Simon Wardley - Situation Normal, Everything Must Change (EN)

[00:00:15] Now the issue of strategy, so where I started with mapping, uh then I'm going to talk about maps themselves, then I'm going to talk about being trapped by context, particularly stories. Then I'm going to talk about patterns, how to go go about mapping. And then a particular subject which keeps on coming up at the moment, uh which is the issue of sovereignty and digital sovereignty. Uh, I mean, there's many places we could go, organizational structure, um, we could go into depths on culture, but I, I'm going to pick on this area. Because I have to use a lots of different areas of mapping and a bit of nation state competition to explain it. Now, before I can start, I see some of you have already put thumbs up. Um, can you put a thumb up if you you know my form of mapping?
[00:01:10] Or raise a hand if you know, okay, a few.
[00:01:18] The majority of you, this stuff will be fairly, fairly new to. Okay, that's good, that's good to know. Um, so let's just started, therefore with the basics of where this all came from.
[00:01:33] So the issue of strategy. So many, many years ago, uh this was about, oh gosh, um 2002 roughly, 2003, I was working for this company. Now this company, uh, was an online photo service, I had about 16 different lines of business, uh, one of which was online photos. It was very profitable, uh, revenue was growing, uh, it was doing fine, but it had a major, major problem. And that problem was this person, uh, the CEO. The CEO of the company didn't have a clue what they were doing. They were making it up as they went along. And I know this because I was the CEO. So, um, I used to create wonderful statements, uh, vision strategy, all this sort of stuff. Uh things like this, uh for Tango 2003. Our strategy is customer focus, we will lead an innovative effort in the market for our use of Agile techniques and open source. Uh we were um heavy supporters of the Pearl language, I hired most of the Pearl pumpkins, so we actually um uh pretty much developed the language as well or helped develop it. Um, and we used to use a lot of open source and provide a lot of open source, we adopted extreme programming. I, I, I, I know Kent Beck, I am a big fan of XD. Uh so we'd adopted it throughout the organization by about 2002, 2003. Um, but the problem with the statement was that I'd literally pinched it from another company and changed a few words.
[00:03:15] So I used to go around recording other CEOs talking about strategy. I, I'd take a, recorded it, zoom in, you know, write down the words that they used, I'd look for what I call business level abstractions of a heroic strategy or blaz for short. And I've done this every few years, uh so 2014, I think this is from. Here are the common blas, uh digital business, big data, disruptive.
[00:03:43] Innovative, collaborative, competitive advantage, ecosystem, open source, blah, blah, blah. I mean if you did it today, um, you would definitely have things like AI, uh, blockchain, probably get a bit of digital sovereignty coming in there as well, uh industry for, that'll be a, that'll be a cool word they throw in there. But anyway, I grabbed um a whole bunch of company different documents and and created what I called the blah template, it was like my strategy generator. So, uh, our strategy is blah, we will lead a blah effort in the market through our use of blah and blah to build a blah. And then I would take the blah and the blah template and smash it together. And from this, I would generate 64 random strategies. Things like this. Our strategy is innovative digital business, uh we will lead a growth effort in the market through our use of customer focus, competitive advantage, destructive social media to build a collaborative revolution. It's just pure gibberish.
[00:04:52] But I, I would send them around. And I, again, done this many times, and last time, and I haven't actually done this for about four or five years, so I, I maybe I'll get around to doing it again. But last time I got 400 responses back of three basic types. Uh the first type was, uh this is the exact wording from our business plan. And the second type was, I've seen two of these used already. And the third, and this was my favorite, was, are you for hire? So I, I not only knew I didn't know what I was talking about. I, I started got a suspicion that I wasn't the only person, I mean, um as I did this test more and more, that suspicion just grew. So I didn't know what I was doing. Uh I mean, I, I bought every strategy book I could find, it wasn't helping me. Um I, I, I sort of thought, well it must be they teach something secret to to MBAs which help you understand this stuff and I couldn't find anything. And then I was in a a bookstore. The seller said, have you ever read Sun Su The Art of War? I was like, actually, no I haven't. And so she persuaded me to buy two copies, two different versions because they're both, uh, you know, they're translations. And the argument other than increasing book sales, uh, was that, you know, maybe I would see a difference. I thought, oh, sounds sensible.
