Deb: ⁓ everyone, ⁓ to this episode ⁓ ⁓ Space Between hosted by me Deb and by my esteemed colleague Seth Hawthorne. ⁓ are absolutely delighted for this episode to have Morten Wendel ⁓ The Space Between with us. ⁓ met Morten about 12 months ago when he started looking at
Seth: Hello?
Deb: some of our data and some of the stuff that sits behind MSQ. And we were blown away by some of the ideas he came up with and some of the perspectives he had on our own data. ⁓ He works typically at enterprise level, but I know he's happy to work with even scrappy startups like us. He works at enterprise level, really helping businesses look at the intersection of business strategy, organisational development, data and technology. Hopefully I've done you justice there Morten, but I'd love you to just introduce yourself and tell our listeners from, you know, in your own words, exactly where you're sitting in business structure.
Morten Wendelbo: Yeah, yeah, sure. Well, thank you for having me. I really appreciate it. This is a very interesting endeavor and I'm happy to be a small part of it. ⁓ Well, my background is as an economist, as a behavioral economist and ⁓ have done a lot of work on sort of behavioral economics and political psychology. ⁓ decision making, the psychology behind decision making in organizations. And I am also a data scientist and combining those two sort of skill sets is what I build on when I help organizations make better sort of strategic decisions about ⁓ really everything from IT and data.
Deb: Thank you.
Morten Wendelbo: the high level, not like, you know, specific IT systems or anything like that. But how do we organize ourselves as a business, all the way from that to sort of strategic HR and larger strategic commercial decisions ⁓ that are going to have consequences that sort of reverberate through the whole organization.
Deb: I hope you enjoyed. Brilliant, brilliant. And we're delighted to have you along. And in fact, it was interesting in the briefing session, I think it came out, you we've all been involved in data in some form, good or bad, my mind has been in small and large businesses. But the one thing that you said that absolutely stopped me in my tracks in the briefing session was that, you know, and ⁓ it was just an interesting comment based on the commentary about AI. is that AI is actually a hindrance on data, not a help if a business doesn't have a strategy or an architecture underneath it. I'd love to unpack Seth and I'd love to unpack that with you because you know, what we really heard was, know, what we're hearing is, know, AI sort of might be the answer to everything, but that sort of stopped me in our tracks. So expand, please.
Morten Wendelbo: Yeah, sure. Well, I I think AI can be the answer to a lot of things. Certainly not to everything, but to a lot of things. And particularly when it comes to business, there are a lot of advantages to AI ⁓ that can help us improve productivity, quality, remove a lot of the sort of manual. work kind of work that we often have to do even in highly specialized physicians, how many of us don't spend hours and hours every week doing something that really should be automated and so on, right? So there's great promise. But the reason why I brought it up when we talked is because I think there's a misunderstanding sort of between what AI is sort of on the margins of the business. And so here I'm talking, you know, chat, GVT or Claude or something like that. We are we are using AI as individuals in our work days, some kind of like AI assistant of sorts.
Deb: Yeah.
Morten Wendelbo: quite different from having AI in the core of the business. So putting AI into ⁓ automating certain tasks, having AI do quality assurance. having AI be part of production decisions. And there are different levels of AI here. There's the kind where it's AI in a machine. We have these incredible machines where artificial intelligence in a hundredths of a second can take a large resolution picture of things flying through the machine and with high precision get rid of anything that looks like an abnormality. So there is AI in the core of the business in this sense. But if we want to use sort of AI in a more comprehensive way in the business, really have it be part of the way that work gets done overall, then there are some challenges that we have to overcome and there are very few businesses that have done it. And where they don't, and we can talk about what these are, where they don't, AI is going to do more damage than good.
Deb: And I know that we, I'd like you to talk through, you gave us an example and obviously we want to keep our clients confidences. So if you could tell us that story and maintain your clients confidence, that would be amazing. So it was just interesting around the architecture and the strategy and their approach. So I would love you to touch on that for our listeners, if you don't mind.
Morten Wendelbo: Yeah, so in a very general sense first, and then we can dive into the specific case. In a very general sense, the limitation on using AI and using technology generally, really, because it's not AI specific in some way.
Deb: Yeah.
Morten Wendelbo: The challenge is that in a complex organization, there are lots of systems, lots of data sources, lots of rules and standards, some of which are written and some of which are not. The way that work gets done is influenced about by
Deb: you
Morten Wendelbo: let's call it business architecture. It's the kind of structure and systems and everything that sits not in an individual department, but that sort of governs the business overall. There are some aspects of IT, some aspects of HR, IT systems, all these kinds of things, everything that sort of is the connected tissue in a business. So sales will tend to have control of ⁓ customer system, production of production system, logistics have some kind of warehousing system and so on. All of these systems tend to sort of sit out in their individual departments, but they all have to connect with each other to make a coherent business. And what happens often is that that sort of connecting layer, the business architecture, is very immature in most businesses. So what happens is when we come in and we sort of plug in AI in the core of the business, so not in an individual production setting or in an individual sort of administrative setting or something like that, but we want to bring it into the overall process of the company, then it needs to interface with lots of different systems in lots of different departments. But that architecture in being very immature means oftentimes there aren't standards and rules for how we connect to data. What, you know, sort of the truth of the data is which system owns it. know, we tend to have customer data in like four or five departments and four or five systems. We keep them sort of up to date, but not fully up to date in any of the departments. so as AI dives into our data, the data quality and its access to data is incredibly important. But what happens is because there is no sort of like coherent architecture, is that we then kind of from whatever AI solution we have, we're going to sort of have tentacles into lots of different systems and departments with varying definitions and standards and all this kind of thing. It's doable. But typically it requires a lot more to set up an AI initiative like that than one might think, because instead of integrating into the company, you have to integrate into seven business units. Really troublesome aspect here becomes, even if you have a good commercial case for spending the resources to do that, that pays off. What happens is you actually make the future harder because those six or seven integrations you made into different departments,
Deb: Yeah.
Morten Wendelbo: If you ever want to develop your AI solution, whatever initiative is, then you have to redo those integrations. If you want to add in another AI initiative, you've got to do another six or seven integrations. And so what happens is that as we put AI into the business, it exponentially increases complexity. in a way that's very expensive to maintain. It's very fragile because if one of those connections breaks, then your solution also breaks. So we have to maintain seven systems correctly in order for your AI solution to work. And if we ever want to sort of change up our organization, then we also have to change all of our various solutions to match however that changes. So it does a lot of things. It makes us very vulnerable. It's very expensive. And it kind of what what I call it with a colleague of mine, it sort of pours the organization in place. It becomes sort of like digital concrete. And it becomes incredibly hard to escape. Very expensive, very involved, ⁓ and many businesses get really hurt, and we can talk about a case, get really hurt when they try to do so and they haven't planned for it. ⁓ So one example of this, and this is the one that I brought up, it's a large, ⁓
Deb: Yeah, yeah.
