ugo actress, olloS repeatedly. schle embedded in the what is called A-Players, analytics champions every day. And yeah. And me. Also, and also you, Sten. And one of my customers, Pandora, my contact person there, Marte, actually started working with Sten. So we talked a bit about, maybe we could do some stuff together. The thing is that when I go out and meet our customers at Acutics and talk with them about what they're doing with web analytics, there's just a lot of things that can be done in a smarter and better way. And yeah, I wanna make sure that every one of our customers gets as much value out of the web analytics as possible. And that's also the reason why I invited Sten, because he has quite a bit of experience in this space, having worked more than 20 years with web analytics, I guess since it was called Urchin and the world was a whole other different kind of beast. Yeah, so he has a lot of stuff to share. The webinar here is live. So we are sitting in Copenhagen on Bauergade. So we will also have the Rådhus clock up in the background sometimes when the time shifts. So you can hear that everything is very live. And with that, I guess you can also just say hi in the chat. We have a question coming up here in the beginning. I just wanted to see how people are actually seeing this. So the question goes, could you use your web analytics smarter? So that is just gonna go up on the screen now. And while we're looking at the, while you're answering into that, I want to welcome you Sten. Thank you. Could you share a little bit on your role at IIH Nordic? Well, that's hard to say. And your journey into data innovation and the campaign tracking. Absolutely. So basically my role today is very much, so I have this wonderful title that I invented myself, Director of Data Innovation, because I've been doing web analytics basically since 2002, something like that, in different variations, back in hit counters and click trackers, and then really basic stuff. Moving from there into urgent and Google analytics, I'm certified in Omniture for the really old people will recognize that even, yeah, Omniture and so on. All of them I've been messing around with. But I think my practical role today is really much, is very much focused on the commercial return on analytics. So going in and actually saying, okay, so we have this investment, we invest in analytics, what do we get out of it? And I think so, and that's kind of the core things also why I think it's you as a product are really interesting because for me it's, why do we track? Why do most people track? So in eight out of 10 cases, they track to document their marketing. And as late as Saturday in Munich, I was standing on a stage saying, so if this is the most important part, why the hell does it suck so bad? So the data quality in the marketing channel is normally one of the first places I go to to define, to gauge the maturity of a client. It makes a lot of sense. It's all of the reporting and the direct attribution and that's maybe not the best way to use a web analytics platform. Anyways, it looks like we have a 100%. I kind of like joke that this would probably be the result of it. Yeah, but we talked about that and I think this is true, right? So if the answer was no, then you probably wouldn't be at this webinar because it's like, I'm not wasting my time seeing this because I already got this under control. And so the risk is that we're actually preaching to the choir, right? So everybody there sitting in the audience are the people who will say, yeah, this is really important. So everything we say needs to kind of see if we can add an extra layer to the onion. Yeah, exactly. Get them out, get it out and live in the organizations then. Talking about that. So Sten, you've been very focused on GA4. I know you just started also to have a partnership with Amplitude for their web analytics. And Pivot, just to say. Yeah, so now you are a bit broader. But I know you have a lot of experience with GA4 and of course there's a lot of stuff happening since it went from the old Google analytics, whatever it was called before, Universal Analytics, I guess, and into GA4. And there's still a lot of involvement on the tool. So my question is basically, how has campaign tracking your Google analytics involved here in 2024? And what does that actually mean for decision makers? So it's kind of a good stuff, bad stuff and coming stuff. We can say we'll take that later. So on the challenging side, we still have the data loss and the cookie consent issue, right? So there is something now with the consent mode, adding data, but actually, yeah. I have a funny slide on that if you wanna see it later. So, no, you can actually see, don't show the slides yet. No, no, you shift the slides, but don't show it yet because no, next to further, further, further, further, that's too, no, that's continue. Yeah, this is really interactive people. Continue, continue, couch, there, this one, show this one. So I use this one to illustrate saying that with consent mode, right? So Google built in this engine that says, okay, if you have data in there, then we'll model based on the data you have in there, right? And the problem is that if you look at the couch, it looks like a fairly ordinary couch, but it's modeled actually, it's the same couch model on half the couch. So if you shift the slide, the problem is that there was actually an elephant sitting in the couch as well, but we completely missed that in the data set because we only modeled based on the data we have. And I think that the core problem is, if you change again, is what we see is some massive data deviations in relations to the model data because there are some core problems with the representativity of the people who give consent versus the people who do not give consent. So normally, especially from a marketing perspective, the people who give consent are much further down the funnel, right? So you basically, the reason why you measure is to measure the marketing, but the thing you don't see in your data set is the marketing. The executives extrapolate on existing customers and all the other stuff that is in there, right? So you're missing a massive opportunity. And this is something that I think Google is working on right now, trying to resolve this. They can never resolve it