Legendary Analytics with Steen Rasmussen.

Hello and welcome to E Commerce Growth Stories. This is your hostess with the mostest, Juliana Jackson, aka the CLV lady. And also I am the Chief Evangelist for OmniConvert.

And this is Valentin Radu, CEO of OmniConvert. Thank you for tuning in today for a new episode of Ecommerce Growth Stories.

Do you like snooping around? I know I do.

I don't know. Did you ever wonder what so only.

Uses to power their e commerce business? Or Coca Cola or Motorola or other big names in e commerce and retail? Well, today the secret is gonna be revealed, and I'm talking about Vitex. Vitex is the first and only fully integrated commerce marketplace and OMS solution that offers fastest time to revenue and no upgrades like, ever. So if you're an e commerce brand and you're looking to to find a software that can really become a powerhouse to your business, providing you the best go to market speed, flexibility, daily updates and reports, and overall platform stability and agility, you really have to check these guys out. So make sure you go to visit vtex.com that's vtex.com today and give that.

E commerce business a bo. Hi. It's Thursday. We are here for a new episode of Ecommerce Growth Stories and I have an amazing. We have Valentin. My bad. We have an amazing guest. It's me, me, me, me. Yeah, I know, I know, I know I'm selfish. But we're here with Steen Rasmussen. He is the founder of IIH Nordic, and this is a tram passing down the street. As you guys can see, I am in my kitchen, but I'm warm. You know, I'm here giving Steen a warm welcome from. I just cooked some food. So, Steen, welcome to E Commerce Growth Stories. Thank you for being here.

I've never been invited to a place where I felt more at home.

There you go. I have some warm Mac and cheese for you right here. I just did it. I'm super happy to have you. It's been crazy how we just met, like maybe a year ago officially, and we have been seeing each other on Clubhouse at Webanar. You have. You've known Valentin for a long time, and I was jealous on your friendship, so I had to make my friend as well. Super happy to have you here. We're always asking people when they start, when we start the podcast, how did you get into E commerce? But in your case, I know you're like the typical. You're not the typical. You're the atypical data guy because you have interest in networking and socializing and commercial stuff in business and marketing. So you're not the typical one. And I want to ask you, how did you get into data analytics?

It is the classic a guy happening to be at a place at a time, right? So when I studied, I finished university back in 95 and I had a degree in corporate communications and that was fine. So in 99 I needed to do something else. At that point I was the head of PR event for chain of movie theaters in Copenhagen. So absolutely commerce related, not a lot of E in it. We did build their first website, but I, one of my friends had just started the first, one of the first agencies in Denmark and I actually moved in as an online copywriter and that was kind of my introduction to the digital world. But the interesting thing was over the next couple of years I started moving into usability and that was horrible because I had done this master thesis on commerce and business and what I saw in usability at that point that I still see in usability old school. It's like if you make it user friendly, nobody will necessarily buy. So it's the business case for the users but not for the business. Right. So in those days, and I think it still is, if you do usability, you're kind of the ambassador for the users, but nobody's the ambassador for the business. And I think that annoyed the hell out of me saying this was a. Because at the end of the day all these digital activities were an investment. So no matter if you're in E commerce or whatever, you're in E commerce, it's just the purest form and the easiest way to see the return. So we moved in and started working with data and I started, actually the first stuff I did was a lot in the line of content and conversion rate optimization. So that's kind of my angle. I remember reading the first white paper back in the days where Brian Eisenberg, that was the first place I came across it used the term conversion rate optimization. I actually think it was the white paper where they coined the phrase so. So it's like we were working with it before that. I was like, oh yeah, this is what we're doing. And it was so simple in those days. I think that that's kind of the, the beauty of it. And this was in what, 2003 or something? So. So in that day it was just, yeah, you needed to optimize your conversion rate. And that was kind of, that was big juju back then. It was really amazing being able to, oh yeah, you can improve the conversion rate. Oh, my God. So that is how I ended up doing it. Then I moved more into data and more on a general level. And as you know, things just got complicated. Every day we add another layer to the complexity that is online business.

It's true. And I remember. I know about the white paper that you're talking about because I was ashamed some time ago by my friends that I didn't know who Brian Eisenberg is. And I was very ashamed for a whole month about that. So now I know. It stuck to my head. I always thought that conversion rate optimization was invented by volunteers. No, but I think. I think it makes sense. A lot of people that are doing CRO when they're dealing with website conversion and optimization, I think it's a natural switch towards more complicated stuff. It's the whole mentality of experimentation and testing and understanding what happens behind the customer behavior that you see on the website. You have to be ridiculously curious to be able to see. And this is an.