Morten Wendelbo: ⁓ company here in Denmark, I won't mention the sector because then it very quickly becomes evident which company it is. It is a very large company. They've done lots of acquisitions over the last three or four years and by lots I mean like dozens. ⁓ Basically, their growth strategy is through acquisition. So they buy out lots of small competitors ⁓ that are peripheral in their
Deb: Yeah, please don't.
Morten Wendelbo: in their market. And what has happened there is they didn't have an overall architecture. It wasn't designed to include ⁓ new companies at this sort of rate. So they are actually at the for the time being, last I heard they were maintaining 44 separate, what we call ERP systems, which is sort of supposed to be a core business system, most businesses that should have one ERP system, a very complex business businesses might have three or four. But 44 is, I was going to say is not the norm, but actually that's not true. the size, it probably is the norm, but it shouldn't be, right? Incredibly expensive to maintain, and it makes the business very fragile.
Deb: Can you just elaborate on when you were telling us this on the briefing session? later on, I'd like to give a story about how it impacts on marketing. But you did give some real examples of how this impacted on operations, on customer service. Can you just talk through a few of those for us? And I don't think, I think 44 systems is a lot. But I think many businesses have multiple systems that they've just built. rubbish on top of rubbish. ⁓ you know, think it, but just the impact of doing that from an ops perspective, from a customer service perspective, those stories are pretty good.
Morten Wendelbo: Yeah. Yeah, yeah, absolutely. so a couple of interesting things. So 44 systems, it's 44 ERP systems. If we count all of the IT systems that they have, they're somewhere in the several hundreds. ⁓ And in very large complex businesses, that's not so atypical. ⁓ If they're very specialized, then they do many different things. ⁓ One of the largest company mergers of all time was between a Danish logistics giant called DSV. They bought DB Schenker. It's a German or it was a German ⁓ logistics, you know, shipping company. And ⁓ DB Schenker had when DSV was looking at DB Schenker, when they had sort of tallied it, they had 1184 systems that DSV had to somehow integrate into their own business on top of their own sort of complexity. So it very quickly gets very big and very complicated. So in this case, the 44 ERP systems is accompanied by... I think there were at least seven or eight different customer, like CRM systems. So customer relations management systems, right? And so the, the really problematic point here isn't just that it's expensive and difficult to maintain that sort of, ⁓ that's for the financial officer and for the, you know, the, CIO, the, the IT chief to manage, but it has a very clear impact on customers. So what happens is this company is quite large and they have sort of functions, they support the same industry, but a customer may be a customer in two or three or four or five divisions of this company. But because there's no coherence between them, they're going to have as many different customer experiences as they have customer relationships. So the business doesn't know from one division to the other that this is a customer everywhere else. On top of that, So, you know, they can't provide the best customer service. can't sort of provide a loyalty bonus for, you're also, you know, a customer in four other divisions. We're going to give you, you know, the extra good treatment. The problem is also because they have different ERP systems, they get billed four different ways. with four different visual expressions, four different account numbers, four different customer representatives. Like in terms of creating a coherent customer experience, which is incredibly important in that sector, it's impossible. The system, because there is no sort of architectural coherence, the business simply can't do it.
Deb: or actually for the business, a lifetime value of each customer. Sorry, Seth, I've butted in.
Seth: was listening to all of this, I mean, it's absolutely fascinating, Morton. And I'm really intrigued and interested in if you're faced with that situation, you're faced with that spaghetti locked in that concrete, how do you unlock it and unravel it? mean, what do you do?
Deb: Where'd you stop?
Morten Wendelbo: Yeah, so very carefully. And this is where it sort of goes wrong for a lot of businesses. And I'll give you a concrete example that's a bit scary and hopefully drives home the importance of doing this. But generally speaking, it's one thing to know what good looks like. It's another thing to get from where we are now to what good is, right? So there is a lot of... Decisions that have to be made there are there's a lot of careful process But I actually I think the most important thing is that decision makers need to understand that architecture is When it's not managed well because it's there whether we want it or not, right like business architect
Deb: Yeah, it'll be there of some description. It's just the pile of spaghetti, what they described it, what that...
Morten Wendelbo: Yeah, exactly. We call it sort of a spaghetti landscape in fact. ⁓ What has happened is if we haven't managed it, it's sort of been poured bit by bit. Lots of small decisions that weren't made with the overall architecture in mind, know, sales decides on the sales system. ⁓ Maybe they decided on a customer ⁓ service system, but also then they acquired another customer service system that sort of got included in the system landscape, but not fully integrated or combined. So it's all of these small decisions that looked rational and reasonable on their own, but they all accumulate to create this digital concrete. So you have to unpick it very carefully. And the reason why is, so I mentioned, if you have an AI solution that relies on access to seven different systems, if you undo one of them without understanding the consequences for everything else that's tied into it. that's going to sort of have second, third, fourth, fifth order consequences that you're not going to be able to foresee. So if you don't know that this data in this particular format in this system is actually what drives this customer facing AI solution, for example, the day you disconnect this or even just change the way that the data looks, like you just changed how something is defined. that AI solution breaks, right? And so if it's customer facing, that's particularly bad, but there are also examples of, and I'll give the concrete one, there's a sort of a mid-size manufacturing company. they, the board, the senior leadership sort of decide we need a new ERP system. The system that we have is kind of end of life. And in a mid-size company, this kind of system typically has all of the sort of financial things, finance and operations. It had logistics, it had sort of production coordination. In a mid-size company, ERP systems tend to do a lot. And they just sort of said, our business has gotten really complex. We're going to simplify it. We're going to buy a one IT system, everybody's gonna work with it, it's gonna be great. And what happened is that they underestimated drastically how complex their existing architecture, that they hadn't built, but existed nonetheless, how complex it was and the ways in which it impacted the business. So what happened is they paid in excess of 10 million pounds for the software itself. Including sort of implementation of it, know, right consultants come in and help implement it. But what happened was They hadn't understood just how many places it was connected It turns out sort of in retrospect it was connected in in dozens and dozens of other systems many of which they'd overlooked So the business actually stood still for seven weeks Logistics didn't work production was shut down Billy like bills couldn't be sent out. So they weren't getting paid
Deb: Thank
Morten Wendelbo: And so a number of things happened here. One ⁓ very painful one, of course, is that they're not producing and so they're losing their production and their future earnings. But they had to hire extremely expensive consultants to come in and fix it on an emergency basis because every minute it's not working is a minute we're not producing and we're losing money. And so their implementation costs quadrupled. the money for the implementation quadruple overnight and there was nothing to do. It had to get done. The business didn't have it was existential. You must do it. But when they did finally get it to work, what they had also overlooked was the way that these IT systems sort of work together. What really separates that business from a lot of its competitors is the way in which it develops its products and the way in which it sort of interfaces with the customers.
Deb: Yeah.
Morten Wendelbo: and the new IT system did not allow for either of those. So what happened is that we made sort of a large scale strategic decision without understanding the architecture and how it impacted the business. And not only did it cost us a lot of money because we didn't do the planning and execution of it well, but even if that had gone well, what we'd actually done is we eroded what defined the business's competitive advantage. because we didn't understand that an IT system was at the center of how production was done. in the production unit, they literally could not produce the way they used to because the development unit, it's a specialized product. a customer comes in, it goes through development and it's produced for that customer sort of custom. So they lost market, they lost market share. They lost 25 % of their revenue, which was 13 years worth of profit.