completely because it's still a consent issue. It's also the entire problem of saying, well, if you start a campaign today, but it models based on what happened yesterday and the last week, then suddenly there will be an error in the modeling because it won't take account into the differences you're making. So Google is actually now, it's one of the reasons why they don't commit to updating the data so fast because they need to be able to kind of feel the changes on your site so they can enhance the model. But it's a big problem that a lot of people are not aware of saying that modeling actually goes in and undermine your data quality. Is this a GA4 problem? Well, yes, it is, but I guess it's everybody who has a consent banner on their site problem. Yeah, definitely. And I think I've seen it quite a few places that, yeah. The data is not real and people think it's real data, but it's actually a model data that you're looking at. And they're trying to give you the best version of what they can show you, right? Yeah, yeah. But often it's just extrapolated on what you've already done. So that's a big problem. Yeah. On the other hand, what I see with GA4, what they're doing on the positive side, right? It's they're adding more campaign codes, more UTM codes to kind of enable us to increase the precision in the data. And I think that is a really valuable thing. So I'm actually right now when I discuss with people, one of my favorite things is for you shuffle two slides forward, I think. No, stop. That's the wrong direction. Ah, sorry. There, this one, yeah. So one of my favorites right now is the UTM marketing tactic. Because it's really, it has been a blind spot in so many people's campaign tracking, saying, okay, I'm running this campaign, but everything gets measured on the same conversion points. Right? So we send in 100,000 people to the website, but it's a branding campaign, but we measure the conversion rate in relation to how successful it was in selling. Right? So it gives some really big problems where this adds a granularity in relation to the purpose of the campaign that has really been missing before. Yeah. I think a lot of our customers, Acutics, have had this as part of their structure when they're tracking, because it is so fundamental. But I can also see for a lot of people that are not living in that world where it's more like from case to case or campaign to campaign, and they're just building out the tracking, maybe a bit ad hoc, then they might miss this very important piece and it kind of ruins a lot of the data because you don't know if it was awareness or if you're just trying to convert people. Yeah. And I think that that's actually one of the bigger problems in general in relation to web analytics, even if we call it marketing analytics, is this, the purpose of the traffic, because really we have a bigger traffic and we have a bigger issue. This is just going in and saying, I think your product is amazing at solving the problem on a higher level if I want you to have it structured. And I think it's getting more and more needed because the campaign budgets are getting more and more squeezed. So if people wanna get more out of their budget, they really need to do better tracking. But even to the extent of saying, could even be focusing on saying, well, if we look at our website, we have this idea of saying that everybody's on our website for a reason, but most people are actually not on our website because of marketing. They might be there for a thousand. So the bigger website we have, the more complex the reasons that people will be there. It might be existing suppliers. It might be consultants that wanna sell something. Might be somebody looking for a job. So there's a massive imbalance in the signal to noise ratio that we go in and we try to measure just, so we set up campaign tracking for one activity and it's like saying, okay, cool. I'll throw in a handful of peas into this pot of boiling soup and I'll see if I can monitor what happens with them. And it's almost impossible if we don't have the other stuff under control. And I think that imbalance is where a solid campaign tracking foundation becomes really important. And also because it actually goes beyond the marketing department. So people who run a lot of campaigns are, it's HR. What we see, some of the biggest deviations in performance on websites is, well, HR is running a recruitment on seven people and then suddenly traffic explodes and none of them are buying because they're there for another purpose. So going in and being able to work that and recalibrate around the, well, basically going in and saying, to what extent can we define the segments we have on the website based on their activity? And what we will have left after we defined everything is our prospects. It's, yeah, you need to segment it down basically, right? So you know which segments you're working with. I'm also thinking like from, if you're taking the executives, the marketing directors and in many cases today, the chief financial officers that starts to look into what are we spending all this money on? We have the attribution within GA and I know that there's been quite a bit of a shift in the approach to attribution. I believe now it has an unknown MTA of some sort. Can you tell a bit more about that? Also just for our audience to help the executives understand what can you actually use attribution in GA for? So I think that we're in a place where we need to take control of it ourselves. Attribution has been one of the most complex things. It's always been one of the most complex things and with the loss of cookies and data and other relations, then I think attribution has become even harder than it is. And for Google, they have gone and changed from having like, I think they had eight or nine different attribution models that you can go in and look for. They ball it down to two now, right? So you have the last click because that's what we can always see, people coming to the site, where did they come from and what can we attribute based on that. And then there's the data-driven attribution, which is Google's magic black box, where they based on the model data and your historical data will go in and try to spread people out and say, this is the secret sauce, we're not gonna show you what's going on, but this