You just muted yourself. Okay, so from my perspective, I think one of the things I've learned by working with conversion rate optimization for a long time is that it's. I think it's a lot more. So right now, my theme of the stuff that I'm working with is very much about data needs context, or it loves context. And one of the things that I found is most of the stuff that we're doing on the website of optimizing flows and changing prices and doing paths and stuff that is only like a quarter of the stuff that you have to do to optimize your conversion rate. Because very often the main part of the online sale is it's out of your control. So you need to be aware of saying. I think I've used this example with you before saying. So you're sitting there and suddenly your data tells you that, oh, my God, our most popular sneakers are not selling anymore. Call in a conversion rate expert. And you call him in, and he looks at the data and he looks at the side and say, oh, you have to change this, this, this, and that. And this button has to move here. And you have to do all this stuff. But the problem is, it's because the competitor is running a campaign on your sneaker, right? So it's actually. It's such an obvious problem. There's nothing to fix. There's only a thing. Maybe we should tone down our marketing in this period until they run out of money or we should change the price point. But you cannot fix a Market problem with optimization. Right. And if you keep trying to only focus on the data that you have in your bucket, then you're not adapting to the world, you're just trying to force your data to do the stuff that you wanted to do. Does that make sense? That's just one part of the thing. The other part is saying what we see also when we say, people come and say, oh my God, our conversion rate is tanking, you need to fix the data design. But then it's like, no, it's because your tech is rubbish, your site is too slow, there's nothing to fix. It's just because you're in the wrong service. And these customers are actually coming from Norway and they have horrible bandwidth. Right. So having these dimensions and looking outside just the website, because we have this, this really narrow focus of what we can see on the screen and this is the entire, this entire, is the entire relations with the customer. I think we need to look outside that and say, what is the context? What is going on? Valentine, you're sitting agreeing or.

No, I was just wanting to add that context is great, but I think many data analysts are lacking the some of the greatest tech that they have available, which are the questions, the answers from the questions. And recently one of our data analysts had an aha moment when he suddenly said, come on, what you have here with these freaking questions is more impactful than my whole work in the last three weeks. That happened due to the fact that he could finally understand what was going on after blending the insights from the funnel analysis and the customer journey with answers to some questions. So mainly it was they've blended jobs to be done answers. They've blended the jobs with the qualitative research by understanding who's the ICP and then with the behavior on the data. And I think that is that you need the full story. So maybe what you're saying, and I'm totally with you on this, is that unless you don't have the context, you are trying to fix things, but you're missing out the 3D layers. Let's say you're looking at things and you think that it's a B dimensional thing, but it's actually having other, other dimensions.

Yeah. And I think one, one other aspect of that is, is if you, if you look at it in a greater context, right. And, and also say, I think we, I think we touched upon this before.

Right.

Sometimes the biggest source of insight is actually the people who don't convert. Right. It's the people who fail saying okay on even on the best of website, you have like 16, 15% people buying, right? And so it's not like the other 85% was just sitting there saying, okay, let's waste some time, right? They have an agenda. There's a reason why. One of the things that we have found out is saying people basically don't waste time if they're on your website, they're on your website for a reason. And then that's kind of the key thing. So if, if they come there, then they, the likelihood of them buying somewhere else or coming back and continuing the journey, understanding that process, it's really critical because saying if the majority does not end up buying with you, why don't they end up buying with you? What is, what is wrong with your offer? And especially if your offer is something generic saying, yeah, you're selling the same sneakers as everybody else, you're even cheaper. Why the hell don't they buy from you? Because then there's something else to the story that you need to understand that is not on about the price, it's about the value.

Yeah, the whole, yeah the whole, the whole context of it. And I think the customer journey optimization, I guess narrative is becoming more and more adopted and broadly adopted in general. And this makes me want to ask you a question. So because you are one of the first people that adopted the conversion rate optimization, you're amongst the first guys that actually further, I want to ask you, how do you feel about how conversion rate optimization is today and is it, is conversion rate optimization enough to sustain a business long term?

It's a trick question and because you kind of know the answer and I would say no, it isn't because what is happening right now is the prices of marketing has gone to a level where the single conversion cannot pay. So you need to think beyond that and look at customer lifetime value.

I said to not have for that.

So you need to add the pieces together because that is where the value is, right? So you need to look at the full journey and not just the single conversion because in most cases this convert and burn, just saying, yeah, I sold something and then I will be hunting for the next person to buy. And because I'm so amazing, they will naturally come back. I don't have to check or anything. Of course they will keep coming back. But for most companies these days, getting a profit on the first sale is not a given. And especially if you go through several marketing activities saying, okay, they click on three AdWords and they click on a banner and you also have to share with an affiliate. It's like, okay, that's not a lot of profit you have depending on the product, but.

It'S real and it's very hard. And a lot of people are aiming for this, you know, profit on the first purchase and very misplaced focus and interest because it's really, really hard to get it. I mean, of course, if you're manufacturing a product that costs you $5 to make it and you're selling it with 20 and you, your ad is 10, okay, maybe you're making some profit, but that's not how the business world works. And which, which once, you know, makes me want to ask. And I promise this is not a trick question when it comes to data in general. I mean, this is something that we've been, you know, I think touching upon and I talked about with June as well, because she's here. So. Hi, June. I asked her like, how do you feel that E commerce and retail is quarantined, disappeared. How do you feel that E commerce and retail is embracing data? This is the same thing I asked Simon. He came to internal AMA and I asked him, like, how do you deal with people that do not want to look at the data? They don't want to see, you know, the reality. And there's like, let's be honest, there's companies that don't track anything and they're on mountains of data and then they're going, going ahead and doing the same mistakes all over again. So where is the space and the position of the data analysts inside the modern E commerce and retail company, the direct to consumer companies?