[00:06:25] And I was so, so grateful for that because it was actually in the reading of the second version that I actually saw something. So here I am in this world of, uh, you know, blah template strategy. And and by the way, I, I forgot to mention this, a friend of mine's put this all online. Uh so if you ever need one of these random strategies, just type in the URL, it will, uh, this will automatically create you one based upon nothing whatsoever, I mean, you can. pretend it's got AI blockchain, whatever, it doesn't matter. I mean, it's just gibberish words, our strategy collaborative, we will lead an open effort in the market through use of big data, blah. If you don't like it, it's really simple, just press refresh and it will automatically generate you a new one. Anyway. So back to the sun zoo.
[00:07:18] Um, what I noticed, uh, were these five factors that Sun Su talked about. Um, understand your landscape, the environment you're competing. Understand the climactic patterns and how the landscape is changing. Understand doctrine. So that's your principles of organization. Uh understand leadership, which is all about how you play your game.
[00:07:43] And of course, you've got your purpose, your moral imperative, what you're trying to do. And I, I was quite excited by this. And then I came across John Boyd, John Boyd is a US Air Force pilot. He was a US Air Force pilot. And he came up with uh something known as uh the U-loop. So you've got your game. The first is to observe the environment, that's what landscape and climatic patterns are about. Understanding the landscape and how it's changing and what's happening. Then you need to orientate yourself around that space. And there are a lot of factors involved in that, but also principles and doctrine are a key part of that.
[00:08:25] And then you need to decide where you're going to attack and and you act. And a lot of that is where leadership comes into play, including my own context specific forms of playing, you know, flanking an opponent or whatever it happens to be.
[00:08:40] And of course, it's, it's a cycle. And at the heart of this are two wise. The why of purpose, your moral imperative. And your why of lever, why do I make this choice over that choice? So, if you've ever played chess. You know your why of purpose might be to win the game. Your why of movement is do I move this over there or that. And it's through that movement that we learn more about how to play the game. Assuming, of course, we can understand the landscape.
[00:09:19] Now, to me that was fascinating. Um, it sort of made sense, uh, I, I, I, but I started questioning this whole area of strategy.
[00:09:31] So I started off, um, looking at history and I came across these, these. Uh very frightening people, Vikings. Uh I don't actually have any real pictures of Vikings, um, so so there you are, that is my Viking. And what I found fascinating about Vikings was the epic stories that they would tell. From Ham and Had US towards Heart, I've added a few instructions in bits and of course, it's not in this language. Uh but you know, enormous epic stories. They were used for navigation. Um, and in part, instructions of how to go from A to B. Now, this particular epic story, and it's quite a long story, actually translates to this on a map. And so what I was thinking about, I started to really think about the importance of maps. And I was going to myself, you know, what would I use to navigate? Would I use some sort of story or would I use some sort of visual map? And I was looking around my business and it was like we can use a lot of stories, we, we, you know, we tell, tell each other stories all, all the time.
[00:10:41] But obviously there's a difference here with navigation. Um personally, for short journeys, if somebody needs to go to the pub from my house, I'll say out the door, turn right, turn, I'll use a story. But anything complex, somebody needs a map.
[00:10:58] And so I started looking more into maps. This is a map of a Roman, uh, town. On this particular map, there's about 30 points of interest or so.
[00:11:11] And if I describe the journey between each point, just simple, so no, um, multiple permutations, just simple A to Bs. Uh, there's about 870 paths, about 20,000 words. And that's not including the permutations, because that would make the figure astronomically big, and it's not, um, including going off the path, going to explore the spaces, it's just for those 30 points. And I thought, well, if I gave somebody a journey as well. A to B, and I showed this is the path. It's quite easy for somebody to look at that and go, well, hang on a minute, there's a quicker way of doing this journey, there might be reasons why you can't, but at least we can have that discussion.
[00:12:02] Now, something I noticed with my company was that wasn't happening.
[00:12:08] Um, one of the things and this has been ongoing is we spend an awful lot of time saying that good leaders are great storytellers. And it's really important to be a good storyteller. And and by that, I mean, um we tell people that, you know, your idea didn't succeed because you sold it in the wrong way or you didn't sell it well enough. It's really good to be really important to be a good story. But that also means that when you challenge somebody's story, then they often react somewhat negatively.
[00:12:41] And they do so because you're actually challenging them. So it's the same way that if I gave you a verbal story of navigation and you question it, you're challenging me, so the guy's like, I know I'm right. But if we put it on a map and you go, oh, you see that A to B, go in that route. You're not challenging me, you're challenging the path on the map, it's the map that's wrong, not me that's wrong.