is what we think happened. And if we take the first one, what do you see if the issues with this? Because I understand if you are a chief financial officer, you wanna know did my marketing actually create some value? Because it is an investment and you are like, is my investment paying back? Because there's no reason to just give money to Google or to Facebook or whatever, you can spend your money on these days. But what are the issues with the last click? So the last click is, the problem is that the moment you have something that is a little more complex of a customer journey than just saying, I saw an ad, I clicked, I bought, then you will be missing those steps, right? So that's kind of the challenge is that you don't have a customer journey, you have a purchase experience. So breaking that and saying, okay, doesn't matter that I saw this ad somewhere first and then I visited the site. We had an amazing case back in the days with Telia. Back in the days when iPhones were really hot, we would have the same user coming on a website, looking at the same phone, maybe between 20 and 40 times over two, three months before they ended up buying it. Yeah. And that was kind of a, they were just looking at this phone, dreaming of the day they would have the money to buy it. And so, and from a last click, we will only see the last time the person came and that they bought it. The entire dream phase would be completely invisible. Yeah. So in your marketing budget allocation, you would just think that everyone would just come directly and if you use the last click, but in reality, we need to make people dream. I think I've seen some like the Jaguar stories and where you can buy a really expensive car and you go on your dream about it for 20 years and then you actually end up buying it. But it was the commercial 20 years ago that had you. That started the dream. It's also why I saw a presentation from Aston Martin and they say, nobody goes to our website to buy the car. It's the fan base, it's everybody else that will go and look at the car. The people interested in buying the car will go to a dealer and get a test drive. Or they will actually interact because for them, it's not the dream phase isn't necessarily that long. No. And if you look in the other direction, I guess there are some cases where you can actually use the last click, right? Because if you have a product that's very just something you need on a, let's say umbrella, for example, it's rainy, you need our umbrella, then you go buy it. Yeah, yeah, yeah. Yeah, and so it's really, like I said, it's dependent on the customer journey saying how long is it? How long is the decision process around this? So in a lot of cases, like, buying a computer game for some kids at least, right? It's something they go, they look at it and either they buy it or they don't. Yeah. So that process is so fast. Yeah. But I think the complication is what we also see is that sometimes when people come to you based on these searches, actually it's an indicator of something going on in the market. So if you look at it from a marketing intelligence perspective saying, okay, you have this product, I'm searching specifically for buy these shoes and I come to your website and then I don't buy the shoes. Now that's a signal that there's something wrong with your placement in the market, right? Because the purchase intent is there and you still end up not buying it and that means you probably bought it somewhere else. Yes. That you were actually in a competitive situation where you completely overlooked saying somebody is selling this cheaper or with more fringe benefits or better loyalty program or whatever. But I'm failing if these high purchase intent keywords are actually not delivering any sales. Yes. Then you have a really big problem because that's just eating your bottom line. Yeah. Completely right. Yeah, I mean, and those are the sales you're really counting on. Definitely. I think there's a whole story here on also just making sure that you are investing in the top funnel and making sure that, yeah, you're not only looking at the conversion path and path because that is, I think from like a executive point of view, direct distribution, it's probably not something that you should look that much at. I don't think there's many umbrella cases out there. Most cases are probably a bit more. And I think so, so from my perspective, one of the things that we like to talk about saying is that if you are in a situation where you have a more complex purchase journey, right? Then you need to set up something different. And I think one of the big misunderstandings that we see there is that, I lost my own train of thought. That's okay. No, what was I saying? I was talking about. You were just going into, so we were talking about the broader setup. So you have some marketing. Yeah, yeah, yeah, yeah, thanks. So, and one of the big misunderstandings is saying that even though we cannot identify the individual users in Google Analytics, we can still identify the Google and the Google Analytics, we can still identify the Google, the individual user in another system. So for us, it's really about saying, you need to tie another system in so you can get this identity resolution. So you can identify and build your own customer journey. And this is next webinar talking about the bigger project processes and how we're actually moving people in the process and identifying them. Because that is desperately needed. Definitely. And I also think it's like, if we're looking at Google search and there's a lot of branded search out there and there's a lot of agencies that have this tendency to forget to mark that it's branded search and you're just looking at the overall budget and Google looks really good. But it's again, sometimes a lot of that budget could maybe be invested in a better way. Because again, the branded search might have happened anyway. We have the Airbnb case where they stopped all brands search and basically make the same amount of money. They just didn't pay $250 million to Google. And I mean, these kinds of cases, it's just so important that the businesses out there, at least from my point of view, they start asking that this is part of the data that they're looking at. Yeah, and even to the extent that we actually have cases where saying