Seth: Wow.
Morten Wendelbo: in seven, eight weeks.
Seth: So you need to, before you ⁓ embark on a project like this, hearing that you need to sort of understand three things ⁓ as an organization, you need to understand the cost of doing it, the physical cost of doing it, the loss, potential losses involved in doing it. And then importantly, I think you need to understand the gain. that you're going to get from delivering it? Is it a case that organizations don't really understand those three things before they embark on these projects?
Morten Wendelbo: very, very often, yes, I would say, because there are three aspects to interacting with or dealing with business architecture. One is the architecture itself, right? So if it's just been sort of poured bit by bit and we hadn't managed it, it's generally going to be very difficult. Now, if we actively manage it, we have... know, rules and standards for what systems work with what systems and what different departments are allowed to do. Let me give you an example of, imagine this instead of like thinking about business architecture, let's think about literal architecture. Think about a company building a new headquarters. What happens is we want to build new headquarters. We want sort of a building or a campus. that ⁓ not only enables us to do what we do today, right? The amount of production and the amount of warehousing and everything we've got today, we're imagining this as a manufacturing company for the sake of simplicity. We also need that campus to sort of support who we think we're gonna be in five or 10 or 15 or 20 years, at least to some extent, right?
Deb: you Thank
Morten Wendelbo: Maybe we don't build the buildings we need in 10 years, but we need the buildings to be sort of reconfigurable in such a way that we can pivot if we need to. We can expand production or expand warehousing. Maybe it's just that we make sure that the land that we've bought is big enough that we can just build a new building next to it. But we've thought through the sort of architecture of making sure that the land that we've set aside for an extra production unit isn't, you know, half a mile away on the other side of campus. It's the plot next to the existing production unit, right? Like some, a lot of thought goes into how do we go from what the board and the CEO sort of strategically wants, what their ambitions are, and how do we translate that into an architecture, in this case, like a literal architecture that enables that in the future. Yeah, and so what happens there is, and there is no business leader in the world
Deb: of physicality effectively.
Morten Wendelbo: who would sort of go, okay, we have these ambitions, we have a plot of land, now everybody, you just build whatever you need for your department, right? Because that is not coordination made.
Deb: ⁓
Morten Wendelbo: that's going to be a mess, right? So, so what's effectively happening in most businesses today is that we are letting sort of department heads and so on make decisions for their department that in their department makes perfect sense for their incentives, their level of insight in the business, you know, their goals and strategic goals in front of them. But that sort of accumulates to be this sort of concrete port in place where everybody kind of does their own thing. And that becomes a very incoherent architecture. Now, what we do is we hire an architect to help us translate our ambitions and expectations of the future into a coherent sort of flexible, managed architecture. And then what happens is then we give the production unit, we give them their building and within that building, they do their thing. We've defined where they need to output their things to warehouse, right? So that the warehouse knows where to pick it up. That's defined in the business architecture, but how they get from receiving to warehousing, like through the production unit, the production unit decides that. What's important is that the production unit doesn't decide to do it well away with like some of the shared utilities, like they don't move the electrical main, they don't move the doors, they don't, right? And so, but this happens in businesses, in business architecture all the time, right? One business will remove, sort of a load-bearing wall, because it's in their way. And what's going to happen is that the department upstairs is going to come through the floor next to you, right? Because they don't understand that it's a load-bearing wall. Because our functional leaders aren't, and probably shouldn't be, prepared to understand the sort of business architecture implications. What we really should have
Deb: Yeah.
Morten Wendelbo: is a coherent managed architecture and that takes a literal business architect in a large enough company there should be somebody aware of function with that responsibility.
Deb: Yeah, I think in, I remember doing this a long time ago, like it was in my last corporate role. had 37 offices and five brands. And we ran a project like this, you know, we had, we had a very specific team that did it. And it took us three years to do it. Marketing, you know, was CMO at that point and marketing was very much part of that team. It was fascinating to watch how this team worked because they also had to do a massive, I think probably the first 12 months was education, education, you know, whole education, education of the business, education of the key stakeholders, scope of work. So it's, it's no, it's no easy feat. And I was actually really pleased. I was there at the end when I saw the results that the business had, ⁓ because, but it was hard work. So. but totally worth it. And I love the retaining more story. ⁓ We just looked at it exactly in the way that you described it. So that's absolutely perfect. ⁓
Morten Wendelbo: But it requires leadership that understands it because one of the really important aspects of this is imagine that we think of the execution of the company. So out in the individual departments, whatever they're working on when they're executing, if they are what we would call architecture aware, so they understand what their decisions are doing to the rest of the business.
Deb: Yeah.
Morten Wendelbo: even if they are, they don't have the authority to make decisions that are coherent for the whole, right? That sort of authority has to come down from the board and the CEO. and can't be made, can't be, I mean the IT, the head of IT can make IT decisions, the head of HR can make HR decisions, but making decisions together that work. Turns out, know, in the world that we live in, say remote work for example, an HR policy and the IT policy kind of go hand in hand. They're very important that they sort of work together, otherwise it's not really a possibility to do remote work. But what happens a lot is that senior leadership, imagine we have a company where architecture is perfect, of like it's perfectly managed to the degree that it possibly can. ⁓ We have the right complexity that sort of defines what the business competitive advantage is. we've done away with all of the unnecessary complexity that, you know,
Deb: Thank you.
Morten Wendelbo: Even then, if you have senior management that makes strategy that doesn't sort of depart from this architecture, like understands the limitations that the architecture puts on the business, it doesn't matter how, you know, architecture where everybody out in the executing departments are, they're being handed a strategy that the architectures literally cannot support. And so what happened is that. It's a good thing that they're aware because they can cope better, but all they do is they cope. They try to, within the architecture, the IT systems and the HR rules and everything they're sort of given from above, they're trying to make this strategy work, even though the strategy is very, very often developed in a way that doesn't understand the business's limits. And so we end up spending lots of effort on these coping mechanisms, sort of workarounds. tend to be manual workarounds and all of this. Actually, AI can help us probably with quite a few of
Deb: Thank
Morten Wendelbo: those. But even where they do, and this is a tricky thing about AI is a lot of the places where businesses see an opportunity for AI to make a difference are as coping mechanisms, doing, you know, taking care of automating things that maybe shouldn't need to get done, streamlining them, making them faster. they tend to be efforts that sort of are in response to bad architecture or bad strategy, as opposed to being again, sort of in the core business, developing, setting the business apart. Whereas if we have sort of like very good coherent architecture where we know where all of the data lives and how the systems work together, the consequences when we have an AI initiative, we only need to do one integration into the overall architecture, not into the individual systems. It becomes easier to take them away. If we need to, becomes easier to sort of like plug in a new system and define who owns what. But that can't happen. if either the architecture isn't managed or the senior management is making strategy that isn't aware of the limitations of the architecture.
