 What an incredible stage this is, isn't it? Absolutely amazing. Listen, it's great to see all of you here. It's wonderful to be back in Helsinki for Slush once again. We're gonna talk about performance marketing now. That in itself froze us a little bit of a loop because performance marketing has about two and a half million definitions, I think. It is true. Yeah, so I think what we should do first is why don't we just go around and you tell us what performance marketing means for you. Okay, so for me, it means a way to do marketing that is really data-driven. It doesn't mean that it's only certain time channels. For example, you can do even brand marketing and still being a performance marketer. So it really means making decisions just based on data and being analytical and trying to have the right KPIs for the right team. That's my definition at least. Okay. Yeah, I would say it's goal-driven. So you know you have a goal at the end that you want to achieve with your activities. In my roles, it's always been sales, but I mean it could be just an install or registration. And the interesting thing is it's totally transparent. So that comes with the data. Yeah, and I would say it's everything you can measure in marketing. So basically you can make almost every channel measurable. And if you look at this data on a daily or weekly basis, you have a very good view on your marketing efficiency and cost. All right, so that all sounds very utopian to me. Let's be honest. Marketers have been asked to be data-driven for many, many, many years. And then we get the hippo in the room, right? The highest-paid person's opinion. And they come along and they say, well, I don't really care about your data. I'm paying the bills. You'll do what I want. How do we actually deal with that? I mean, that's a really good basis for the first thing you should do when you're dealing with performance marketing is setting the expectations, setting the right culture and making sure that the hippo doesn't get a say. How does that work for you? So setting the expectation on the KPI that people need to reach. I think you need to teach them. So I think every person in a company, independently from where they are and which department they are, they should know about the business and really understand fully what is the impact of what they do every day on the business side. I don't, for example, for my team, I don't want to have like a Cosper-acquired customer goal that they need to reach and they don't even know why. Why is it this and it's no more? They need to know exactly the reason. And I think as soon as you onboard everyone and you explain why and the long-term goals of the company and your transparent, that is the best way to create a really good performance marketing culture, in my opinion. Do you agree with that? Yes, I do. And I actually, in my team, I don't have targets against them working with different channels because they change so much on a monthly basis. So actually, it's not how much work you put in to get the best results always. They're really unfair, so I try to create a culture where they work together. And if one is struggling, you leave whatever you're working on and you help out. And I think if you don't do that and you have very strict targets coming from above, you can nurture the culture that, yes, it's data-driven and you can show whatever you want to show in the numbers, but numbers can also be fake and not so great for the business long-term. So I think that's what happens in a lot of large corporations where people have targets that they need to reach and in the end they're just building up a pile of lies. And then no one dares to break that because then you end up with bad results. I think it's super important always to have that conversation within the business that you should be telling the truth. Yeah, I've been expanding on that for a second. You know, I see that a lot and that's because a lot of CMOs within companies are under such a huge amount of pressure to produce results and they'd love to try really amazing creative, content-driven plans, but they end up doing lowest common denominator ad buying because they know that on a CPI basis, they can get a guaranteed return for their money. You know, that's a problem because they can't do the really high-impact stuff. They always end up doing lowest common denominator advertising because it will save their jobs and CMOs right now actually, I think have the shortest tenure of any executive. I think the average is like 18 months or something before they change jobs. Like, how do we give CMOs the ability to be creative and do incredibly powerful performance marketing that is data-driven that, you know, does show results and give them the freedom to be able to do that? I think, I mean, maybe if I can jump in. So I think kind of from Zalanda, our learning was to split budgets a little bit, right? So you have smaller budget for branding and creative stuff where you do exactly what you said, crazy stuff, you don't ask anyone, you don't measure it in the beginning and then you throw it out and this is a bit more about gut feel of the CMO and actually kind of it's better not to test it, right? Because otherwise you always get to a mainstream result and then the biggest chunk of your marketing budget should still be on a performance basis, right? So what you measure every day and, you