I think it's not complex because it's really about to be a valuable analyst, you need to have, you, you need to have a value for the organization. It's one of the things that I keep saying, saying. So why did anybody, why does anybody hire us as consultants? Why does anybody hire a digital analyst? Right? And they hire them for the same reason. They hire them because. Well, there's two reasons. The, the right reason is they hire them because they are a good investment for the company and they will make the company earn more money. Right? And that is the value of the data. There's also the ones that hire an analyst because they're supposed to hire an analyst that's kind of like a trophy wife, your trophy analyst, that you have a analyst on your staff. So you can say, oh no, no, we have three analysts. It's like. But they're not doing anything right. So they're just sitting there being, being fancy and being brought up in meetings because they're so valuable, it makes the company look good.

Like.

Oh, exactly. Yeah, yeah. We have a data department. We're not using it for anything. Right.

Yeah. We're not listening to them. We're just giving them. I don't know, we're asking them to give us random segments. June said here in the chat, she says the data people need to internally market themselves and what Steen said show their value. And I think it's true. And I really don't see too much data people really being out there in the open. I mean, there's you, there's June, there's Jim Stern, you know, with Simo, of course. But there's not too many people that make it really cool to be a data. Like our data guys. I mean, Adrian, we have two data guys. We have Matei, which gave that book to Valentin, and we have Adrian. So Adrian is much more out there and he will post some stuff, you know, on social media and make it cool. But, mate, I don't think we will ever see the day that he's gonna go on LinkedIn and talk about his, you know, trophy. Yeah. And I think this is a great point. I never thought about it like this. By marketing the data people. I love this. I'm gonna say this is a great.

Point, but how do the companies see.

The value besides the trophy stuff? Like me and Valentino, we are shocked every time we have these calls and we're coming with data and insights and we're saying, look, this, your business is in deep shit. You're losing money. Save yourself. Now we're making money from ads.

Yeah. I think part of the problem. And it's actually going back to another thing. It's the name calling somebody a digital analyst. It sounds like you're the accountant. Right. You're just a guy who's gonna deliver a report. You're not gonna have any opinion about it. This is the numbers of this month. So a lot thing. And this is why June and me and other commercial people actually can get to tell a story here. Because the story is not about the data. It's about the business. And it's where we overlap. Right? So saying if you have a lot of data and you're not using it for anything, it's not about. It's not the one who has the most data when he dies, who wins. It's the one who have used the more the data they have best during their lives. Right. So kind of flipping that and saying the naming a digital analyst basically sucks. It should Be data driven business developers saying it's going in, finding something to develop the business based on the data. Because analyst is so passive. It's like it's the safest job in any organization being the analyst. Because it's like, like you said, you.

Have a truck, you don't have to do anything. You just need to analyze and to send reports and that's it.

Yeah. So when somebody comes and asks you, so what was our bounce rate last week? Hang on for a second, I will do my mojo and magic and I will conquer this number for you. And it's like, oh, that is so impressive. Yes, I know.

Yeah, we're a treasure hunter. And I have to say something about it like three, four years ago ago, I've been into a meeting with, there were like three data analysts. There was our data analyst, there were all the heads of the departments and whatever. And there was a mess over there and those, those guys at some point, I mean our data analyst identified like a huge flow in how they reported the data. Like a thing regarding an average, at an average. And it was a mess. I mean you couldn't rely on the data. And one of the data analysts over there said that's how we've got the brief to build a report. And he was like justifying it by the fact that the business asked for the report and he wasn't even bothering to verify this. And I think that's the danger zone for a lot of data analysts because they are either, let's say on a routine, the business asks and we deliver, we ship these reports. Tell me how many customers that did that and that and the other are from Greece and that's it. They simply go, they've made that query and that's it. Maybe they do like a nice graph over there or a animated graphene flourish, but that's it. And I think the job description of them being like treasure hunters or gold diggers or data driven people, business developers, developers that could help them because that label, that new label, that new role would help them. But to be honest, a lot of companies, once they reach a level where they are, they afford, they hire an army of people that are not being trained to make the company thrive. They are being trained only to deliver numbers. They throw numbers everywhere. Numbers, numbers, numbers. But.

I just came from a meeting with a potential customer, right? And he said he was so tired.

Because.