Seth: So Morten, to sort of bring it back to the world in which we work in, which is working with independent agencies and they tend to be, you know, not enterprise businesses, they tend to be smaller businesses. ⁓ Is there a size of business at which this should be started to to think about? Or, ⁓ I mean, you know, There's a lot of use of AI in small agencies, but probably the tool that gets used by agencies most of all is probably still Excel on a daily basis, ⁓ an Excel spreadsheet. But there's independent agencies ⁓ that merging with other independent agencies. So you start to get these small multi-entity groups. So they obviously have this issue, but is what you're talking about, is it a
Deb: . Thank
Seth: only applicable to large scale organizations or is there a point at which you should start to integrate?
Deb: So thank ⁓
Morten Wendelbo: Yeah, no, at all. It's just it. Yeah, it's just the place where it's easiest to illustrate because the complexity is so clear and so obvious and it sort of like mushroom very quickly. ⁓ So in a large complex organization, you we'd want to have a business architect or a business architecture function. But even in very small companies, mean, like with a dozen employees, two dozen employees, we'd want somebody to have responsibility for the architecture.
Seth: Yeah, sure.
Deb: you
Morten Wendelbo: Right? Because part of the important thing is to just ask the right questions. So even if we don't like have coherent procedures and all this kind of thing, which a small company is not going to have the bandwidth to do, and it's not really necessary. But if we make sure that we ask the right questions sort of throughout the year about how our systems work together, what consequences is it going to have when we get this new system or we integrate with a customer system or customers pull data directly from our systems? Well, then we have a great responsibility for making sure that that data is always available consistently without errors and so on. So do we understand how our own sort of system landscape shapes our ability to keep our promises to our customers? ⁓ So the board, if there is a board, should ask sort of different questions about risks. do with systems and architecture and so on. ⁓ But I think generally it's also, I mean, it's important in a few different ways as an agency that works directly with a customer in so far as there is any AI solutions involved in what you're providing, for example, any other technology solution for that matter. ⁓ It's important to understand what are the limitations of the customer's data that we're going to sort of be integrating into it, right? Because often they have limitations. It's also worth understanding. you know, where's this data coming from? it in systems that are going to get changed and this kind of thing? We should make a sort of a robust solution that can ⁓ handle a little change as opposed to just have something that breaks entirely. I think lots of companies are hesitant to buy any sort of digital service that ties into their systems because they are aware that it's yet another thing that kind of adds to their complexity and and makes it hard to break away from it eventually. So the degree to which agencies in their solutions and integrations with the customers are able to sort of not just change the solution because oftentimes the solutions aren't ⁓ very sort of invasive, but articulate how the solutions aren't actually adding to the complexity and so on. think a lot of customers will respond relatively well to that.
Deb: Bye.
Morten Wendelbo: I think it's also worth understanding how this adds, just how this adds to the customer's complexity because being able to advise them on that complexity is something I think will send a very important signal to most customers about the maturity of the agency and their ability to see the rest of the business and understand the rest of the business because...
Deb: I, yeah, and I think just before we wrap up, couldn't agree more. think this is an opportunity, you know, we're talking a lot to agencies about, you know, how they shift themselves in, you know, in the minds of their clients. ⁓ And I think, you know, to have an understanding of how to put down, I've got example after example of how, you know, we've gone into businesses ⁓ and try to help. that we actually can't deliver a great marketing solution unless we truly understand the data ⁓ and how they structure their data. ⁓ personally, I think this is a massive opportunity, not only for agencies to sort of clean their own house, but actually be fully, show customers how to do it and their clients how to do it. I think that sits at the heart. I think that will help them sit at the heart of their client's business. So just before we close, ⁓ What are the three, you know, when you walk into a business, what are the three questions that you ask a business, a CEO and their leadership team about the structure or the architecture or the, know, what do you ask them to get a really good picture? want, whoever's listening, we want them to be asking them the questions of themselves. What would be three questions?
Morten Wendelbo: Yeah. Do you have clear insight into where all of the necessary data lives? So do you know what system it belongs in, not just which systems have it, but which system has the correct one? Which system owns the truth? It's a very important thing because otherwise we end up using data from systems that have information that looks relevant to us.
Deb: Yeah, great. single source of truth.
Morten Wendelbo: Yes, single source of truth is incredibly important. Most companies don't have it, but you can kind of figure out different data points will have different sort of sources of truth and you can hone in on it. So that's one thing. ⁓
Deb: Question two.
Morten Wendelbo: Do you have clear definitions of the data? Do you understand how the data comes to be? Oftentimes, it's a literal equation. Most of this data is stored as some text or a numerical figure. But where does it come from? Who produces it? Do we trust that source? Because the reality is, even if we think we have a very good system and our system is very clean, most of the data that our systems contain come from other places.
Deb: human.
Morten Wendelbo: Human, yeah. And there are like really simple things that we can do to help humans do better data entry or after the data has been entered to just verify that the data has the shape that we're expecting. Simple sort of. ⁓ data validation things that I would say probably with a regular Excel data set, half an hour of work and we can improve the quality of our data several hundred percent ⁓ because we can trust it more because we've just kind of like made sure it's shaped correctly. ⁓
Deb: Question three.
Morten Wendelbo: And then probably in this particular case, who is going to own it moving forward? Because one of the things we want to understand is whether they have a clear plan for what's about to happen. Who's going to own it? Who has the authority? And so on, right? Like we need to understand who has the authority to give us this data and maintain it and so on, ⁓ Because without that, whatever solution we have is sort of liable. liable to kind of die if nobody on the customer side owns it.
Deb: Yeah, thank you. Morton, just as we're closing, you know, you are a prolific publisher, writer, producer of white papers, author of books. Where can people find you if they want to reach out to you? I think this to me is possibly one of the most important topics that needs to sit in a business with a leadership team. So if they want some help from you, where can they connect and find you?
Morten Wendelbo: Yeah, the easiest thing is probably on LinkedIn. So just look me up and you're the only one in the world with my name. it should be relatively easy. I'm very happy to.
Deb: You are? I can confirm that.
Morten Wendelbo: to have a, even if it's just a brief chat, ⁓ I think my years of experience and having seen this across more than 100 different companies of all sizes in all sectors at this point, I can probably for most in half an hour give them more and set them in the right direction ⁓ in half an hour. And what's that?
Deb: Yeah. Untangle their spaghetti and concrete. Untangle their spaghetti and concrete.
Morten Wendelbo: Yes, yes. And within, hopefully within about six months, ⁓ I'm writing a book with a colleague that's called Port in Place. That's about organizational concrete. ⁓ so I hope people will keep an eye out for that. That's going to have sort of... That sounds great. Yeah. And that explains, you know, why, how we got in this situation, what good looks like, and then how do we actually get there.
Deb: Thank maybe we'll have you back on in six months and you can talk us through the book that would be awesome. Oh, cannot wait, cannot wait. Thank you for your time this morning. As always, it's a delight to talk to you and we cannot wait to, you know, start working with you ourselves. Brilliant. Thanks a lot. Thanks, Morten.