that the campaign tracking actually goes to the extent of one of the curses if you're sitting as a marketeer somewhere. Then you have agencies and you go there and say, hey, I wanna run this campaign. And they say, sure, I'm gonna run this campaign, but I forgot to check it. It makes grown men cry. But we actually have a lot of cases now where the clients will go and ask and say, yes, we want you to run this campaign and we will pay you this. But there are two elements to the campaign. There's the traffic that you deliver and then there's the data. And if you don't deliver both, then we won't pay you the full amount. Yeah, exactly. So you get this, that the data is part of the business case because this is what you will be making decisions on next. 100% and it's so important again to understand also incrementality and stuff like that. Where the real, how much did I invest? How much did I actually get out? That's what you wanna look at. And you need to have the data there to be able to find and figure out if that's actually the case. Because I also think, and maybe this is just a general thing, we've had third-party cookie and we've had a lot of, we have put a lot of our decisions into Meta and Google's hands. But they're basically putting their cold hand on every single purchase that you're making. And then when you have one purchase in your order system, you actually have five purchases in the media platforms systems. And that makes it very hard to make any decisions, right? So what you need as a business is definitely to know what actually drove that conversion, what actually did it. And that you can only see in your own analytics platform. Yeah, or some extension of it, right? So it's really about building your own infrastructure. And I think also when it comes to a lot of this campaign tracking saying, people assume that the, that in DA that Google know what they're doing. And I guess Google does, but they assume that in their best interest. So for instance, some kind of that annoys the hell out of me is one of the maturity signs I look at for people is, are they using the default channel groupings? So is, do you think that Google is the best to define your traffic channels? No, no, no. So, and that's kind of the first level maturity is going in and defining saying, okay, let's look critical at this. Okay, we know there's a lot of people coming from different places in the organization, but they will be coming from our channels and not just the channels that Google will box it and fold it into. And that can be like PDFs and it can be a mail signatures and it can be, we see in, it can be customer service. There's some really big traffic drivers where people are really fine to the process that you will never see if you just use the standard channel groupings. So that's kind of a huge part of the exercise is actually going in there and saying, what are our channels? How do we drive visitors to the website? Because then if you have that, then you can also add the purpose and you can. Yeah, you get the segments and you can really understand what's actually working. We have a really good, we have a case with a customer where they print their QR code on all their products so that you can go and get the recipe or like how do I fix it if it's wrong or you can help me fix it and stuff like that. And it drives a lot of traffic because again, they send out millions of products every week. So there's a lot of these QR codes in play. And again, if you're just completely not thinking about that traffic channel, as you said, then you have a soup and you can't really understand what the hell is going on in there. That's just a lot of driving traffic. But also if you just look at it, one of my favorite examples is from a customer. They had for a long time, they had really pushed saying they wanted people to download PDF on their website. And but at the end of the day, what's the PDF worth? It's not like they can take it to the bank and say, if somebody downloaded a PDF, please deposit eight kroners. So what we did to actually enhance that was we on top of all the PDFs was saying, put a big banner on the link saying click here to see if there's a new version of this PDF. Nice. And with a campaign code, because if there's no campaign code then it will actually be registered as direct traffic because it's coming from a PDF. But with a campaign code, you can actually go in and see, this PDF is still alive and people are still using it out there. So you suddenly get a lifetime on your PDFs, how they're still being used and still being supported. And you get these life signs of people coming to the site. Yeah, exactly. And we all know it because sometimes we end up reading a PDF that's quite a few years old, right? And it creates a lot of value for us and because it's something we spent some time to find, I guess on Google before. Especially, but if there was, that's a question to you saying, if there was a link on top of your favorite PDF saying, check if there's a new version. 100% I would click it. Yeah, it's great. Just to get the confirmation that it's still there, right? It's a brilliant hack. I need to move our conversation a little bit forward and let's look forward into 2025. Google is investing so much into AI and we are starting to see, especially in the US, quite a lot of results that have the AI results coming together with it. So that if you're asking a question, you're actually getting an answer from Jiminy. What do you see in 2025? What is going to happen with Google Analytics and that kind of stuff? So there's something out of the box already. So Google is investing massively on the marketing side. So what they are offering now is like, I don't know if they've been launched or still in beta, but it's just on the steps. So one of them is predictive campaigning. So go in and you will be able to set a target for your campaign saying, this campaign is supposed to generate this amount of downloads or sales or whatever. And it will actually forecast whether you're on track or not as a built-in functionality. Yeah, it is brand new. It's something we've been crying and screaming for since 2007 or whenever the first version came, right? Yeah. So, but that's super interesting because that enables you to kind of align the activity. The other thing that is even more