[00:13:12] So I started to look more into the subject. And I started to look into the whole area of military history.
[00:13:19] And military history, maps have been used for learning for a significant amount of time. This is the Battle of Thorp. So Themistocles, ancient politician Greek general, Greece three independent city states, uh Persians were invading, there about 140 to 170,000 Persians invading. Now what they decided to do was to block off the straights of our, force the Persians into along a coastal road into a narrow pass which is named the lock and there a small number of troops could defend against the larger force. There are about 4,000 Greeks, including 300 Spartans, and that's where we get the story of the 300 from. So they used their landscape very basically as a force multiplier.
[00:14:08] Because, you know, it's difficult to move a huge number of troops through a narrow area.
[00:14:15] So I was looking at this and thinking, okay, I can see that we can talk about, you know, the space, what potential moves, we can scenario plan, we can discuss it, we can learn from the battle, we can organize and coordinate. And so I started looking around my company and I thought, what, what sort of tools do we use to help us design? And one of the tools we used was something called a SWOT. So I thought, well, let's do a SWOT of it. Uh, strengths and well-trained Spartan army, high level of motivation not to come a Persian slave. Weakness, the Efors might stop the Spartans turning up, a truckload of Persians are turning up. Opportunities get rid of the Persians, get rid of the Spartans, uh we're an Athenian, we actually hate the Spartans. And threats, the Persians get rid of us and the Oracle says a really dodgy film might be produced a few thousand years later.
[00:15:06] And so what I did is I simply put the map and the squat next to each other. And said, what would you use to communicate and determine strategy in battle?
[00:15:17] And um it was like position and movement, oh of course I'm going to use a map.
[00:15:24] And uh what I was using was SWATs. Well, what I thought I was using was SWATs. So I thought, well, okay, where are my maps? And so I started to look around my business and I had loads of maps. I had business process maps, I had mind maps, I had systems maps. Okay, so I took one.
[00:15:47] And I took a systems map. And this is a systems map for uh the online photo service. And um, you know, has customer and photo storage, other bits. I took one component, CRM customer relationship management.
[00:16:03] And I moved it. I moved it from there to there. And I thought, how has the map changed?
[00:16:11] Well, it has, the components are there, the nodes are still the same.
[00:16:16] So I thought to myself, well, hang on. If I took a geographic map and I took and moved, say, I picked up say Australia and moved it next to England. Does that actually change the map? And the answer is, of course it does. So why doesn't it change the map here? What I realized after time was we keep on using that word and I'm afraid it doesn't mean what we think it means. Pretty much everything in business, business process map, mind maps, even customer journey maps, systems maps, all of them, strategy maps, all have one thing in common, they're not maps, they're graphs.
[00:16:56] Now to explain the difference fairly simply, the three images at the top.
[00:17:01] are identical graphs. Nottingham, London, Dover, Nottingham, London, Dover, Nottingham, London, Dover connected by two roads, M1, M2.
[00:17:11] The three images at the bottom are different maps.
[00:17:18] And the difference between a map and a graph is pretty simple. In a map, space has meaning. So you can't move a piece in a map without changing context.
[00:17:31] So I thought, well, what I need to do. Is I need to somehow create a map of business. But how on earth do I do that?
[00:17:45] So I looked at the characteristics and I thought, right, a map has three basic primary characteristics.
[00:17:50] One, you have an anchor, like North, magnetic North. Secondly, you have the position of pieces relative to an anchor, north, south, east or west of this. Lastly, you have what we call consistency of movement. So if I go north, I'm heading north, if I go east, I'm heading east. It's not a case that I'm going west and suddenly I find I'm going south, not unless your compass is broken or map is broken somehow. So you've got anchor, position, and movement.
[00:18:23] So I thought I had to recreate those in business. So I started off by thinking about a tea shop. And I thought, what's the anchor in a tea shop? Well, we've got hopefully the public who want to drink tea, and we've got the business who wants to sell tea. So I'm going to put the public and business as the anchors. I can do more, I can add regulators and others as well, I could put shareholders. The public has a need to drink cups of tea, and the business have a need to sell cups of tea. But a cup of tea itself has needs. There's the process by which we create a cup of tea. So we need tea, cup, hot water, hot water needs cold water, needs kettle, needs power. And actually, this gives me position. Because to the public consumer, the cup of tea is very visible to them. Whereas the power to heat the kettle to make the hot water to make the cup of tea is very far away from them. It's very invisible, less visible, can't be seen, it's very distant.