know, then draw your conclusions. I think you always need to have a smaller budget for this experimental branding and creativity. Otherwise, kind of, you know, your brand gets a bit boring, mainstream, you can't stick out. So you need some experimental money for that. And I think there was also like an article from Andrew Chen for a couple of months ago and he was actually saying that paid acquisition can kill your startup, exactly for this reason because paid acquisition is not easy, but let's say that it is easy, right? So you invest and you have a result right away and the risk is that you say, why should I go for the creative idea when I can do this and it will give me money right away? So you need to look at the long term because there will be a point where you reach a threshold where it's not possible to grow or it's gonna be really, really hard. So you need to think beyond about like, okay, if I wanna be successful in the long term, it's not only performance marketing. I need to build a brand. I need to have an organic acquisition squad, for example. I need to have a referral program. So I think probably thinking about the long term is the right strategy to not just stay in your paid acquisition bubble. Because ultimately what we're trying to do is elicit emotions in consumers. Because unless they feel an emotion, they're not gonna care about what we have to sell them, right? Exactly. And also maybe translate this in business. Happy customer, higher lifetime value. So you can make this a data-driven decision as well if you want. Yeah. And I would say even you can kind of measure the brand impact, right? So if you say, what are attributes to my brand? And then you do a certain campaign and afterwards you measure it. So you know, did I achieve my goal in whatever becoming more kind of funny or emotional or fashionable, whatever the brand is? And then you somehow get into the same measurement in terms of seeing the results and optimizing afterwards. Makes sense. Absolutely. I mean, you know, if we think about this from just a purely practical point of view, if you're going to do data-driven marketing, you need to use marketing technology. When I first started analyzing marketing technology, there were maybe 750 products that mattered. When I gave up and couldn't do it anymore, I had on my database over 9,500 products that mattered. And these are the ones that were funded, that are making money. There were 700 mobile analytic programs alone, all competing with each other. You know, how on earth do you navigate the marketing technology landscape and choose the right solutions to give you what you need but that connect with each other so that you really, truly have a good marketing stack? And maybe, you know, tell people out here what some of those tools are. You know, what are the things that really help you to do data-driven performance marketing? I mean, let's start from the basics, mobile apps, so you need an attribution tool. In our case, it's a just break. There are a lot of different players and I think you need to understand which one is better and which one, the reason is better with the infrastructure that you have around you. That's the first thing. But also, we have a really strong BI and engineering department and they've always been really, let's say, informed about what we need from a marketing point of view, so we develop our own tracking solutions for mobile and web and that's really cool because we have all the information we need inside our warehouse and that's awesome because that's everything that we need and everyone is marketing, is aware of how the structure works and then we build on top, we actually, let's say, boat some tools that will help us to make better decisions. Some of these tools are, for example, Matillion. There's an ATL tool so that we can take the data, manipulate it and put it in a form where you can easily access to you and then we have a visualization tool for Dashboard. Again, if you need one BI person to do the Dashboard for everyone, you go to a certain tools. I want my entire team to build a Dashboard that they need, so with SQL. So I need a tool that will let me do that, for example, Periscope data. So yeah, these are probably the most important tools that we use at the moment. Excellent, thank you. Yeah, I would say it's very similar. I mean, I guess everyone using an attribution provider, I find it quite difficult that the one that we use, they don't provide all the cost, for example, so aggregating all the cost was a big problem for us for a long time and we sold them that by going to an external company that can help us with that and connect to all of the different third-party tools APIs. We have a business analyst in the team, we use a visualization product as well so I would say it's very similar. Very cool. And yeah, I think similar. I think having the BI team in place in the very beginning is crucial, right? So, because if your numbers are a bit shaky, then the risk is very high, you go in the wrong direction. So I think having that in place, and even if it's in spreadsheets that kind of your team does themselves in the very first days, it's also fine, right? As long as you have transparency over your numbers and then once you grow bigger, you iterate and find other solutions for that, but it's just core to have the fundamentals. Actually, spreadsheets is the old-fashioned