complicated, Google will make is looking at building like a marketing mix model where you can actually turn up and down in relations to how things are gonna be performing and saying, okay, I'm increasing my spend because I wanna reach this target of the sale. So these are built-in components, but the problem is that shit in, shit out. So if you don't campaign track decently, then none of this will work. It will just be misinforming you even further than you already are. Right, so that's a huge problem saying that the foundation needs to be right for this. And regarding the AI stuff, there's actually, one of the other things coming is they call it generative analytics. So you get this GNI interface where you can sit and talk with your data and ask it questions and then have it make analysis, but the exact same problem, right? So if you campaign tracking is not in order, then asking it if you should turn up the volume of your ad words spent, then it's gonna look at the data that you put in there. And that will most likely be wrong if you haven't done anything. 100% sure it will be wrong. There'll be a lot of holes in the cheese, right? That you just can't see and understand. Yeah, but that's kind of one of my messages at the moment saying, don't assume that this is okay. Yeah, exactly. I had a good friend who was working with AI and these large language models. And one of the things he said to me, actually a few years ago, you need either a ton of data to make it any way smart or you need very good, precise data. And in most organizations, you do not have option one here, having a huge amount of data. So you need to, the data that you do have is to be of very high quality for the AI to actually being able to help you. And I've seen it, so we did a webinar with Andy Crestadena about using AI on your Google Analytics data and what you can see and get out of that. And it's really interesting and he's very good at making sure all their campaigns are tracked in the right way. So he can actually get good answers out of it. But I think for most organizations, they're not really there. So if they want to have that capability in the organization, they need to start labeling what is actually going on because that's what the AI is driving on, right? Good labeling means... Yeah, good thinking. Yeah, exactly. Yeah, and I think it's, again, people just looking for silver bullet. They just want something, okay, I don't have to do anything. And then suddenly somebody's doing my job for me. I think that it's a general risk right now in marketing, in general, is saying that, see, managers are letting their people go because their tools now have AI. But the problem is that the AI and the tool can't help you if it's not calibrated and maintained. So this entire governance thing is really critical. Yeah, yeah, definitely. So before we talked a bit about the investments, and of course, for me, it's very important because, again, a lot of organizations, they try to create a revenue engine that makes their companies spin a little bit faster. And you can say marketing is the start of that because if you don't really have any marketing or don't think about marketing, you're not gonna get a lot of growth out of your business. If we look into that, how can companies maybe start using GA4 in a bit different way? Because, again, a lot of the stuff that I see happening in GA4 is that you end up with the reports and they're very reactive and they're not very often used for any decision making, really. Maybe small changes, slushes, but are there better ways to use the data? There is, but it still then requires you to have a good level of data. If you jump to one slide, hang on, continue next one, maybe this one, yeah. So this is just an example, but it's basically saying that if you have the Google Analytics data and you build an infrastructure around it, then you can actually export it to BigQuery and you can actually start using BigQuery as a CDP or as a marketing engine to actually go and do stuff. So you get this superpower. So it's not about the data and analytics, it's actually about what the data and analytics enables you to do, but you need a massive governance structure or you don't need a massive governance structure, you just need a governance structure that is managed and that is governed. Yeah, exactly. So one of my key words in relation to this is, I use it all the time, so people are probably getting tired of it, but I hope it's a new audience you have. When I say, the problem is that very often when I ask, okay, who in your organization is responsible for campaign track? People look at me and say, Steam, don't be silly, in our organization, we're all responsible. Yeah, well, no. And that means that nobody takes the ownership, right? And having this ownership that somebody is actually accountable for the quality of data in the marketing tracking is so critical if we want to actually get these 20, 30, 40% extra out of the budget that is there if we just track the right way. The person also needs a mandate to kick other people in the, yeah. Yeah, take them all the way up and make them report to someone higher up, right? Yeah, so a mandate and maybe a budget to get things fixed. Yeah. So that ownership with those two functions, the mandate and the budget and the right platform is something that can radically supercharge your return on marketing. I think this is something any CFO that does the calculation make sense of. I think part of the problem when I'm talking to CFOs, if they're old school, they say, well, I cannot consider analytics as an investment because it's in my cost column. And that's kind of sad because it's like, yeah, but if you pretend it's an investment, then you can actually start focusing on what do you get out of it? Do people get more time? Do we get better financial results? Are we increasing sales? What is the purpose of analytics that we are actually looking for? Yeah, and that's exactly that, right? Because you're not just doing it for the fun of it. You're actually trying to drive some business results. So for all of the listeners, go talk to your CFO. Tell them to start thinking about this as a business opportunity more than a cost because I think that's one of the places that we really need to start. I had another question, but yeah, maybe I can just get it