[00:19:24] So actually, by creating a chain of needs, we've got a semblance of position within the chain. What about movement?
[00:19:33] Well, it turns out there's a common pattern by which capital evolves.
[00:19:38] Um officially it's stage 1, 2, 3 and 4, but that's fairly meaningless, um so I use labels instead. You start off with the genesis of novel and new items, custom built examples, products and rental services, commodity and utility services. And so I simply took the map and took pieces and put pieces in their relative position. are away from them. It's very invisible, less visible. can't be seen. It's very distinct. So actually, by creating a chain of needs, we've got a semblance of position within the chain. What about movement? Well, it turns out there's a common pattern by which capital evolves. And officially, it's stage 1, 2, 3 and 4, but that's fairly meaningless. Um, so I use labels instead. You start off with the genesis of novel and new items, custom-built examples, products and rental services, commodity and utility services. And so I simply took the map and took pieces and put pieces in their relative positions. So now what we've got is an anchor, and we've got position, and we've got movement. And this is a map, and if I move a piece in any direction, it changes the context. So what can I do with this?
[00:20:15] Well, somebody can look at my map and say, I'm missing staff. And somebody might say, oh, staff that should be more commodity robots. Somebody else can come along and they can add financial figures because these are stocks of capital and flows of capital. So they can add financial figures and put a PNL together. Somebody else might come along and say kettle. Why are you custom building kettles? Why aren't you using standard kettles? That might be a good idea. Somebody from marketing might say, oh no, it's brand exclusivity. We promote our use of these special kettles or whatever. We make the kettles more visible to the users. Which is why I put another chain. Because if you know, in a normal chain, kettles are far away, we might be promoting them somehow. And the point about this is fairly simple, it doesn't matter whether you're from engineering, operations, business, finance, marketing, even HR. We can all actually all talk about the same space using a common language.
[00:21:22] Which brings me on to trapped by context.
[00:21:29] So this is to do with insurance companies.
[00:21:35] And this is a process flow for a particular insurance company. So the first map I did was 2005. This, well, this was about 7-8 years ago. So this is their process flow. They needed to compute, they ordered service, service would go into goods in. They would modify mount rack them. And they had a bottleneck to do with modification and mounting of service. And so they'd spent six months working on this, and they were trying to work out how to get rid of this bottleneck. And they came up with this plan of using robotics and they'd invested into looking into companies and got them in and presented. They'd produced a wonderful business case. Fabulous return on investment sub one year, um, in terms of because there were also failure rates involved, and this was going to improve the process flow and speed up their ability to get new capacity because they were growing. It was fantastic. Okay? So, sub return on investment sub one year, wonderful business case, lots of spot diagrams, lots of competitor comparisons or customer comparisons, or vendor comparisons, all that all that sort of stuff thrown in there. Beautiful. Now, the problem is, I couldn't go and say, why do you need robots? Because I'm challenging their narrative. So we give straight into that punch up, that fight, uh, because I'm stepping on their saying they're wrong. It's a political battle. So I asked them to simply map it. And this took 15 minutes.
[00:23:04] They went user needs compute. Compute needs order service, service that needs a server and that needs goods in. Now ordering service, server goods in, they said is more commodity. Compute they put in product, don't know why, but there we are. Uh, we can challenge it a bit later. And then they went rack mount modify.
[00:23:23] I was able to simply look at that and say, um, why have you got rack in custom built? And the answer came back, they had custom-built racks made by another company. Fair enough. What are the modifications you're doing to service? So the answer came back, they don't fit our racks. So we have to take new, the cases off them, drill new holes, add new plates in order to get them to fit our racks. And that's what we need robots for. Yes. To which somebody in that room went, hang on, why aren't we using standard racks? To which the whole thing changed. It was like, huh? So I wasn't having to challenge, they weren't. Now these people are not daft. They're very sensible people. The problem is, there was a time in the past where using custom racks actually made sense. And they were trapped by that context. So they were constantly trying to improve their process flow, which I've shown here in blue.