and we don't use them anymore, but we did use them for a very long time and I think the good thing with spreadsheets is that maybe every marketer is very comfortable with manipulating or visualization tools or SQLs and I think it gives a very good way of digging into the numbers and throw them around and really analysis. I think for us, actually, Excel is still one of the main tool, mainly because, so we have this fancy dashboard and they're really beautiful, but then if something happened on the backend, for example, so we fixed a bug or something, the dashboard and the numbers will change. So I saw that some people are just downloading the data, storing there, so that they can say, wait, this changed and I didn't notice what happened and then they go deeper and analyze. So Excel is still good. That's a job. You heard it here, first ladies and gentlemen, we're still using Excel in marketing. Yes. I thought we'd burnt that with fire, but apparently not. That's a little bit like all these blockchain companies. Apparently, whereas most applications are developed on things like C, all blockchain companies are developed on PowerPoint right now, I understand, so yeah. All right. Thank you for the laugh. So just a reminder to the audience, do go to Slido and make sure you put in slash 18 and join the Evergreen stage. We're getting some really good questions here and you can actually put your name in so that I can call you out. So don't be that person who doesn't put your name in. We'll get to some questions in a moment, but everybody here who wants to do performance marketing, all these founders who really want to get results, what is the very, very first thing that they should be doing to make sure they set themselves up for success? You know, what is step one? So I'll say step one, set up your BI infrastructure so you can track everything. Step two, I would say is, test out a few channels that might work for you and then go slowly, figure out which campaigns are working. If you see one is working, put much more budget behind it until you might get to a ceiling, then you optimize and you go on and on, right? So I mean, at the London, we always had the rule of open budget, so we never set ourselves kind of a fixed budget. It was more that we said, okay, this is the cac we can pay for customer, which is like a lifetime value of 180 days, so we're okay in financing customer for half a year. And as long as we reach this cac and then stay under this, it's fine and we put as much money behind it as we can, right? And I think if you have kind of such framework, it helps you in navigating through your marketing journey. And I think that's like a very important thing that you should never talk about marketing budgets. It's a sales budget if you want to do performance marketing. So it should always deliver returns. What they are is different, right? But if you want to grow, you should look at it as a sales channel and hopefully you can have at least the people I work with that know, I will always ask for more. It's the only way to grow. Yeah. For me, the step one is hire someone that knows more than you. As a founder, maybe you have a marketing experience and that's awesome. But I just hire someone that knows more and can guide you, trust this person. Because you as a founder, of course, you know the product, probably you know the target audience and so on, but this person that you will hire will know the channels and what to do to make your product successful. So that would be the first step for me. And if you really have some knowledge regarding which channels at the moment is working better, maybe hire an expert on that channel. I don't, I think, I mean, of course it depends on how big is the organization. In our case, the performance team is still like nine people, so I can still be hands-on and I love this part. I love to be hands-on. Maybe I'm not the best at campaign management. Sometimes my team is like, no, just give it to me. But I need it to just like keep my feet on the ground and I need to understand what they're doing every day and I think for me is really important. So do not lose touch. And then, of course, the data part. For example, we didn't have a BI department at the beginning, but we still had people that were really into the data part. So we actually made it happen and we made the right decision and it really fade off in the long term. So, you know, thinking about that, one of the things that was definitely a trend in 2017 and has continued to be throughout this year was that there wasn't a week that went by. In fact, almost there wasn't a day that went by where we had a artificially intelligent marketing technology come along. And, you know, AI has worked its way into marketing technology. It's one of the biggest areas for artificial intelligence. It is doing some really interesting stuff, you know, from an SEO standpoint. You know, it would not be a good idea to get a job as an SEO consultant right now because AI can do the job a hundred times better than you and a thousand times faster. You know, how has it affected what you're doing? You know, has AI come into your marketing stack and what do you do using it for and what are the pitfalls? So regarding AI, I mean, it is kind of inside the current platforms that we are using. On Facebook, you have automated bidding based on in-app events. So you don't do anything, I