Morten Wendelbo: Yeah, thank you for having me.
Seth: Thank you, Morton.
Seth: Hello?
Deb: some of our data and some of the stuff that sits behind MSQ. And we were blown away by some of the ideas he came up with and some of the perspectives he had on our own data. ⁓ He works typically at enterprise level, but I know he's happy to work with even scrappy startups like us. He works at enterprise level, really helping businesses look at the intersection of business strategy, organisational development, data and technology. Hopefully I've done you justice there Morten, but I'd love you to just introduce yourself and tell our listeners from, you know, in your own words, exactly where you're sitting in business structure.
Morten Wendelbo: Yeah, yeah, sure. Well, thank you for having me. I really appreciate it. This is a very interesting endeavor and I'm happy to be a small part of it. ⁓ Well, my background is as an economist, as a behavioral economist and ⁓ have done a lot of work on sort of behavioral economics and political psychology. ⁓ decision making, the psychology behind decision making in organizations. And I am also a data scientist and combining those two sort of skill sets is what I build on when I help organizations make better sort of strategic decisions about ⁓ really everything from IT and data.
Deb: Thank you.
Morten Wendelbo: the high level, not like, you know, specific IT systems or anything like that. But how do we organize ourselves as a business, all the way from that to sort of strategic HR and larger strategic commercial decisions ⁓ that are going to have consequences that sort of reverberate through the whole organization.
Deb: I hope you enjoyed. Brilliant, brilliant. And we're delighted to have you along. And in fact, it was interesting in the briefing session, I think it came out, you we've all been involved in data in some form, good or bad, my mind has been in small and large businesses. But the one thing that you said that absolutely stopped me in my tracks in the briefing session was that, you know, and ⁓ it was just an interesting comment based on the commentary about AI. is that AI is actually a hindrance on data, not a help if a business doesn't have a strategy or an architecture underneath it. I'd love to unpack Seth and I'd love to unpack that with you because you know, what we really heard was, know, what we're hearing is, know, AI sort of might be the answer to everything, but that sort of stopped me in our tracks. So expand, please.
Morten Wendelbo: Yeah, sure. Well, I I think AI can be the answer to a lot of things. Certainly not to everything, but to a lot of things. And particularly when it comes to business, there are a lot of advantages to AI ⁓ that can help us improve productivity, quality, remove a lot of the sort of manual. work kind of work that we often have to do even in highly specialized physicians, how many of us don't spend hours and hours every week doing something that really should be automated and so on, right? So there's great promise. But the reason why I brought it up when we talked is because I think there's a misunderstanding sort of between what AI is sort of on the margins of the business. And so here I'm talking, you know, chat, GVT or Claude or something like that. We are we are using AI as individuals in our work days, some kind of like AI assistant of sorts.
Deb: Yeah.
Morten Wendelbo: quite different from having AI in the core of the business. So putting AI into ⁓ automating certain tasks, having AI do quality assurance. having AI be part of production decisions. And there are different levels of AI here. There's the kind where it's AI in a machine. We have these incredible machines where artificial intelligence in a hundredths of a second can take a large resolution picture of things flying through the machine and with high precision get rid of anything that looks like an abnormality. So there is AI in the core of the business in this sense. But if we want to use sort of AI in a more comprehensive way in the business, really have it be part of the way that work gets done overall, then there are some challenges that we have to overcome and there are very few businesses that have done it. And where they don't, and we can talk about what these are, where they don't, AI is going to do more damage than good.
Deb: And I know that we, I'd like you to talk through, you gave us an example and obviously we want to keep our clients confidences. So if you could tell us that story and maintain your clients confidence, that would be amazing. So it was just interesting around the architecture and the strategy and their approach. So I would love you to touch on that for our listeners, if you don't mind.
Morten Wendelbo: Yeah, so in a very general sense first, and then we can dive into the specific case. In a very general sense, the limitation on using AI and using technology generally, really, because it's not AI specific in some way.
Deb: Yeah.
Morten Wendelbo: The challenge is that in a complex organization, there are lots of systems, lots of data sources, lots of rules and standards, some of which are written and some of which are not. The way that work gets done is influenced about by
Deb: you
Morten Wendelbo: let's call it business architecture. It's the kind of structure and systems and everything that sits not in an individual department, but that sort of governs the business overall. There are some aspects of IT, some aspects of HR, IT systems, all these kinds of things, everything that sort of is the connected tissue in a business. So sales will tend to have control of ⁓ customer system, production of production system, logistics have some kind of warehousing system and so on. All of these systems tend to sort of sit out in their individual departments, but they all have to connect with each other to make a coherent business. And what happens often is that that sort of connecting layer, the business architecture, is very immature in most businesses. So what happens is when we come in and we sort of plug in AI in the core of the business, so not in an individual production setting or in an individual sort of administrative setting or something like that, but we want to bring it into the overall process of the company, then it needs to interface with lots of different systems in lots of different departments. But that architecture in being very immature means oftentimes there aren't standards and rules for how we connect to data. What, you know, sort of the truth of the data is which system owns it. know, we tend to have customer data in like four or five departments and four or five systems. We keep them sort of up to date, but not fully up to date in any of the departments. so as AI dives into our data, the data quality and its access to data is incredibly important. But what happens is because there is no sort of like coherent architecture, is that we then kind of from whatever AI solution we have, we're going to sort of have tentacles into lots of different systems and departments with varying definitions and standards and all this kind of thing. It's doable. But typically it requires a lot more to set up an AI initiative like that than one might think, because instead of integrating into the company, you have to integrate into seven business units. Really troublesome aspect here becomes, even if you have a good commercial case for spending the resources to do that, that pays off. What happens is you actually make the future harder because those six or seven integrations you made into different departments,
Deb: Yeah.
Morten Wendelbo: If you ever want to develop your AI solution, whatever initiative is, then you have to redo those integrations. If you want to add in another AI initiative, you've got to do another six or seven integrations. And so what happens is that as we put AI into the business, it exponentially increases complexity. in a way that's very expensive to maintain. It's very fragile because if one of those connections breaks, then your solution also breaks. So we have to maintain seven systems correctly in order for your AI solution to work. And if we ever want to sort of change up our organization, then we also have to change all of our various solutions to match however that changes. So it does a lot of things. It makes us very vulnerable. It's very expensive. And it kind of what what I call it with a colleague of mine, it sort of pours the organization in place. It becomes sort of like digital concrete. And it becomes incredibly hard to escape. Very expensive, very involved, ⁓ and many businesses get really hurt, and we can talk about a case, get really hurt when they try to do so and they haven't planned for it. ⁓ So one example of this, and this is the one that I brought up, it's a large, ⁓
Deb: Yeah, yeah.
Morten Wendelbo: ⁓ company here in Denmark, I won't mention the sector because then it very quickly becomes evident which company it is. It is a very large company. They've done lots of acquisitions over the last three or four years and by lots I mean like dozens. ⁓ Basically, their growth strategy is through acquisition. So they buy out lots of small competitors ⁓ that are peripheral in their
Deb: Yeah, please don't.