[00:24:20] And what they should have been dealing with is evolutionary flow. This is probably the single most common problem that I see. A huge numbers of projects which are a complete waste of time and effort, and money, and resources, where we're trying to get rid of bottlenecks. and improve process flow, when in fact, the thing itself has evolved.
[00:24:41] And again, this isn't because people are daft, this is because they cannot see the environment they're working in.
[00:24:51] So I'll give you another example. Um, this is HS2, high-speed rail, big heavy engineering project UK.
[00:24:59] Um, this is James Finley, who's the CIO. Now, before I go on, can somebody in the chat say hi? Just to make sure I'm still here and you're still here. Oh, fantastic. Thank you.
[00:25:11] So, James's problem was this.
[00:25:16] We needed to build the entire railway in a virtual world. Because it's cheaper to dig up a virtual world and go, oops, we've got that wrong, uh, than it is to dig up the English countryside. So they put a plan together for building the entire railway in a virtual world. And this is the systems diagram they came up with. And James's problem was this.
[00:25:39] Which bits do I outsource, which bits do I build off the shelf, which bits do I build in house? Now, those three questions in that diagram have 387 million possible permutations. So he's going, which one do I do?
[00:25:56] So normally, what I used to happen in government, we'd make life easy, we'd just go, let's outsource the whole lot. And then what we would do is we would overlay lot structures where we group together things which sounded similar. So we'd have lot one engineering, let's put that out to tender. To some engineering folk. Uh, lot two, user experience. Let's put that out to some user experience. Lot three, that seems like back office. Okay?
[00:26:24] So, what James did was something different. James sat down on a Sunday, uh, spent a couple of hours and and just simply mapped it out. Then he sent it to me, he's an old friend. And said, um, look at my map, and I took his map and I just tidied a few things up. He said, how do I apply methods to this? And I said, oh, that's easy. I had that problem years ago. So, we'd gone all agile in 2002, 2003. Of course, by 2004, we realized it doesn't work everywhere.
[00:26:55] What we learned is that when you have a map like this, all these components are evolving due to supply and demand competition. They're moving from left to right. So they're moving, it doesn't matter, money, penicillin computing, all starts on the left in this uncharted space, chaotic, uncertain, unpredictable. Eventually becomes industrialised, ordered, standard, boring, dull. Actually there's three changes of characteristics. And what we'd learn was these as these characteristics change from the uncertain, unpredictable, changing to the ordered, known, measured, stable and dull. that there is no such thing as one size fits all methods. We'd learned that extreme programming was very good on the left-hand side because it was all about reducing the cost of change and the cost of change is normal. On the right-hand side, we'd learned that Six Sigma sourcing was best because what we needed to focus on was reducing deviation. Whereas in the middle, well, we were using different methods, but you you would now definitely say Lean, Scrum, uh, uh, Kanban, MVP, all that sort of stuff, um, where you want to be focused on learning and reducing waste.
[00:28:08] Now the problem is, if you go to a conference like I sometimes do, an Agile conference and say, not Agile everywhere, or go to a Six Sigma conference. And say, not Six Sigma everywhere, then people will just go, burn him heretic. Um, but the reality is, you need to use multiple methods.
[00:28:27] So I just sent him back and we had a discussion, that's what we did. Outsource the stuff on the right. Off the shelf products in the middle. Agile in house development on the left. Use appropriate methods. Now this was 2012, uh, this ended up in front of the public accounts committee, uh, because it was way ahead of schedule, massively under budget.
[00:28:48] So it was praised, very simple.
[00:28:51] Now,
[00:28:54] normally, as I said, we do the outsource everything, overlay lot structures, what would have happened? Well, let's take them out and outsource the whole lot.
[00:29:05] Okay?
[00:29:07] Let's overlay one of those lot structures, engineering. So we want to put this in a big contract and because we want to know what's being delivered, et cetera, et cetera. Well, there's your problem.
[00:29:20] Because the bits on the right inside that lot will be efficiently treated because you can specify and define them. But the bits on the left will always incur excessive change control costs because you cannot define them. And you will get into a fight with the vendor, and the vendor will say, well, it's all your fault because you kept on changing your mind. And then somebody on your side, if you've got people who are actually daft, will go, next time we need to specify it better. No, that'll just make the problem worse.
[00:29:49] You need to use appropriate methods.