They had no alignment on their data. So he had just come from an hour and a half meeting with the marketing and sales team and they had basically been sitting Presenting sales team was presenting numbers based on less click attribution and marketing team was basically presenting numbers based on first click attribution. And there was no alignment. He was just sitting there and no, nobody was in agreeance of anything and all the numbers were different. It was like. So nothing happened, it was just the number porn, right? Is it like just throwing numbers out there and saying my number is bigger than yours and doesn't really providing any value and absolutely nothing actionable. So it's same situation. But I think part of it is also because we know in analytics that the data sucks. We know all the flaws and all the holes in the data. So we think it's supposed to be really scientific and something that we could defend a master thesis on at mit. But that's not the data we have. We have some numbers that if we're lucky show some clear tendencies. Right. And then that's what we have to act on. But once we try to do this and somebody saying so how come this number shifted power by 0.5 decimal points the other day? Who knows, dude? It's like that isn't the point because. So I think for me, what I try to teach people saying, except that the numbers are always wrong, but they're always equally wrong. So the trends are correct. Yeah, because so you have this, you find this in meaning. So somebody, this, this guy, the cool dude working with emails, right? And you go in and you present your numbers and blah, blah, blah. We had like 90 people sign up for emails and he like hang on, that can't be right. I can only see 72. And then it's like suddenly it becomes a precision game. And that's not the quality of data we have. It's absolutely not what we can deliver. We can deliver trend saying yeah, but the trend is that it's upward going and we're actually doubling our numbers week by week over the last seven weeks. So what we're doing is right, not going in and discussing. Okay, so who are these 7, 72 people? Who gives a sh. Who cares, right?

It's cool. Steve, this is how I like you unleashed.

Yeah, yeah.

This is how I'm used to. So I want to ask you something. So do you think the term data driven has made a lot of consultants or technology companies have like ginormous balls and go ahead and people just for the sake of being data driven.

Yeah. So I think actually where it really started was in programmatic advertising. I think at some of the levels, if you read some of the stuff online in relationship to People deconstructing programmatic. I respect Programmatic a lot, but there's a lot of black boxes there. Right. And the challenge with any black box is that people can tell you anything. It could be really hard science, but it could be a really lovely Ponzi scheme, basically. Right. So you don't know what's going on. And if you have like a lot of the programmatic stuff is you have like programmatic and programmatic and programmatic on layers. So you buy from somebody, they buy back. And so the lack of transparency has opened doors for a lot of people to deliver amazing services that might not be as amazing yet. And it might be that they're trying and it might be that it coming. There's this. One of my favorite apps when I'm traveling is a company called Vivino. So you can scan the layer the label of a wine and it will show you what is the price of this wine other places. So you get this fixed price. But in the beginning when they did this, now they have a lot of data and people are scanning it and entering prices and what they paid for it. In the beginning, it was when you scanned the image, there was a dude sitting, trying to Google what the price of this wine was so he could give you a data. So you get this. This feeling that this was a. This was the service they wanted to build. They just hadn't built it yet. So. So the dream was there. It was just a dude in the meantime, while the real dudes were trying to get financing to build. To build the engine. And I think that is the place we are right now. There's a lot of dudes sitting out there where that has been sold by doing amazing stuff. Right. And then honestly, if you're on the client side, and I think actually now we'll go back to if you're on the client side. This is really hard to tell what is what what is what is a good service. What should I expect? And I see it and I actually think I talked to. I talked to June about it at one point in relations to. I got distracted because I read the comment Valentine. Yeah. So in recruitment that we have this, people who are recruiting people in our sector have a tendency to be goddamn amateurs. I put it out there. So they will hire somebody, but they will have no idea how to evaluate their skills. Right. So they will hire somebody who comes from another big place that hired somebody from because they had something else going for them. So how do you. If you're sitting in HR or if you're sitting in purchase and going to buy a tool. How do you evaluate a tool that is supposed to solve a problem for you if you have no idea what the problem is to be solved? It's really complicated. So it's like going to a restaurant and then the chef comes and you say, yeah, I would like the omelette. Okay, so. So. And the chef comes to your table, say, okay, so how do you usually make your omelets? And she's like, that's not my job, dude. You make me the omelette. And that's kind of the challenge here that, that we, we buy services from a perception that people are able to deliver them, and then they need to show us the money. And I think now I'm ranting, but we see a lot of clients who have been burned badly by buying tools they didn't understand what they were supposed to do with. And it's partly. Could be a wrong tool, could be a wrong team, could be a bad implementation, could be a bad staffing. All these things can turn a tool into a turd. So I think what we see when. When we talk tools, it's really about training the organization to use it for value. Because if we don't, and if we don't have this focus on embedding the value in the organization, then in a couple of years, the tool will be gone. Because, yeah, it's that tool we bought, and now we're going to pay for it again. But what is the return? Yeah. Wasn't it something Jim bought back in 2017? Yeah. Okay. But he's gone now. Okay, let's get rid of it. Because nobody knows what it is. Right. So it's not the tool, it's just the team that hasn't been. Yeah. And then for me, I think this is a thought I had the other day. So this is my first rant with it. I saw an article on LinkedIn that I sadly missed, but it inspired me, saying, maybe managers do not need to be able to work with data. And I think if so. So I. I didn't get the point, but it did trigger a thought. It triggered the thought saying that maybe managers don't need to be able to work with data, but what they need to. They need to be able to recognize crap when they see it. Right. So. So they don't need to be able to understand the data. They're just able to. To. To call out and say, well, that's wrong, or that sucks or that's bad. That's not science. That's not statistically Viable. They don't need to be able to go in and do an analysis. They just need to be able to call bullshit.