out of this one actually. No, you know what? We're gonna go further because I can't remember my question. I lost the chain of thought when telling to go to the CFO. So I wanna go a little bit further here. So do you have some good examples where you have been part of building out like an effective strategy for actually taking home some of these effects? So like a customer case or anything like that that you could share? I think one of the interesting cases is specifically if we don't look at the... So it's... Is it a campaign track? Yeah, so okay, I'll try you anyway, right? So, but we did a proof of concept for Skype where we actually looked at saying, okay, we have the visitors on the website and this retail good data quality. And then they were interested in reducing churn. So they were looking at... They wanted to reduce the number of people calling and canceling their broadband subscription. So it's a wonderful case, a very linear and we had managed to based on the data because the data was good, we could see that this was an existing customer. And we then were able to based on the behavior on the site going in and predicting and saying, okay, with 90% likelihood, this person is gonna call within the next seven days to cancel that subscription. And if they cancel, then there were two scenarios. If they called to cancel, there were two scenarios. Either you could convince them to get a new router or they would just cancel, right? So either there's a 50% chance of extending customer lifetime. But we were actually based on the predictions able to go in and say, okay, now we know that they're gonna call. So instead of waiting for them to call, we'll send them a message saying, hey, we can see that you're not getting enough bang for your bucks. So we have already shipped a new router to you. So you just need to install it, put the other one in the box and send it back. And then you're up and running again. And that suddenly moved the needle to what? 80, 90% continued that subscription. Wow. That's a brilliant case. Also because it goes in and it just takes the data and it looks at it in another context and then just the marketing side. Yeah, 100%. Now I've seen some from, and this is like a referral campaign. So I think Jessica who leads the DoorDash, which in Denmark is called Valt, she was looking at the referral traffic they're getting in from one customer that is referring another. They have this bi-directional, so both get something out of it. You get 50 Kroner and the other gets 50 Kroner. And so it's quite an cheap channel when you look versus marketing in general. But for them, they just saw that there was a huge cohort of the users. They would just go in and cheat. But they could use the analytics and they could actually figure out who is the good users that would just refer some friends and they would actually become new good customers. And who was actually cheating them and being able to then change the flow for those type of users. Right? Also the case that I think one of the boldest case I've seen has been Boost going in and saying people not being able to buy on their platform because they were returning too much. Yes, exactly. Saying you're not a real customer. You're just somebody who borrow our clothes and then send it back. Yeah. Yeah, but there's so many of these like, and it's very tactical, but it's also like, it's so important to like start thinking of it, right? And when you have the data again, it's a lot easier to understand what is actually going on. Again, if you just have a big soup, you can't see anything. You can't find your vegetables in there. Yeah, yeah. And I think the interesting thing, if you start can break it down into the segments, you can actually open the analytics talks with other parts of the organization. You can go and talk to HR about reporting for them because they have a big budget that they're actually spending on bringing traffic in. You can talk to product because product also have massive interest in relation to how are people actually interacting with my products. And we've done some analysis in relation to focusing on the basket instead. So people always have abandoned basket campaigns, but nobody abandons a basket. This is the same being both. You don't abandon the basket. You abandon the products in the basket. The basket in itself has no importance. It's the products. So we need to do the product analytics in relations to what's in the basket. What is it that's being abandoned? So these entire product journey is going in and saying, okay, we've actually done it for customers in relations to marketing saying, okay, what product attracted the visitors to the site? What was the first product in basket? But also what was the last product taken out? So actually understanding the dynamics of the products and how they support the journey. And we could see that some of the products were really good at driving traffic in, but often it would not necessarily these products that get bought. But if you just look at your data, you see, okay, somebody came on this ad and they converted. And because we as humans are lazy, we assume that the conversion was the product that we advertised. But if you advertise something with a high click price, for instance, and people buy something with a low margin instead, then you're actually- Having a really bad business. You're actually having a really bad business even though your conversion rate looks amazing. Yeah. So those dynamics, having that in there as well is amazingly interesting. Yeah. I think there's something for the marketing people on this call as well that go to speak with your finance people, figure out how do you actually make a good business case? Because I see sometimes that people are like, I just needed three X horse and then I'm happy. And I mean, yeah, it might be fine on the product where you are getting a very nice cut on the cake, but in many cases, if you're getting a three ROAS on your sale, then it's a really bad case. Yeah. Probably throwing away money, right? Yeah. So I actually, I've been using the term saying that ROAS it's the way that marketing shows that we don't really understand business. Yeah, exactly. Because you can actually have a, so the first thing is saying that in most cases, for instance, paid media is without VAT and the product you