[00:29:54] So I'll give you another example, emergency services mobile communication platform. Critical infrastructure, radios, this sort of stuff with police and uh, uh, fire service. So again, big contract about to go to lot structure. Um, I turn up. I'm not the challenge, I said I'd just ask a question, what's the user need? Understand that the user need is the first thing you need to do in mapping. Uh, to which they pointed to the contract, it must be in there somewhere. All right, let's just map this out.
[00:30:27] So they started off, you know, the user needs point-to-point communication, emergency function, job dispatch. And then they added all the underlying components. Um, yeah, that took a couple of hours. Once you've got that, it's pretty simple. Uh, you can overlay the contract structures that you need. Uh, you need to outsource this stuff, we need to use agile techniques over here. We need to use off the shelf products where possible, being mindful that stuff will eventually evolve and the techniques will have to change as it does so.
[00:30:59] Um, that was like too complex, so they decided to go with the old lot structure. And of course, that ended up in front of the, it's been a massive cost overrun. And all the rest of it as well. So, you know, the interesting thing was I was actually able to tell them which which lots were going to be a massive cost overrun based upon this map before they started. So I often do that as well.
[00:31:23] But it's not lost. Because one of the beauties about the maps is you can share them. So I had borders in the immigration and people in the police map shared them, and of course we started to discover common components. And uh, um, that we were building what everybody else was building. So you can just take the components and build a profile diagram. And start asking your question, why are we building multiple user registration systems? Why are we custom building over here when it's when we're buying a commodity like service over here? Now, if you think this is a pop-up government, don't. The worst duplication I've ever found in government is 118 workflow systems doing the same thing. We've managed to build prisoner registration 118 different ways. I've got a farmer company who have 350 teams building enterprise content management systems. And five global efforts to build the global enterprise content management system, none of which knows the others exist. The worst I've got is in the banking world where they managed to build risk management over a thousand times. So, you know, people go, oh, well, government is inefficient. Government is a uh, um, a virtue of efficiency compared to the nonsense that goes on in the private sector.
[00:32:41] So now we're getting into patterns.
[00:32:45] When I look at a um, this this strategy cycle. Once you start to understand the landscape, you discover there are patterns. There's about 30 common economic patterns of climatic patterns that are used for things like anticipation of change. There's about 40 universal principles or doctrine, uh, which I use in terms of organizing companies and also in M&A. There's about 110 different forms of gameplay, ways of manipulating the market to your favor, which I use in things from nation state company to nation state, sorry, to individual companies. Um, most people are oblivious to all of this because you will not be able to see it if you cannot see the environment you're operating in.
[00:33:28] So let's deal with climatic patterns, the rules that influence the game. And by the way, all of this stuff is creative commons, it's I open source it all about 15 years ago, it's loads of stuff you can read online. So these are the rules that influence the game. To do this, I'm going to use a
[00:33:48] obviously enough, a map user application, best coding practice, runtime, operating system, best architectural practice, built on compute. So it's just building an app using good coding practice, runtimes, operating system. It's all basic standard stuff. Right. First pattern you learn, everything evolves.
[00:34:07] If there's supply and demand competition, you end up going from a world of computers as product to computers as a utility. Of course, compute started somewhere, Genesis, so 1943, there's three. And custom built systems lately here.
[00:34:22] Second pattern you learn is you have inertia. Uh, mostly because of pre-existing capital, past success, you know, pre-existing practices.
[00:34:32] Best example of that is probably Blockbuster Netflix. Blockbuster, first with the website, first with video ordering online, first with video streaming, first to go bankrupt. Why? Well, it wasn't because it didn't out, didn't innovate, it out innovated everybody. The problem with Blockbuster was it, uh, was very successful. Because of late fees, people bringing those cassettes back late. That's where it made most of its money. And that's what created inertia. And Netflix didn't have that problem.
[00:35:05] The next pattern you learn is as things evolve, you get coevolution of practice. Particularly as we shift stages. So we end up with best architectural practice for computers as product. Uh, and then we had an emerging architectural practice, which in 2008 was named uh, Andy Patrick called it DevOps. 2008, 2009, something like that. Um, that was for computers as utility.
[00:35:30] And the reason you get a change in practices is because of the change of characteristics. So we go from what's called high MTTR, high meantime to recovery. Your machine goes bang, it takes you weeks to get a new one. So we do things like N+1, uh, disaster recovery test, uh, lots of capacity planning, to a world where you have low MTTR. So it takes seconds to get a new machine.
