay, who needs which data, and what’s the common ground? And we also made ourselves aware that there’s a very different understanding of data in different people’s heads and in interdisciplinary collaboration. Product management, for example, works with different data anyway, but also with different kinds of data than, say, marketing or e-business. So I can’t just call out in the meeting: “The conversion rate was bad, it was only such and such, and the bounce rate was this and that.” Then I’m only ever addressing exactly the specialists and experts who work with it—but not the whole group. And then someone tunes out every now and then, and I just can’t bring the majority along.

That was actually the first step. Then we went and really sat down together once and thought—as I’ve now said twice already: What do we need to know at the end of, for example, an initiative, a promotional period or a campaign? What do we want to measure when the campaign is over? We’re very performance- and sales-oriented. So we probably want to be able to measure revenue, for example. Or we want to be able to measure how many orders were placed. But then we want to know other things as well: Were we able to sell the products or the product range XY—or the anchor bolts, to name names—to customers? And then all of a sudden a completely different KPI is important: no longer revenue, but how many customers bought product X, Y and Z in period A, B or C.

And then you break it down—away from always just “revenue, revenue, revenue, we need a few percent growth here and there, and another great KPI here”—and ask yourself: Okay, what did we really influence? And what does it look like in total at the end?

Maria-Lena Matysik: I actually imagine that you approached it like this: Okay, what’s the result? And what—to put it very simply—is the funnel leading there? So, for example, users, cart values or rather “added to cart” and, in the end, revenue—like a kind of mini funnel up to the result. To keep it a bit contained and, let’s say, to simply keep the number of KPIs small.

Marco Jenzer: Exactly, that’s exactly the point. That was actually our approach. Because I’ve said again and again—and still say, because we’re not done; we won’t be done in a year or in two years either—that it’s a huge, long process. Never ending.

We were sitting in the meeting—that could have been today, but it could also have been two years ago, it doesn’t matter—and the statement was: “We need to sell more screws, we need more revenue.” Then you stare at the task: What do we do now? Of course we need to sell more products. But revenue is ultimately nothing other—as you also said—than the result of combining different variables. Because who generates the revenue? First and foremost, the customers. That would already be the one KPI you mentioned earlier.

Starting from the end product, it’s like the rule of three, actually like a math problem: I have the result, but still two, three unknowns. How can I calculate that backward so I can break it down into the sub-areas? In our setup, for example, we generate no or only very few new customers through the email marketing channel. That’s down to the process—we only sell to businesses and so on. That’s not bad, it’s just the way it is in our case.

Then on one side I have revenue, and I set up the equation with our three variables. We already know: One variable is the number of customers. And the second and third unknowns are, for one thing, purchase frequency—as you already hinted: How often do the customers buy? And that gets multiplied again by what you also already said: by the order value of their purchases with us. If I multiply all of that together, I end up with exactly the KPI everyone’s talking about: revenue.

Then I can say: Okay, now I’ve broken down a complex construct that everyone was sitting in front of open-mouthed two minutes ago—“Oh, we need half a million in revenue!”—into three variables. And I can work on those. The complex task becomes something I can break down, and then I can say: Okay, this channel’s strength is acquiring new customers—ergo, social media and SEA now focus on new customer acquisition. In the newsletter, or rather in email marketing, our strength is encouraging customers to buy again—so email marketing focuses on the KPI purchase frequency. And on the website and in the online store, we’re strong at upselling—so there we focus on the KPI order value, for example by adding a trigger or a pop-up: “Hey, only 10 or 20 euros to go until free shipping” or whatever.

Maria-Lena Matysik: Something that just occurred to me while you were talking: I could imagine that this makes it so much easier for you to really involve the entire team with all the marketing disciplines and all the experts. Because when—in quotes—the directive comes from above: “Hey, we need this much revenue,” that’s of course relatively abstract and intangible for a social media manager. A social media manager thinks more in terms of posts, impressions and so on, but not primarily in terms of revenue. But a simple equation like that, I could imagine, builds a great bridge between an abstract number—the goal—and the activities that the individual marketing disciplines can actually carry out.

Marco Jenzer: Absolutely, I’d agree with that one hundred percent. That’s exactly the case for us, because it simply makes it more tangible and concrete—and also simpler. Because as you rightly said: Nobody’s sitting there thinking in terms of revenue. Someone in direct sales might do that more, for us, for example, a field sales rep who’s in the car on the way to the next customer and thinks: Okay, today I need to write an order with at least such and such an order value, blah, blah, blah.