Morten Wendelbo: in their market. And what has happened there is they didn't have an overall architecture. It wasn't designed to include ⁓ new companies at this sort of rate. So they are actually at the for the time being, last I heard they were maintaining 44 separate, what we call ERP systems, which is sort of supposed to be a core business system, most businesses that should have one ERP system, a very complex business businesses might have three or four. But 44 is, I was going to say is not the norm, but actually that's not true. the size, it probably is the norm, but it shouldn't be, right? Incredibly expensive to maintain, and it makes the business very fragile.
Deb: Can you just elaborate on when you were telling us this on the briefing session? later on, I'd like to give a story about how it impacts on marketing. But you did give some real examples of how this impacted on operations, on customer service. Can you just talk through a few of those for us? And I don't think, I think 44 systems is a lot. But I think many businesses have multiple systems that they've just built. rubbish on top of rubbish. ⁓ you know, think it, but just the impact of doing that from an ops perspective, from a customer service perspective, those stories are pretty good.
Morten Wendelbo: Yeah. Yeah, yeah, absolutely. so a couple of interesting things. So 44 systems, it's 44 ERP systems. If we count all of the IT systems that they have, they're somewhere in the several hundreds. ⁓ And in very large complex businesses, that's not so atypical. ⁓ If they're very specialized, then they do many different things. ⁓ One of the largest company mergers of all time was between a Danish logistics giant called DSV. They bought DB Schenker. It's a German or it was a German ⁓ logistics, you know, shipping company. And ⁓ DB Schenker had when DSV was looking at DB Schenker, when they had sort of tallied it, they had 1184 systems that DSV had to somehow integrate into their own business on top of their own sort of complexity. So it very quickly gets very big and very complicated. So in this case, the 44 ERP systems is accompanied by... I think there were at least seven or eight different customer, like CRM systems. So customer relations management systems, right? And so the, the really problematic point here isn't just that it's expensive and difficult to maintain that sort of, ⁓ that's for the financial officer and for the, you know, the, CIO, the, the IT chief to manage, but it has a very clear impact on customers. So what happens is this company is quite large and they have sort of functions, they support the same industry, but a customer may be a customer in two or three or four or five divisions of this company. But because there's no coherence between them, they're going to have as many different customer experiences as they have customer relationships. So the business doesn't know from one division to the other that this is a customer everywhere else. On top of that, So, you know, they can't provide the best customer service. can't sort of provide a loyalty bonus for, you're also, you know, a customer in four other divisions. We're going to give you, you know, the extra good treatment. The problem is also because they have different ERP systems, they get billed four different ways. with four different visual expressions, four different account numbers, four different customer representatives. Like in terms of creating a coherent customer experience, which is incredibly important in that sector, it's impossible. The system, because there is no sort of architectural coherence, the business simply can't do it.
Deb: or actually for the business, a lifetime value of each customer. Sorry, Seth, I've butted in.
Seth: was listening to all of this, I mean, it's absolutely fascinating, Morton. And I'm really intrigued and interested in if you're faced with that situation, you're faced with that spaghetti locked in that concrete, how do you unlock it and unravel it? mean, what do you do?
Deb: Where'd you stop?
Morten Wendelbo: Yeah, so very carefully. And this is where it sort of goes wrong for a lot of businesses. And I'll give you a concrete example that's a bit scary and hopefully drives home the importance of doing this. But generally speaking, it's one thing to know what good looks like. It's another thing to get from where we are now to what good is, right? So there is a lot of... Decisions that have to be made there are there's a lot of careful process But I actually I think the most important thing is that decision makers need to understand that architecture is When it's not managed well because it's there whether we want it or not, right like business architect
Deb: Yeah, it'll be there of some description. It's just the pile of spaghetti, what they described it, what that...
Morten Wendelbo: Yeah, exactly. We call it sort of a spaghetti landscape in fact. ⁓ What has happened is if we haven't managed it, it's sort of been poured bit by bit. Lots of small decisions that weren't made with the overall architecture in mind, know, sales decides on the sales system. ⁓ Maybe they decided on a customer ⁓ service system, but also then they acquired another customer service system that sort of got included in the system landscape, but not fully integrated or combined. So it's all of these small decisions that looked rational and reasonable on their own, but they all accumulate to create this digital concrete. So you have to unpick it very carefully. And the reason why is, so I mentioned, if you have an AI solution that relies on access to seven different systems, if you undo one of them without understanding the consequences for everything else that's tied into it. that's going to sort of have second, third, fourth, fifth order consequences that you're not going to be able to foresee. So if you don't know that this data in this particular format in this system is actually what drives this customer facing AI solution, for example, the day you disconnect this or even just change the way that the data looks, like you just changed how something is defined. that AI solution breaks, right? And so if it's customer facing, that's particularly bad, but there are also examples of, and I'll give the concrete one, there's a sort of a mid-size manufacturing company. they, the board, the senior leadership sort of decide we need a new ERP system. The system that we have is kind of end of life. And in a mid-size company, this kind of system typically has all of the sort of financial things, finance and operations. It had logistics, it had sort of production coordination. In a mid-size company, ERP systems tend to do a lot. And they just sort of said, our business has gotten really complex. We're going to simplify it. We're going to buy a one IT system, everybody's gonna work with it, it's gonna be great. And what happened is that they underestimated drastically how complex their existing architecture, that they hadn't built, but existed nonetheless, how complex it was and the ways in which it impacted the business. So what happened is they paid in excess of 10 million pounds for the software itself. Including sort of implementation of it, know, right consultants come in and help implement it. But what happened was They hadn't understood just how many places it was connected It turns out sort of in retrospect it was connected in in dozens and dozens of other systems many of which they'd overlooked So the business actually stood still for seven weeks Logistics didn't work production was shut down Billy like bills couldn't be sent out. So they weren't getting paid
Deb: Thank
Morten Wendelbo: And so a number of things happened here. One ⁓ very painful one, of course, is that they're not producing and so they're losing their production and their future earnings. But they had to hire extremely expensive consultants to come in and fix it on an emergency basis because every minute it's not working is a minute we're not producing and we're losing money. And so their implementation costs quadrupled. the money for the implementation quadruple overnight and there was nothing to do. It had to get done. The business didn't have it was existential. You must do it. But when they did finally get it to work, what they had also overlooked was the way that these IT systems sort of work together. What really separates that business from a lot of its competitors is the way in which it develops its products and the way in which it sort of interfaces with the customers.
Deb: Yeah.
Morten Wendelbo: and the new IT system did not allow for either of those. So what happened is that we made sort of a large scale strategic decision without understanding the architecture and how it impacted the business. And not only did it cost us a lot of money because we didn't do the planning and execution of it well, but even if that had gone well, what we'd actually done is we eroded what defined the business's competitive advantage. because we didn't understand that an IT system was at the center of how production was done. in the production unit, they literally could not produce the way they used to because the development unit, it's a specialized product. a customer comes in, it goes through development and it's produced for that customer sort of custom. So they lost market, they lost market share. They lost 25 % of their revenue, which was 13 years worth of profit.