Did you hear that, Valentin?

What?

So I'm gonna announce this publicly that I, I am, I'm not a big fan of Excel. I know everyone here listening sleeps with Excel in the bed. I don't. And we have like this crazy report thing that we're doing and I hate it and I lose interest because I, I'm very good at finding very fast. So I have like this detector. So you couldn't have been more timely with your rant. So go ahead. This is why I throw it to Valentin's.

No, no, I just think it's important because otherwise if the manager is able to work with the data, then he can do the job himself. He just needs to be, he needs to manage. You saying that. Okay, this is bad, dude. And you need to do it again because this, I can't present this upwards.

Yeah, what I think is the context here in, in what you've studied is that many, many managers are very stubborn into rigid numbers. But they don't even mean that. They don't even understand if those numbers lead to real progress. So they are looking at some lag indicators without understanding the context. Like they don't see the micro conversions, but they are looking to. For conversions on the other look side. The, the other side of, of it is that many people that are, let's say, tracking numbers, they don't know what to do in order to improve those numbers. And I think if we are talking about the relationship between the person which is managed and the manager, the, the problem that they both have is that they don't know what progress means. The manager is forcing some numbers. The person which is managed is accepting eventually because they have this, let's say, hierarchy, which I think is going to disappear in 20, 40 years from now because the value is not. I mean, for a company, the manager should be more valuable, but is not. I mean, we have is like in a football team, right? Maybe the coach is the best person, but maybe he's the goalkeeper or maybe, you know, I mean, you don't know who's the best. So eventually they are all necessary. Like in our orchestra, yeah, you need the guy at the piano, but also at the violin and the drum. I mean, without all of them, you can't freaking play and that's it. So I think the problem with this hierarchy is that they are not looking at numbers so that they can understand the progress behind those numbers. And, but progress should be measurable, no doubt about it. Because we in a company, we're not here to, let's say, do anything else than to make progress. And if you can't measure it, then you're not improving it. And yeah, you and what doesn't get measured doesn't get managed. The problem is that we have too many metrics and KPIs and blah, blah, blah. Eventually we're not making progress or analysis paralysis. But you need to find the right balance between how many numbers, which of the numbers are the lead indicators and how many lag indicators you're after.

But also in that sense, it actually goes back to another thing saying. I did a presentation a couple of years back with another Danish guy called Worst Case Scenario where we went in and we were tired of every time we went to a conference it was always somebody telling us, yeah, we came and we did something amazing and everything grew by seven gazillion percent. So we wanted to do a presentation. All the stuff that could go wrong. And actually the first basic thing that will go wrong is not having the, the same nomenclature of things. So you don't have, you don't have a definition of what things is. So my lead question to people, if they want to, when they talk about conversions, it's saying, okay, cool, so is that user conversions or session conversions? Because most people will look in Google Analytics and they have their conversion rate from Google Analytics and that is providing them session analytics, session conversions. But if my purchase process on a site is, hey, I come 10 times, or the average customer comes 10 times to this site before they buy, then the conversion rate based on the sessions can never be higher than 10% point. Right? So, so there's nothing to optimize if we're there and we're looking at session conversions. But if it's the user conversion and we have 10%, then there's something to work with. Yeah, people, they come 10 times and out of that we have 10% who converts. Just this silly number that we work with all the time saying people like, what are the two definitions of this? It's like, yeah, because. And then most people are used to the Google Analytics one, the session conversion, which is not very operational. It's saying of the number of people who walk through the door today who bought from a B2B perspective. I actually tried to talk to the developers at Google and they're saying that Google Analytics is really a piss poor true when it comes to B2B because it doesn't support complex purchase processes. It's A very linear and simple tool. G is better because you. It's much more about tailor making things. But you need to be able to focus on the user and not on the session. And this is kind of the stuff that you guys do saying it's customer lifetime value, right? Because session lifetime value doesn't really give that much value. So.