But someone sitting in front of Meta Campaign Manager doesn’t do that. They think completely differently, more along the lines of: Okay, we want to build a campaign now, and I have to reach this many people, or the reach has to be X today. That’s where we calculate backward: I then reach, for example, 150,000 people and have—no idea—an assumed conversion rate of two percent. Then in the end that gives me so and so many orders, times the average order value, and then I generate such and such revenue. And that’s my part, what I can contribute in my area of expertise.

Maria-Lena Matysik: Absolutely, right.

Marco Jenzer: Yes, I’d second exactly that.

Maria-Lena Matysik: Yeah, cool. As Würth Switzerland, you’re embedded in an international group. What role did that play in developing your best practices—that is, in how you work with data?

Marco Jenzer: A central role, to answer it very directly. For one thing, it’s a great setup—I find the collaboration in the group very cool. It’s not at all the case that something gets dictated from above: You have to do this and that. On certain topics it’s clear, there you follow a common strategy that everything is supposed to contribute to at the end of the year, the month or the day. Otherwise, though, it really is a cool community—you could almost say: well, the Würth family. You really notice that across national borders, too. Now I’m doing another plug for my employer here—of course I wouldn’t have to say that, but it’s actually true.

And there really are also solutions provided in the sense of a collaboration framework, including a toolbox, which we then started working with. For example, this very “Super Simple Strategy” with these three KPIs was presented there as a best practice: “This takes the complexity out, try working with it.” It really didn’t take long; it was recognized relatively early and also well received. You really have to give people a lot of credit for that. There’s also an appreciation toward one’s managers, you can see it that way—I certainly see it that way: When I tackle something with the team, it isn’t questioned much at all, and if it is, then constructively. And then people work with it, because at the end of the day it simply makes the whole topic easier for everyone.

Maria-Lena Matysik: I think that’s really cool. It’s often the case that corporate groups are big, there are lots of rules, everything is slow and so on. I think it’s really nice to hear that there are positive sides too—namely, that in a large company you can learn from each other: How do others do it? How would we do it? What do we adapt, what might we not adapt, and why? I think that’s really cool!

Marco Jenzer: Yes, it’s really a great thing. We call it “Steal with Pride.” For example, if I see at another Würth company, say in Portugal: They’re doing something mega cool! Then I just go and steal all the images and the source code. For that, I quickly write an email or make a quick call and say: “You’re doing that really nicely—could we maybe steal it from you?” That’s Steal with Pride. And then they say: “Yes, help yourselves, of course you can have it, cool, please go ahead.” And then everyone’s fine with it.

And when topics like that are then presented and shown as best practice, you simply have to check: Does that work for our market too, does it work for our company? Because just because something went through the roof at Würth in Denmark and tripled revenue there or I don’t know which KPIs, that doesn’t automatically mean it works just as well for us in Switzerland. Of course it may not work at all. It may work quite decently. But it may also go through the roof for us, too. But then everyone just has to find their approach that works best for him or for her. And of course that differs from country to country, sure.

Maria-Lena Matysik: Do you, or rather do you as a team, still have open points on the to-do list where you say: Okay, the process already works pretty well, but we might do this and that differently or even better after all, we’re still working on that?

Marco Jenzer: Yes, as I said: In my opinion, this is definitely never finished. Something fundamental changes every now and then anyway, whether on Google’s or Meta’s side, or somewhere something gets turned completely upside down. Then you have to look for a new way again. We know that by now—be it the GDPR or, for us in Switzerland, the new Data Protection Act two years ago. That’s normal in our working environment.

Otherwise, when it comes to the internal processes and how we work: We’re not done there yet, not on the home stretch either, but definitely in the middle of it. Becoming aware of, or rather working even more with, where each channel has its strengths and weaknesses—that’s very much the direction things are going right now. That’s of course a long process and not an easy one. But fortunately we’re not alone in that; we have cool colleagues to our left and right. That’s why I’m optimistic that we’ll manage it very well in the long run.

Maria-Lena Matysik: Cool! Do you feel that your strategy of approaching this with data—that is, identifying simple levers—will help you with that? Or do you feel here and there: Okay, we might not get any further with the Simple Strategy after all, and we need to tackle our data strategy again—if I may call it that—so that it helps you reach this goal faster?