[00:35:51] So now we distribute systems designed for failure, chaos engines. And that's the new emerging practice.
[00:35:57] Now, along with the efficiency and the speed that the new practices create, uh, they also enable you to build new needs quickly. So, uh, we get things like Netflix. And so what you also learn is efficiency enables innovation.
[00:36:12] And that's simply componentization. Herbert Simon's theory of hierarchy, nothing more than that.
[00:36:20] Okay, and those higher order systems create new sources of value or worth. So, for example, with electricity, we got as that went from uh, products like Siemens generators to more utility provision with Tesla and Westinghouse, uh, we got things like uh, Hollywood, teletyping. We got computing, radio, um, entire new industries being created. So higher order systems create new sources of work. But remember they are in that uncharted space, so they're also uncertain as well.
[00:36:53] Now, because of this, you don't have a choice to adapt. And this is known as the Red Queen effect. If uh a market is changing and your competitors adapt, uh they get the benefits of efficiency, uh speed, uh new sources of value. that creates pressure on you to adapt and as more adapt, uh, you the pressure mounts. Right? This is why things like cloud, you never had a choice. It used to be wonderful, I used to hear so many myths back in 2006, 2007. People running around saying, Amazon's selling its fair capacity. It was complete gibberish. I mean, it was built as a standalone service. And you used to run, Jesse Robbins used to run Amazon's data centers, wouldn't allow this AWS stuff in their data center, so it's built standalone. And then you hear people say, oh, you've got a choice. You had no choice.
[00:37:46] Now, from my point of view, it was good for me to identify where to invest. So, um, you know, um, where to focus, focus on cloud, emerging practice. So this is back in 2007, 2008. Uh, it also tells me where not to focus.
[00:38:04] And so I say 2007, 2008, well, 2008, 2009, uh, that's because I used to run strategy for Ubuntu. So using the map, I spent half a million, took us 18 months, we went from 3% of the operating system market to Red Hat and Microsoft. And to go over 70% of all cloud computing. We did it in 18 months and half a million because we knew exactly where to focus. That's what the maps brought us.
[00:38:31] Now, of course, the emerging practice then continued. Uh, eventually becomes more of a good, heading towards death. Uh, best practice has got a name, DevOps. And uh, best architectural practice in the computers of product world gets the name legacy.
[00:38:47] Runtime, uh, that started off in 2014 with AWS Lambda.
[00:38:55] We're now getting emerging practice, so very much, uh, uh, FinOps, the combination of a capital flow within systems.
[00:39:02] New needs being created. I mean, that's where you need to focus.
[00:39:07] Today. All the rest of this stuff is now the new legacy. Um, so the first lesson from this is what you learn is one strategy is iterative. What was right in say 2008, 2009 is not right in 2020.
[00:39:24] I mean, and people come and say to me, uh, we're going to go cloud and do DevOps. Yeah, that's great if you're in 2010. Uh, but 2020, well, you're going to take at least seven, eight years to implement that stuff. Uh, assuming you're as fast as Netflix, because it took them seven to eight years, uh, we'll we'll say you're not as fast. So we'll say 2030 by the time you get that up and running. In which case you'll discover the entire world shifted to serverless and there you are having built the new legacy. So great plan for 2010, terrible plan today.
[00:39:57] Of course, if you go to a conference and tell people today, DevOps is the new legacy, they will just go, burning him heretic. 2008, 2009 is not writing 2020. I mean, and people come and say to me, we're going to go cloud and do DevOps. Yeah, that's great if you're in 2010. Uh but 2020, well, you're going to take at least seven, eight years to implement that stuff, uh assuming you're as fast as Netflix, because it took them seven to eight years. We'll we'll we'll say you're not as fast, so we'll say 2030 by the time you get that up and running, in which case you'll discover the entire world shifted to serverless. Uh and there you are, having built on your legacy. So, great plan for 2010, terrible plan today.
[00:39:55] Of course, if you go to a conference and tell people today, DevOps is the new legacy, they will just go burn him. Heretic.
[00:40:07] So now I'll get on to how to map. First you, could somebody say hi on the chat? The chat seems to work all the time, just to say hi if you can still see me. Yes, fantastic. Thank you. Right. Um, so how to? First of all, all maps are an imperfect representation of the space.