So, and I love what Jif go Google Analytics 4 is doing. I had the. We have the. I had, I had actually because I think Valentin was not in the, in the office that day. I had a great session with June and we're talking about Analytics 4 and you gave me a great segue in a conversation. So I mean, obviously you know that with Analytics4 the conversation became. The conversation started around predictive lifetime value and our reference segmentation and netting and you know, billing and everything that I like a lot. I don't want to talk about that though. But I see that, you know, advertising because of, you know what happens with privacy because of iOS 14.5, because iOS 15 now with email, a lot of people are misusing and leveraging the, I guess the concept of predictive customer lifetime value and predicting customer lifetime value. It's a pre. It's a difference there. And I want to ask you, how do you feel, and I'm choosing my words very wisely right now. How do you feel about companies that can claim that they can help you with predicting future customer behavior if they don't have access to all the data? Because this is how we do it. We don't have access to everything, obviously. We have access to post purchase data, historical data. So to be able to predict correctly customer lifetime value, you have to mix it with what's happening in analytics, you have to mix it with what's happening post purchase. And it's a whole combination of data from multiple sources, not foreign data, data that you have to be able to predict something like it's not something that we just click our hands and drop it out of our software. But then, you know, in E commerce you if you use some buzzwords like cov, predicting cov and high order value of customers with high order value, people just get excited. But I don't think there's no one that explains the mechanisms of what predictive customer lifetime value. And they're defying the laws of data science and data analysis. So why don't you just. Because there's a lot of people watching on LinkedIn right now, by the way, we got a lot of comments there. So if we're on LinkedIn and you know, this is something that my following likes. Would you like to walk us to what predictive lifetime value is? And how do you correctly create an audience that is informed by customer lifetime value and future customer lifetime value that comes from you and not from Juliana.

Yep. So what I think to break it down first, right, so what we have is historical data that we can try to write forward and that is also. That means that, yes, the predictive part being turned into a qualified guess. Right, because we're only looking at the historical data. If we travel back two years in time pre Covid and we took all the, all the predictive models in the world and say, okay, what are sales going to look like next year? Probably maybe 0.02 of them, which were actually broken at the time, would have predicted the market development. Right. So we are only looking at historical data and we cannot take shifts in the market going on forward, like going back to what I originally talked about with the external factors. So predictive customer lifetime value is. It's a very good qualified guess. It's not something you can just do. You need some data points to trust. But it is saying. And I think the reason why people want to say they do it is because it sounds so much more convincing than doing qualified guessing. Well, I'm the best qualified guesser in relations to your expected future earnings. Doesn't sound really impressive. And consultancy. So you have this and it's kind of predictive customer lifetime value. Right now it's kind of the. It's the new attribution. Everybody talked about attribution, being able to go in and look at the data and break it down and say, yeah, this is exactly how much worth your AdWords is supporting your business. This. And then for me it's. And then they always use this lovely football analogy saying, yeah, you want to know who, who played who. It's not about who scored, but how many times the midfielder actually supported the gold scorer. And the challenge is that that's nice if you have a closed environment, if your website is living in a bubble that has no outside forces. But in reality, this football game is being played on the deck of an oil tanker sailing in a storm at night, in a thunderstorm. So you get these fragments of who has the ball and what's going on. And at the same time, you have like a tennis game going on across the field.

No, we've lost. Thank you. We just asked you for a few seconds.

Okay, so you missed my entire rant. That's always it.

You were talking about the tennis court yeah, yeah.

It's just saying that attribution. Everybody has been saying attribution is easy and it's not. It's, it's probably. And now I've been saying attribution is the hardest thing to do or precise attributes or qualified attribution is the hardest. But I would say predictive CLV is the hardest to do. So that's something you get out of box. There's a reason why Google had a attribution tool that you paid for. They bought for a million, multi million dollar amount that they pull from the marketplace because it wasn't precise enough. They couldn't trust their own data. They couldn't sell this tool to anybody else. And now the complexity of this is actually saying, okay, to do predictive customer lifetime value, we actually need to be able to have our attribution layer so we know what the cost of the marketing is. And then on top of that, we're going to predict how you will respond to this in the future. Right. So having it in a box, it might be again, the dude sitting inside the tool being really good at guessing. Right. So. And it might be that they're absolutely right because the. I remember when we started working in the, with some travel agencies, they told us they had a competitor in the Sweden. They had this was back in what, 2002? Three.

Right.

They he told where it was. One of their competitors had a guy in Sweden who was an autist. So, so, so he would come in and he would open the papers and he would skim through all the travel ads and he would afterwards say, okay, these guys have raised the prices, these guys have lowered the prices. So he would just by going through the ads, people being able to give them a lowdown on what was happening in the market. And it might be that guy doing the predictive clv because it takes a lot of data or a lot of thinking capabilities to be able to do it is really simple, just linear progression saying, yeah, okay, if this happens today, then in one year this would have happened. And that's kind of too easy.