Seth: Wow.
Morten Wendelbo: in seven, eight weeks.
Seth: So you need to, before you ⁓ embark on a project like this, hearing that you need to sort of understand three things ⁓ as an organization, you need to understand the cost of doing it, the physical cost of doing it, the loss, potential losses involved in doing it. And then importantly, I think you need to understand the gain. that you're going to get from delivering it? Is it a case that organizations don't really understand those three things before they embark on these projects?
Morten Wendelbo: very, very often, yes, I would say, because there are three aspects to interacting with or dealing with business architecture. One is the architecture itself, right? So if it's just been sort of poured bit by bit and we hadn't managed it, it's generally going to be very difficult. Now, if we actively manage it, we have... know, rules and standards for what systems work with what systems and what different departments are allowed to do. Let me give you an example of, imagine this instead of like thinking about business architecture, let's think about literal architecture. Think about a company building a new headquarters. What happens is we want to build new headquarters. We want sort of a building or a campus. that ⁓ not only enables us to do what we do today, right? The amount of production and the amount of warehousing and everything we've got today, we're imagining this as a manufacturing company for the sake of simplicity. We also need that campus to sort of support who we think we're gonna be in five or 10 or 15 or 20 years, at least to some extent, right?
Deb: you Thank
Morten Wendelbo: Maybe we don't build the buildings we need in 10 years, but we need the buildings to be sort of reconfigurable in such a way that we can pivot if we need to. We can expand production or expand warehousing. Maybe it's just that we make sure that the land that we've bought is big enough that we can just build a new building next to it. But we've thought through the sort of architecture of making sure that the land that we've set aside for an extra production unit isn't, you know, half a mile away on the other side of campus. It's the plot next to the existing production unit, right? Like some, a lot of thought goes into how do we go from what the board and the CEO sort of strategically wants, what their ambitions are, and how do we translate that into an architecture, in this case, like a literal architecture that enables that in the future. Yeah, and so what happens there is, and there is no business leader in the world
Deb: of physicality effectively.
Morten Wendelbo: who would sort of go, okay, we have these ambitions, we have a plot of land, now everybody, you just build whatever you need for your department, right? Because that is not coordination made.
Deb: ⁓
Morten Wendelbo: that's going to be a mess, right? So, so what's effectively happening in most businesses today is that we are letting sort of department heads and so on make decisions for their department that in their department makes perfect sense for their incentives, their level of insight in the business, you know, their goals and strategic goals in front of them. But that sort of accumulates to be this sort of concrete port in place where everybody kind of does their own thing. And that becomes a very incoherent architecture. Now, what we do is we hire an architect to help us translate our ambitions and expectations of the future into a coherent sort of flexible, managed architecture. And then what happens is then we give the production unit, we give them their building and within that building, they do their thing. We've defined where they need to output their things to warehouse, right? So that the warehouse knows where to pick it up. That's defined in the business architecture, but how they get from receiving to warehousing, like through the production unit, the production unit decides that. What's important is that the production unit doesn't decide to do it well away with like some of the shared utilities, like they don't move the electrical main, they don't move the doors, they don't, right? And so, but this happens in businesses, in business architecture all the time, right? One business will remove, sort of a load-bearing wall, because it's in their way. And what's going to happen is that the department upstairs is going to come through the floor next to you, right? Because they don't understand that it's a load-bearing wall. Because our functional leaders aren't, and probably shouldn't be, prepared to understand the sort of business architecture implications. What we really should have
Deb: Yeah.
Morten Wendelbo: is a coherent managed architecture and that takes a literal business architect in a large enough company there should be somebody aware of function with that responsibility.
Deb: Yeah, I think in, I remember doing this a long time ago, like it was in my last corporate role. had 37 offices and five brands. And we ran a project like this, you know, we had, we had a very specific team that did it. And it took us three years to do it. Marketing, you know, was CMO at that point and marketing was very much part of that team. It was fascinating to watch how this team worked because they also had to do a massive, I think probably the first 12 months was education, education, you know, whole education, education of the business, education of the key stakeholders, scope of work. So it's, it's no, it's no easy feat. And I was actually really pleased. I was there at the end when I saw the results that the business had, ⁓ because, but it was hard work. So. but totally worth it. And I love the retaining more story. ⁓ We just looked at it exactly in the way that you described it. So that's absolutely perfect. ⁓
Morten Wendelbo: But it requires leadership that understands it because one of the really important aspects of this is imagine that we think of the execution of the company. So out in the individual departments, whatever they're working on when they're executing, if they are what we would call architecture aware, so they understand what their decisions are doing to the rest of the business.
Deb: Yeah.
Morten Wendelbo: even if they are, they don't have the authority to make decisions that are coherent for the whole, right? That sort of authority has to come down from the board and the CEO. and can't be made, can't be, I mean the IT, the head of IT can make IT decisions, the head of HR can make HR decisions, but making decisions together that work. Turns out, know, in the world that we live in, say remote work for example, an HR policy and the IT policy kind of go hand in hand. They're very important that they sort of work together, otherwise it's not really a possibility to do remote work. But what happens a lot is that senior leadership, imagine we have a company where architecture is perfect, of like it's perfectly managed to the degree that it possibly can. ⁓ We have the right complexity that sort of defines what the business competitive advantage is. we've done away with all of the unnecessary complexity that, you know,
Deb: Thank you.
Morten Wendelbo: Even then, if you have senior management that makes strategy that doesn't sort of depart from this architecture, like understands the limitations that the architecture puts on the business, it doesn't matter how, you know, architecture where everybody out in the executing departments are, they're being handed a strategy that the architectures literally cannot support. And so what happened is that. It's a good thing that they're aware because they can cope better, but all they do is they cope. They try to, within the architecture, the IT systems and the HR rules and everything they're sort of given from above, they're trying to make this strategy work, even though the strategy is very, very often developed in a way that doesn't understand the business's limits. And so we end up spending lots of effort on these coping mechanisms, sort of workarounds. tend to be manual workarounds and all of this. Actually, AI can help us probably with quite a few of
Deb: Thank
Morten Wendelbo: those. But even where they do, and this is a tricky thing about AI is a lot of the places where businesses see an opportunity for AI to make a difference are as coping mechanisms, doing, you know, taking care of automating things that maybe shouldn't need to get done, streamlining them, making them faster. they tend to be efforts that sort of are in response to bad architecture or bad strategy, as opposed to being again, sort of in the core business, developing, setting the business apart. Whereas if we have sort of like very good coherent architecture where we know where all of the data lives and how the systems work together, the consequences when we have an AI initiative, we only need to do one integration into the overall architecture, not into the individual systems. It becomes easier to take them away. If we need to, becomes easier to sort of like plug in a new system and define who owns what. But that can't happen. if either the architecture isn't managed or the senior management is making strategy that isn't aware of the limitations of the architecture.