Marco Jenzer: Yes, definitely. I think one goes hand in hand with the other. It’s not an either-or. Because our Super Simple Strategy doesn’t solve all problems either—and when we have bigger challenges, maybe not conclusively. But it does take care of, I’d say, the bulk of it. You can maybe compare it a bit to 80/20, that is, Pareto: We have a big chunk we need to tackle. What do we do? First, we approach the topic with our Super Simple Strategy. Then we already have 80 percent.

When we say: Okay, we need to go deeper here because it’s a super important topic—it comes top-down, from the group or from somewhere, and we need to go the extra mile there; and we do that very often, regularly in any case—then we sit down again with the specialists from the controlling team and break it down really deep. Then we can really look: How many orders of this screw came in through an Android device in the last 30 days? And with our automated storage and vending systems: What was often stored in compartments one through ten? Aha—that was the batteries and the cable ties. Then we dig even deeper there. We do that too, and nobody can define and analyze that with these three KPIs or three variables. You do need something additional for that.

And when we then break that down and take another look at our entire analytics setup, our entire reporting setup and whatever else, then all the more so. There we have to define very clearly: What data do we collect, what don’t we have? And what’s the long-term goal of our setup? We ask ourselves that question, let’s say, twice a year, and it actually always moves within a spectrum from here to here. Sometimes in a quarter or half-year the focus leans more in one direction; in the fall it shifts more in the other again. But as long as we stay within that spectrum, that’s completely okay too.

Maria-Lena Matysik: Very interesting! Thank you for sharing. One last question, kind of my closing question: What would you recommend to other marketing team leads—how do you best get started working with data, and in a way that’s actually productive?

Marco Jenzer: Yes, now I could just blurt out: Just do it. That would be a bit boring—but it really is like that: just do it. Of course a concept is part of the task, that’s very clear. Just doing it and only coming up with a strategy for it afterward—that probably rarely turns out really well, and if it does, then more by chance. But I’d really say: a hands-on approach. A rough concept already helps. It doesn’t always have to be a seven-page document that covers every eventuality. Instead, a rough concept is part of the task: sitting down, finalizing the key parameters and then simply thinking: Okay, what do we need to accomplish this assignment or this task? What don’t we need, what’s more nice to have? What might we also want? And making sure that as a team, as a department, as a construct or whatever, you find common ground. Because then you already have a large share of the people running in the same direction, and from there you can keep planning.

And I’d also strongly advise not to over-engineer anything. I’m of the opinion that you don’t need a 360-degree solution right from the start. Depending on the setup, company and task, you might need one—but in very few cases. At least I haven’t yet come across a case in which one was needed from day one. Instead: just do it, gain your own experience and don’t fall into the over-engineering trap!

Maria-Lena Matysik: Very well summarized.

Marco Jenzer: Thanks.

Maria-Lena Matysik: Great! And with that we’ve already reached the end. Thank you so, so much for sharing all your insights and learnings from the process at your company.

Marco Jenzer: My pleasure. Thank you too.

Outro

Maria-Lena Matysik: What I took away from the conversation—and which statement from Marco immediately stuck in my head: Using data becomes much harder when the level of complexity is too high. Chaos needs to be structured. We need some kind of guidance, a guideline, a framework, simply to find a starting point and to really be able to work with the data.

And Marco also made a suggestion for what can serve as guidance: namely, the funnel. The concept behind it is also called funnel thinking, which is of course a natural fit in marketing.

That’s why my takeaway from the conversation with Marco is: Find yourself a guideline, a framework or some kind of system and structure that you can lay over your data like a template—or rather, structure your data so that it becomes part of this structure—to simply bring structure into your work with data.

And the first framework that comes to mind for me is the Pirate Framework, that is, the acronym AARRR. We also use it with our clients. The acronym stands for Acquisition, Activation, Retention, Revenue and Referral—which is simply the funnel.

I find it super helpful, because it—in quotes—forces us and marketing teams, or rather helps them, to really think in a focused way. Instead of having a hundred different metrics or KPIs—in quotes, because when you have a hundred, you can hardly call them KPIs!—so instead of having hundreds of metrics coming from everywhere, it’s super helpful to think in maybe five groups: the ones that are really relevant, step by step through the funnel. Because only that enables a focused, structured analysis in the end: Where are things stuck? At what point in the funnel, in which area? And where can we really take action?

And that’s it for this episode today. Feel free to follow me on LinkedIn and send me feedback on this episode anytime—or, of course, ideas for future episodes. Ciao, ciao!