Yeah. And it's like everyone rediscovered lookalike audiences and they're trying to make it sound like it's some, I don't know, some fucking new shit on the, in the market. But it's something that's done for, you know, as far as I used to sell shoes in a shop when lookalike audience, that was my first job. I was selling shoes, but I was working in the call center with lookalike audience to start. So it's like, it's things like this that infuriates me in a way because I think education is becoming more and more important and I think all of us in the space that have any idea about what's going on, I mean, no one knows exactly everything. No one knows exactly, you know, how to. I mean, we all are doing our best, right? It's frustrating because I know like what we're doing for an educational point of view. You go and you try to educate these companies and the brands and the consultants and then you see people that just come up with some random shit overnight and they're like, what did I do? You know, like, how do I undo this? And it's very hard. And I think all of us should do, I guess, more efforts into educating the market when it comes to these metrics. Because cov, right now, when me and Valentin were talking about it, we got one like on LinkedIn from each other, but now, you know, like two likes. Yeah, exactly. And now everyone is talking about customer lifetime value, but they're misusing it and they're making people distrusted because customer lifetime value in the hands of the wrong people is damaging as hell. And even those customer signals that you have that you think they're VIPs, they might be VIPs today, but were they yesterday? And this is the event based thing that happens with segments and groups, but people don't think about that, they just get excited, you know, oh my God, I got 100 VIPs, but were they VIPs? So all these brands and the brands that are listening, like when you're working on this customer segments, make sure that you ask yourself the questions, where were this customer journey yesterday or today? So, yeah, I'm sorry I had to go on this. I want everyone to hear it from you. The lifetime customer lifetime value, as you know, for me, is a very deep subject. And on another note, when are you going to come and teach a course in our academy? Steve?

I haven't been invited. Every day I open my mail and look for an invitation. Every morning I have to take an extra cup of coffee because I'm so disappointed it isn't there yet. So you tell me when I.

He's on the way still.

Okay, good, good.

I wanted to ask live, so you cannot say no to me now. I have people on LinkedIn here. Let me, let me just see if anyone wants to ask something because I got a lot of comments here. Oh, everyone, like the trophy analyst thing. Someone says, interesting conversation, happy to listen to such practical advice. Indeed, application, communication, actioning is much desired. One can see so much talking of data because one is supposed to the whole formal thing. Someone is asking, what's the convert? What's the conversation you3 want to start as you move further? That's a very good question. What conversation do we want to start here for everyone? That's a great question from Andrew Lopez, I think.

So what we all suffer right now, and this is the entire privacy issue, right? And that is kind of the big conversation that we need to have saying, what happens when we have less data? Because what is the next step in data? And then this is where you might be able to go and do some predictive customer lifetime value. So one of the things that I've been looking into, and this is actually, is this virtual users or simulated audiences, right? So you actually go in and you create a horde of users that will interact with a website and create data that you don't have based on user behavior in general. So I think this is kind of the next frontier that we need to be pushing because this, this is saying then we might be able to actually start doing customer lifetime value in the long run if we can get the simulated audiences to actually go through scenarios faster, saying, okay, cool, if this is the behavior and does that make sense?

I know you give us. Valentin is probably thinking on how to develop it right now. That's face.

I know Valentin, he's already thinking, okay, maybe it's about time to market it. He has it in the door.

No, but it's a great idea because you put something to learn this behavior, so then you know how to tackle it. And I think machine learning is going to become more and more of a normality to customer journey and customer experience. And maybe then the predictive lifetime value will make sense because even Google says the same thing, like, you need to do it either probabilistically, which is how we do it, and then you have to do it through machine learning. And the machine learning part takes a lot of time and it needs to learn from something to be able to determine things in the future.

So yeah, it needs some amount of data. Stina, I want to run something with you if, if you think that the, the market at this moment, the whole landscape of tools and companies and whatever, we all agree that the education is the main problem. And that's because the formal education is far behind what's happening in the real market. A lot of spending managers think that they are marketing managers and E commerce managers. They throw money at the problems and they switch jobs if they got caught and whatever. So we all know that this happens. And I don't know how you think that the education will look like in, I don't know, in 10 years from now. Do you think the formal way of getting trained at the job, spending the money that the company allows, the budget that the company allows you to, to, to train with, this is going to be the norm in the future as well? Because I, I think there are only losers in, in this game. On one way on, on the employees are not reaching their full potential because they don't know what to do with the thing on the other end. The companies are, let's say affected because they, they, they, they waste a lot of time into fixing problems or they simply die. If they are startups, they, they can't really survive, not mention thriving. The big companies, they, they don't even think what they need. I mean they have, there are a lot of managers that are completely illiterate about the future and they come up from a different prehistorical reality, but they are there and they are the hippos which are deciding over this, that and the other. So I think we live in a broken world at this moment. It's for, it's no doubt about it. But, and we, maybe we all agree that education will fix it. But how do you think this is going to be fixed in 10 years from now?