sell is with VAT. So the first, you have to cut off the first 20% of the business case. So a five X ROAS will only be a four X simply because you have to pay the tax man. Yeah, exactly. There's so many factors in there. Who's like? I never thought about that, but that's actually insane because you see it a lot, right? Yeah. And yeah, that's just crazy. Okay. I wanna talk a bit more also just because we're leading up to another webinar we're gonna do together with you and Casper Asbjørsson. We're gonna talk a bit about BigQuery. So I think for me, like one of the things that a lot of companies should really start doing is getting that data into another place. Because again, if you get it out of GA4 and you get it closer to your business, you will get more metrics that you can actually start working with. Yeah. And those metrics might be a lot more interesting than the ones you will find within the media platforms. But what are some best practices that you recommend for getting started on this? Like, do you have some? So for me, there is a lot of stuff, but the first thing I would say is, of course, opening up BigQuery and starting your export there. And for most companies, I think that GTM server side, it's kind of go and say the holy trinity right now of analytics is GA4, it's GTM server side, and it's BigQuery. So that is the holy trinity in relation to if you wanna create the value. And then of course, with the BigQuery stuff saying, by all the stuff I can add there, this is a safe place. So it's not something that is being sent to the US, it's something where I can actually identify my customer and enrich the information and do the math and have my cost data imported from you guys. And so I will have all the information there so I can actually start looking at how much is this specific customer costing me. Actually going in and building customer lifetime value models based on specific segments. So my best practice is saying that if you don't have those three things in place, a BigQuery install, a GTM server side, and GA4 with governance on the dataflow, then you're really not getting full value out of your data. There's some massive potential. And I guess this looks a bit different if you are talking PIVX Pro or you're talking Amplitude or you're talking Adobe Analytics or whatever it might be, the package that you've joined into, but they probably all have something similar. Yeah, so Amplitude and PIVX have a built-in CDP where you can go and add those information. So I think PIVX is, because it's European based, a strong alternative because you can build it in BigQuery. Yeah. Because they just have that package there. And the same with Amplitude, they also have a CDP, but I think you can still build it in BigQuery. Yeah. I see that we have a question. So I'll just look here. So Thomas is writing here, you are discussing branding and I agree with the challenges. Sten said, oh, thank you. If you want to get more out of your marketing, you need to get your tracking right. But could you also say that it just gives you even more metrics to explain to your CFO? So slightly checking the next situation, what is your advice on starting that discussion? Yeah, but I think so. So the tracking right is actually more flipping it around from a minimalistic perspective saying, if we track less but clearer than we have a better focus, we don't need to track everything. So yes, we might want to add a metric called awareness, but it's only if we have already defined a value for awareness that it's a purpose. So all this stuff that we can track is not necessarily stuff we should track. It's again, another Sten's wonderful onstage jokes saying that when I come to a customer and ask them, so what do you want to track? They say everything, right? But the problem with everything is that they have to maintain everything and that just costs all the resources and maintenance instead of development. So I think it's really, it's by stepping back and only focusing on where the business cases are and where the situations are. So the answer is, well, you go to the CFO first and ask what metrics he would want to work with and then you roll back from there. Yeah, definitely. And I think like this rolling back is a very, very good reasoning for doing. I mean, a lot of the metrics that you're looking at, let's say impression metrics or click metrics, they are leading indicators, but they are not indicating anything themselves. They are just, yes, it looks like the things that you're doing is working, but if you start putting them as the like holy grail and you start reporting as a KPI, you're gonna have big problems in your business. So you should always start all the way to the end and then go back and say, okay, what are the, if we are looking at, let's say in my business, we talk about marketing qualified leads, right? Or sales qualified leads. What is the cohort actually going to end up buying? Because that is the interesting thing here. Again, you can get 100,000 MQLs and get one customer, then it's, and if the other time you've got 10 MQLs and got 10 customers. Much better. Yeah, but your leading indicator will be completely wrong and it will challenge your business quite a bit. So always go all the way to the end and then go backwards from there. We actually talk even to the extent of the, what we call the true conversion rate, because sometimes as web marketers, we get lazy and we stop with the website. Yeah. So we had, well, Taylor as a customer, and again, so funnily enough, selling broadband. And we were like, we were doing the marketing and optimizing for getting most people, most people possible to sign up for broadband. Until we one day had a conversation with them and they said, well, guys, you are aware that this is actually not a sale. It's not a sale until they install it. Right, and then we kind of had to loop into the system, find another place in the system where we could see that now the installation has happened. And then loop that back to the marketing so we could actually see, okay. And then we actually found that some channels was really good at getting signups, but no sales. Yeah. And other were look mediocre in getting signups, but they were really good at closing the sale. Yeah, exactly. All right, so. Do we have more questions? Yeah, I was just thinking