mean, you don't do anything anymore. You launch the campaigns, you set the target, but that's it, algorithm on the back will do it. From our side, I haven't found like an external provider, like an ad network that is using AI and is successful. That's really bad. But I think what is really important and what I'm using for sure is automation. Like in general, I just like, I'm not scared about the fact that our job will change. I know that it will and I think we're ready for that. We want to get rid of all these tasks where humans can make mistakes. It can be changing the bids, it can be changing the budget, upload the creatives, decide which one is better or not. And there are a lot of solutions out there that can automate for you. We decided to do it by ourselves because it's cheaper. Because also this solution are like a one package that has everything inside. And I actually realized that at some point we were paying for them, and I was using one little feature. And so right now we're focusing on really like automate everything that can be automated so that the channel manager can focus on the strategic part, but also so that we can grow in a lean way. And we don't have like thousands of channel managers maybe for doing a task that is not really, yeah, exciting, let's say. I would also say kind of in many aspects, AI or machine learning is just the continuation of continuous development of algorithms, right? So if you take CRM, for example, I mean, yeah, the more AI or machine learning you apply, the better you get in selecting your customers and sending them push notifications, emails, whatever kind of contact you want. But it's also nothing crazily new, right? It's just you put the AI sticker on it and charge 20% more as a CRM company, but fundamentally it's not so different, right? But I think in the future there will be kind of much more and your marketing will get even more targeted and more efficient with this better algorithms. I mean, that's pretty clear. Yeah. How does AI affect what you're doing? Are you using anything that is actually AI powered? Nothing that's not already like in the network of publishers or things like we're using like that for algorithms. I find it hard when I talk to the different companies to find one that actually does what I wanted to do, including all the channels, because I have to have a lot of different systems for different things. It's not going to help. Then you have a lot of the manual work yourself. Yeah. And then I don't want to pay a premium for it on top because as you scale adding another 10% or whatever, you know, it's going to be a lot of money. So there's a couple of really good questions here. I'd like to say thank you to Rich Elliott. You can wave at us if you like, Rich. He asks, what are your top three marketing tips for a new brand? I like that. Top three marketing tips for a new brand. Hmm. Top marketing tips. Just three of them. Three of them. So first of all, understand who you're talking to. It's, you cannot, like, you cannot target everyone. So start with what you think will work the best. Identify this and then create like the right messaging that we resonate with this person. Make your brand marketing team or creative team accountable for giving you creatives that will resonate with this audience and will make the success of the performance marketing team as well. So split the accountability, the accountabilities, make them accountable for that. And number three, I would say, again, higher talents, higher people that have experience with, like, either the channels that you're currently using or maybe has done that before. Cool. You two have got the harder job now, because you've actually got to come up with the tips with all the same as those three. I think it depends as well what you say when I say if you're a new brand. If that means that you just want to push it as a brand thing or if you're just a small company starts off and you're a new brand. And it makes a big difference as well if you're a completely new product and no one really knows what it is or if you have more competition. Because if you have more competition in a way, it's easier because people are looking for your products. Because they know they're out there. While if you completely knew kind of technology or service or whatever, you can't go for the kind of lower funnel stuff. So maybe then you need to educate a bit more. So then maybe I would say, like, educating. I would definitely say what you said about if you have a brand team and a performance marketing team that they have to work together, because branded assets don't really tend to work for performance marketing. And the last thing, yeah, I think you said a good one with like knowledgeable people. And I think maybe one other point to add is be creative and be bold, right? So I think if you're building a new brand, you have nothing to lose, right? That's a good thing, because if you have a legacy brand, you know, the bigger you get, you have to be more cautious, right? Because you don't want to hurt your brand. You have to be really careful to get more criticism. So if you have a new and fresh brand, you can be crazy, right? So I mean, there's like the chance that you're going to stick out is also very limited, to be honest. So whatever, one in a million of brands is going to be super successful with a viral. But I think if you're a new