Seth: So Morten, to sort of bring it back to the world in which we work in, which is working with independent agencies and they tend to be, you know, not enterprise businesses, they tend to be smaller businesses. ⁓ Is there a size of business at which this should be started to to think about? Or, ⁓ I mean, you know, There's a lot of use of AI in small agencies, but probably the tool that gets used by agencies most of all is probably still Excel on a daily basis, ⁓ an Excel spreadsheet. But there's independent agencies ⁓ that merging with other independent agencies. So you start to get these small multi-entity groups. So they obviously have this issue, but is what you're talking about, is it a
Deb: . Thank
Seth: only applicable to large scale organizations or is there a point at which you should start to integrate?
Deb: So thank ⁓
Morten Wendelbo: Yeah, no, at all. It's just it. Yeah, it's just the place where it's easiest to illustrate because the complexity is so clear and so obvious and it sort of like mushroom very quickly. ⁓ So in a large complex organization, you we'd want to have a business architect or a business architecture function. But even in very small companies, mean, like with a dozen employees, two dozen employees, we'd want somebody to have responsibility for the architecture.
Seth: Yeah, sure.
Deb: you
Morten Wendelbo: Right? Because part of the important thing is to just ask the right questions. So even if we don't like have coherent procedures and all this kind of thing, which a small company is not going to have the bandwidth to do, and it's not really necessary. But if we make sure that we ask the right questions sort of throughout the year about how our systems work together, what consequences is it going to have when we get this new system or we integrate with a customer system or customers pull data directly from our systems? Well, then we have a great responsibility for making sure that that data is always available consistently without errors and so on. So do we understand how our own sort of system landscape shapes our ability to keep our promises to our customers? ⁓ So the board, if there is a board, should ask sort of different questions about risks. do with systems and architecture and so on. ⁓ But I think generally it's also, I mean, it's important in a few different ways as an agency that works directly with a customer in so far as there is any AI solutions involved in what you're providing, for example, any other technology solution for that matter. ⁓ It's important to understand what are the limitations of the customer's data that we're going to sort of be integrating into it, right? Because often they have limitations. It's also worth understanding. you know, where's this data coming from? it in systems that are going to get changed and this kind of thing? We should make a sort of a robust solution that can ⁓ handle a little change as opposed to just have something that breaks entirely. I think lots of companies are hesitant to buy any sort of digital service that ties into their systems because they are aware that it's yet another thing that kind of adds to their complexity and and makes it hard to break away from it eventually. So the degree to which agencies in their solutions and integrations with the customers are able to sort of not just change the solution because oftentimes the solutions aren't ⁓ very sort of invasive, but articulate how the solutions aren't actually adding to the complexity and so on. think a lot of customers will respond relatively well to that.
Deb: Bye.
Morten Wendelbo: I think it's also worth understanding how this adds, just how this adds to the customer's complexity because being able to advise them on that complexity is something I think will send a very important signal to most customers about the maturity of the agency and their ability to see the rest of the business and understand the rest of the business because...
Deb: I, yeah, and I think just before we wrap up, couldn't agree more. think this is an opportunity, you know, we're talking a lot to agencies about, you know, how they shift themselves in, you know, in the minds of their clients. ⁓ And I think, you know, to have an understanding of how to put down, I've got example after example of how, you know, we've gone into businesses ⁓ and try to help. that we actually can't deliver a great marketing solution unless we truly understand the data ⁓ and how they structure their data. ⁓ personally, I think this is a massive opportunity, not only for agencies to sort of clean their own house, but actually be fully, show customers how to do it and their clients how to do it. I think that sits at the heart. I think that will help them sit at the heart of their client's business. So just before we close, ⁓ What are the three, you know, when you walk into a business, what are the three questions that you ask a business, a CEO and their leadership team about the structure or the architecture or the, know, what do you ask them to get a really good picture? want, whoever's listening, we want them to be asking them the questions of themselves. What would be three questions?
Morten Wendelbo: Yeah. Do you have clear insight into where all of the necessary data lives? So do you know what system it belongs in, not just which systems have it, but which system has the correct one? Which system owns the truth? It's a very important thing because otherwise we end up using data from systems that have information that looks relevant to us.
Deb: Yeah, great. single source of truth.
Morten Wendelbo: Yes, single source of truth is incredibly important. Most companies don't have it, but you can kind of figure out different data points will have different sort of sources of truth and you can hone in on it. So that's one thing. ⁓
Deb: Question two.
Morten Wendelbo: Do you have clear definitions of the data? Do you understand how the data comes to be? Oftentimes, it's a literal equation. Most of this data is stored as some text or a numerical figure. But where does it come from? Who produces it? Do we trust that source? Because the reality is, even if we think we have a very good system and our system is very clean, most of the data that our systems contain come from other places.
Deb: human.
Morten Wendelbo: Human, yeah. And there are like really simple things that we can do to help humans do better data entry or after the data has been entered to just verify that the data has the shape that we're expecting. Simple sort of. ⁓ data validation things that I would say probably with a regular Excel data set, half an hour of work and we can improve the quality of our data several hundred percent ⁓ because we can trust it more because we've just kind of like made sure it's shaped correctly. ⁓
Deb: Question three.
Morten Wendelbo: And then probably in this particular case, who is going to own it moving forward? Because one of the things we want to understand is whether they have a clear plan for what's about to happen. Who's going to own it? Who has the authority? And so on, right? Like we need to understand who has the authority to give us this data and maintain it and so on, ⁓ Because without that, whatever solution we have is sort of liable. liable to kind of die if nobody on the customer side owns it.
Deb: Yeah, thank you. Morton, just as we're closing, you know, you are a prolific publisher, writer, producer of white papers, author of books. Where can people find you if they want to reach out to you? I think this to me is possibly one of the most important topics that needs to sit in a business with a leadership team. So if they want some help from you, where can they connect and find you?
Morten Wendelbo: Yeah, the easiest thing is probably on LinkedIn. So just look me up and you're the only one in the world with my name. it should be relatively easy. I'm very happy to.
Deb: You are? I can confirm that.
Morten Wendelbo: to have a, even if it's just a brief chat, ⁓ I think my years of experience and having seen this across more than 100 different companies of all sizes in all sectors at this point, I can probably for most in half an hour give them more and set them in the right direction ⁓ in half an hour. And what's that?
Deb: Yeah. Untangle their spaghetti and concrete. Untangle their spaghetti and concrete.
Morten Wendelbo: Yes, yes. And within, hopefully within about six months, ⁓ I'm writing a book with a colleague that's called Port in Place. That's about organizational concrete. ⁓ so I hope people will keep an eye out for that. That's going to have sort of... That sounds great. Yeah. And that explains, you know, why, how we got in this situation, what good looks like, and then how do we actually get there.
Deb: Thank maybe we'll have you back on in six months and you can talk us through the book that would be awesome. Oh, cannot wait, cannot wait. Thank you for your time this morning. As always, it's a delight to talk to you and we cannot wait to, you know, start working with you ourselves. Brilliant. Thanks a lot. Thanks, Morten.
Morten Wendelbo: Yeah, thank you for having me.
Seth: Thank you, Morton.