actually, because I think there's a few more there. And Thomas, was this an answer for you? Or did we get sidetracked? I don't know if our helper here on the backside can pull up a question that they find is relevant. If not, I can definitely. In the meantime, we'll shuffle. Yeah, exactly. I wanted to ask you another question on the things around, like where to start? What would be like your recommendation? Something that you see a lot of companies are struggling with. What is the first thing they should do getting out of this webinar? I think it's a philosophical thing. So as you go back to the first slide I have. Yeah. This one, yeah. So a long time ago, we talked about NIH having this concept called CSI, campaign scene investigation, where the hypothesis was that all the campaigns were murdered. So the job of the analyst was to discover who did it. Right? And I think so the philosophy in it is saying that if we have that approach, because the reason why the negative approach is that sometimes we get much more value out of finding out what went wrong and not what went well. And we have a tendency when it comes to reporting, we share all the victories, but not all the failures. But the failures is where the learning is. So hence this, even it was the best campaign ever. Yeah, the assumption was it was murdered. So who murdered it? Was it Benny the budget or Connie the copy or? And with digital, the amazing thing we have is that we 100% control the crime scene. We can track everything. So the only reason we're not solving all these murders is because we're not tracking them. It's actually our fault that the potential of getting everything mapped and tagged and activated, if it doesn't happen, well, that's because we're not doing our job to the full extent. So that philosophy of going in and saying, I can track everything. I can, if I, I know the crime scene, I can actually map everything that happens that could be important. So I can give these answers. So this philosophical approach saying, seeing that as the challenge. So what we will be talking about saying, okay, what should we do better in the next campaign instead of how good was this campaign compared to our other campaigns? Because that's not a learning thing. No, no, definitely. I think I have a customer and they have, they are very like, they make sure everyone is tracking everything because that is the value that they get to the organization. So they have a very strong culture of making sure that people are aware that if you don't track, whatever you did doesn't matter. And they wanna see that everyone that is doing something is actually also then driving traffic to the website and actually making value for the organization. And I think this is something that probably needs to be installed more places. Again, because yes, there's the analytics team and the analytics team sits somewhere in marketing, but there's a lot of people that are generating the data for the analytics team. And they are actually a different place, right? Sometimes they even outside the organization in an agency or something like that. Yeah. But I think so one of the things that we have seen working where if it's really difficult to motivate people to support the tracking, it has actually been making dashboards with, so you have this campaign dashboard where you go in and it's a big dashboard that publicly announced everybody have access to it. And you can see how the campaigns are going. And if your campaign isn't there, then it's because you haven't tracked it, but it's the campaign that your boss will see. It's a campaign that everybody will be referring to. Right? So if you're not building on, if you're not supporting the structure internally, then you cannot get credit. Exactly. And then you just need to get out of the bad excuses, right? Yeah. And make sure it's actually, we have a product for this, exactly, the Accutics Validate. And we see some customers that are really getting to a better place because they enforce it at the agency side saying, okay, if you're above 95% of your campaigns actually being tracked, you are gonna get a bonus. And through that, they can then get the agency globally to start actually doing what they're supposed to do. But I mean, when we started it out and we have countries that are like on two or 3% of the campaigns being tracked, you also know like how big holes you've had in the data and how many bad decisions you've taken because you had no idea what was actually going on. Like the murder scene is, there's a lot of unsolved murders, right? Yeah, exactly. And then there's a lot of, so the governance, the quality or lack of quality in the governance is a huge problem. Yeah. I just wanna ask, are there any questions? Okay, good. So, thank you, Sting. It's been a pleasure today. So this webinar is part of a series, as I alluded to a little bit earlier. And the next one is actually next week where you are together with another Rasmussen, Casper from Acutics. And it's actually, I might be wingmanning it, but it's mainly Sebastian from our side who will be spearheading the thinking and because he's actually one of the big brains behind the big brains, the bigger brains behind the big brain. Yeah, he's also back. So he might have something, something that he might need to play some back in the background. Anyways, we're gonna be a lot more technical in this webinar. We're gonna talk about the three very important pieces of making sure everything actually runs how it's supposed to with the GA4 and BigQuery and GTM. So we hope you wanna also join in on that. I'm also gonna do another webinar, which is outside this series, but with one of our customers. I'm joined by Brandon and Neil from Qualcomm. So the global martech director. So he's gonna talk a bit about there, how they have started using Acutics and what kind of implications that has had for them. I promise you, it will be very interesting. I did the pre-run with them the other day and it's quite nice. So that's next month, the 24th of October. But yeah, I just wanna say thank you so much for participating today. Thanks for listening in. It was a pleasure, I hope. It was a pleasure for me being here. Definitely. So I hope you got entertained as well. Yeah, thank you so much, Stim. And have an awesome Wednesday and see you all very soon. Bye bye.