brand, try it out, right? You have all the freedom of the world and maybe you get the right shot. Yeah, absolutely. I mean, it's a great example of that with, you know, there are millions of breweries all over the world, but BrewDog came out as really interesting for everybody because they basically said the word fuck about 45 times in every press release. And they had an attitude and they had that tone of voice running across everything and they made themselves different because they were just sweary and alternate and they got a name for themselves. I mean, there's millions of breweries out there, but we all know BrewDog. But you should not forget they also have a great product. Well, there is that, but it's beer. I mean, how can it be bad? So I really like this one as well. I would also just like to do a shout out to the infamous hacking group for asking this question. This one's from Anonymous. What would you focus on if your main source of growth is virality and word of mouth? I think it's a really interesting question because word of mouth is by far the most difficult thing for us to measure and attribute. So, you know, what do you focus on if you just want to build your business with word of mouth? And virality is interesting, right? Because I don't think you can actually plan for virality. There's so much you can do to release your creativity into the world, but you can't really bank on virality. How do you manage that? Huh, chaff. So I don't believe a lot in virality. I think it can happen, but you cannot plan it. And I think that behind most of the viral stuff that you can find out there, there is a little bit of budgets because still you have to produce something. Still you have to put it out there and maybe put a little bit of money behind. So it's really, really tough. And for me, I've never really made this as like a focus point. But I think when it comes to word of mouth, if that's like the way you want to grow, then you probably need to work on your referral program and make sure that there's a way to track it through vouchers or whatever. And also the engagement with the product is really important, right? Because you want people to talk about you. So I would probably like shift everything into like the engineering part. I'd say if it's a mobile product to really like get the most out of it. That's what I would do. But still, you need a little bit of budget. I'm sorry. Right. And I think to be honest, if virality and word of mouth is your only channel to grow, I think you'll have a problem, right? So, because I mean, now being an investor, I mean, we see so many companies where the pitch is, you know, you ask them how you're going to go, yeah, through virality. And they say, yeah, but what is behind that? Because as you said, I mean, it's very hard to kick it off, right? And I think if you want to try it, you need to understand every little detail. What's the target group? What are maybe like influences that push you? What are the right buzzwords and so on? But even then, you know, if you're brilliant, the chance to work it out is very limited. So I think you should think about what is another scalable channel that you can put money behind and scare because just virality is not going to be enough. I mean, just expand on the word of mouth thing for a second, though. How do you measure word of mouth? How, you know, we need to still explain to our executives, you know, what's actually happening, give them the world's lovely KPIs, they love so much. How do you do that if your main channel is word of mouth? I mean, you can measure in different ways. For example, you can have a survey where you ask people what they heard about the product. Of course, you always need to remember that people are lying and they're just like clicking randomly. I remember that before we were doing TV, there were like still a lot of people saying that they saw us on TV, for example. So you need maybe to build a baseline before. That's the first thing. Then if, for example, if you invite people to share something when they are in the product, just link so you can track everything, that's easy. Maybe vouchers in case you have a subscription model app or something that people have to pay. Hopefully that's the case. So there are still ways to track it, I think. Yeah. I mean, I don't think, when I think of virality, it's not really the things that you've been enforcing too. I'm thinking it's more like things that happens on itself. So either you have that brand campaign that's supposed to look really true, right? And people share it. Or you just have a really great product. Or maybe you figure out something new that everyone needs to have, you know? And then people are sharing it because it's a great product. So I think then you know how to put a multiplier on that because you see they're in your own product. Right. And I think ultimately it always gets convoluted in the whole branded and direct traffic, right? Because at the end, it helps your brand. People come to you automatically for free. And that gets all mixed up, right? So it's very hard to single it out as one channel. This has been a fascinating discussion. Thank you all ever so much. We've been talking about performance marketing. I can only hope that everybody out there enjoyed our performance. And please, everybody, give our wonderful experts a warm round of applause. Thank you. Thanks.