I think so. So maybe it might not take 10 years because I think the competitive companies will actually have a massive benefit in the sense that they. One of the wonderful things that we've done right is we've gone to a four day work week and so people with IH only work 30 hours and get full pay. So the idea is saying that what we do on some Fridays is we have like Learning and Innovation Fridays where everybody come together and learn. We're not allowed to work on clients. It's all about learning. And the same idea is saying that if we're able to do the work we're supposed to do on five days and four days, then some of the companies that we've been talking to is saying they're not going to give the employees the fifth day off, but what they will do is they will make sure they use that day for learning. Because what happens in a lot of organizations is you go out and you spend time and you find an amazing talent and you hire that person and because you always push that person to deliver, the person no longer becomes the best in the field because there's no time to get updated. There's no time to learn. There's no. They burn out, they get outdated, so you fire the person again because they're no longer relevant and you let them go. Right, so. So I see this as learning built into the job, like you said. But. But in another way where it's much more systematic because it needs to be built on a. On at least a company level, on a common denominator. Right. So, so what we talked about saying that what is the conversion rate? What is the goals that we're trying to achieve? What is the point of the activities that we're doing, having that as a shared basis and working from that. And, and for me, one of the big curses of data and what we've been doing, if we just look back, Valentin, the time that we have known each other now saying the curse of that period has been the same. It's actually escalating a bit. It's the lack of continuity and activities that we're always trying new stuff. We're changing the data model, we're changing the setup, we're changing all this shit all the time. So we keep resetting the data that we have to a new reality. And I think that is what we need to change if we want to actually leverage the data we have. We have to refocus on continuity of the data and we have to refocus on learning it together and working in that way around. That was very vote stain.

It was very deep. It was very deep because you're right, we never take time in our lives to analyze what's happening to us. And we're always trying to change stuff and improve. But sometimes the lesson is in the things that we, you know, set to the side with us, to the side, because we want always new data. We want to.

Yeah, yeah. But we get this idea saying that, I think for all of us, every time we wake up, we find out shit. The world is. Is more complicated than we thought. When we went to bed, some new problem, some new challenge, something else has changed that we need to take into account. Imagine that the first one, when I, going back to the. When I first heard about the conversion rate, it was just saying, well, the conversion rate was the number of people who came to the site and how large a share of them who bought anything. That was the original conversion rate. And that was when people were optimizing for how can I get more people to buy from myself? And we know now that that number is so wrong. And in 10 years, what we're doing now, people will think, oh, my God, were they like cavemen or something. What's happening here?

Yes, everything is constantly, constantly updated. That's why, I don't know, like the Windows Server, you know, like you have your Windows installed and every day it's a new update on that Windows.

Yeah, yeah. And I think it's the same what we do, it's just accelerating. So to answer your question, I think we will also in 10 years all have a virtual assistant that is actually doing our analysis for us. So one of the things that Henrika and I'm my partner at IH talked about a long time ago saying nobody cares about analytics, we should have ask analytics. So asking a question. So you would come in and it would be like on your, on your saying, hey Alexa, how many sales did we have last night? And then Alexa would, would answer you.

I would love that, Steve.

Yeah, because this is our first click. Yeah, yeah, yeah, yeah, exactly. So yeah, okay, cool. That's a good question. And then you can move on from there. But having this conversational thing saying this happened yesterday and I think this happened because there was this problem and you need to look into this. So instead of logging into anything and sitting looking at the numbers, somebody would always have already have pre tuned that knows your business objectives and your thinking and where you want to go with thing. So that is kind of where I see. And this is not far off. This is kind of Google Analytics has this, this question bar in the top where you can actually say how many sales did we have in China last night or last week? And they will answer as a question. So this is just easy. Yeah, it's, it's up close. But, but when I show people this function in Google Analytics, like oh, is that new? Yeah, kind of. It's only been there for four years.

Yeah, exactly. Well, it needs some fine tuning. I mean at the beginning their insights were pretty bad. Those got better.

But actually it's, they're using reinforced machine learning. So if you will, when you get an answer, if you say that sucked and then it will try again. So you are actually, by using it more and more, you're improving it. So if you just thought it was stupid in the beginning, it's like your kids, they need some training. They sound intelligent, right? Yeah.

Sting, thanks a lot for your presence in our lives. It's been as always inspiring and we're looking forward to having you as an instructor in our CVO academy.

Oh yeah, you cannot get rid of that now.

No, no, no. What's the term? Bring it on.

Where can people find you? If.

I'M kind of a LinkedIn person so that is probably the best place to follow me. This kind of back in the days I decided that LinkedIn was my weapon of choice and I kind of staying there for now. So this is. Yeah, hook up with me or follow me on LinkedIn and I will try to share whatever I come across as entertaining.

I can vouch for great content and for your weekly article that I always read, I really look forward to your weekly article. I think you put it on Fridays or Saturdays. Yeah, I'm always there, I'm reading it. It's probably one of the few things that I read lately because I don't have time because Valentin had the great idea of promoting me. So I have time for nothing right now. Thank you so much for coming, Steen. This was amazing. I'm excited to see what we're going to do on the Academy and also I'm excited to see what's going to happen in our little, you know, parts of the world here in data analytics. Thank you to everyone for watching on LinkedIn. Everyone that commented on LinkedIn that's still watching, follow Steen. Connect with him. Connect with Julie as well because she's awesome. She's been here supporting in the chat. See you guys next Thursday and Valentin, see you tomorrow.

Thanks for having me. Have a great day, guys.

Have a great one. Bye.

This podcast was brought to you by Vitex. Vitex is the first and only fully integrated E commerce marketplace and OMS solution that offers fastest time to revenue and no upgrades ever. So go ahead, check them out. Vtex.com.