[00:40:33] Okay, this is a map of France. It's an imperfect map of France. In order to be a perfect map of France, it would have to be the size of France. Uh, because it would be one to one scale. Okay, and as a map, that would be completely useless. So, in order to be useful, all maps are imperfect representations. They're also models, so they're all wrong as well. But despite being imperfect and wrong, uh, they tend to be quite useful, especially for learning. So that's the first thing. Secondly, I use this, uh, in creating something called the Better for Less paper, uh, back in 2009, 2010. This helped, uh, creating something called spend control within UK government and supported something called the formation of GDS, Government Digital Services. So I used to spend a bit of time there. Um, we've used mapping government in all sorts of different places. This is Liam Maxwell. Um, they had one particular project which was created in 2015 and was going to save about 425 million, about 1.5 billion in its lifetime. That was just one, one area that mapping got used. Uh, you'd talk to somebody like Mark Craddock who uses mapping at the UN. Uh, they they reckon they saved about 12 billion, um,
[00:41:45] from one system, uh, in terms of UK economic impact. Um, so it's it's used in lots of different places, but that's just money. That's the stuff that really excites me, is is things like R&LI using, uh, let's James, they use mapping to reduce call out times. Um, and and what I like about that is that actually saves lives. Um, so they they reduced the call out time from about 14 minutes to about 18 seconds, something like that, which makes a huge difference, yeah, when people fall into say the Thames.
[00:42:20] Uh I mentioned Mark, that's uh the UN global platform was all built on uh mapping. Uh, a lot of it's used in venture capital. So, um, this is India, Indian venture capital firm, 48 startups, uh all use uh mapping.
[00:42:38] Um, and bizarrely enough in books, so Punch's Grow uh is a um,
[00:42:44] uh a book, uh science fiction book, described as the new Ready Player One being turned into a film. Uh it was all the book was written with maps as in, uh, they'd use mapping to build a security company which sold for an obscene amount of money to Microsoft, uh, and they'd use maps throughout it. Um, they then wrote a science fiction book, uh, based, uh, and plotted the future with maps. The other one on the left is Amazon's Cloud Velocity book, um, which they's their second ever book. Uh it's got about 15 pages of mapping in there. Um, there's an entire section which basically explains how Amazon tears up industry after industry with a very old model known as Innovate, Leverage, Commoditize. That's a well worthwhile reading.
[00:43:29] So how do you go about mapping? Oh, it's really simple. One, know your users.
[00:43:37] I go back to the cup of tea, you got to sit down and say, who are our users? And then number two, you've got to think about the user needs.
[00:43:47] So what do they actually need? Number three, that's enough, not enough. You need to know the details. So how to make a cup of tea? So you create the value chain.
[00:43:59] Once you do that, you need to understand what is being considered. So you need to to now map it over evolution, because of a kettle, the custom build is very, very different from a kettle which is more commodity. Now, once you've done that, and now you've got a map, you've got a common language. So you've got to share it and allow other people to channel it, well, to add components to it and also to challenge assumptions. To ask questions, why are we custom building this?
[00:44:31] And once you've done that, then by the act of sharing, you should be able to start removing bias and duplication within the system. Once you've done that, you can start to use appropriate methods.
[00:44:45] And once you've done that, you can, well, actually slightly before that, you can introduce a bias towards data. And by that, I mean you can use maps both pre-mortem and post-mortem. So one of the things, obviously with something like spend control is the idea is, uh, before you spend the money, draw a map. Somebody will challenge what you're doing, why you're custom building this, etcetera. And then what you do is you can ignore or listen to the advice and then you go off and do your thing, but afterwards, we do a post mortem. Did you listen? Did it change? What happened? What did we learn? What does the map look like? And that's a cycle and that's the way you get to learn the landscape of a space and also learn how to play the game a bit better. Now, these are all doctrine and they're all built on each other. So you can't focus on user needs unless you know users, you can't know the details unless you focus on user needs, you can't, etcetera, etcetera, etcetera. And I put phase one because doctrine are these universally useful patterns. And I said there was 40 of them, and there they are.
[00:45:53] 40 patterns. There's the phase ones at the top. Most companies are dreadful at this stuff actually. You sit down there, they don't they they don't have even effective methods of challenging assumptions because it's all story-based. And so if you try and challenge a CEO, they're like, oh, up in arms and all this sort of thing. I mean, there is no, everybody talks about really important to challenge. It's just not true. And they don't even have really a bias towards data, it's like business cases. You write a business case, we go do it, we never look at the business case again. So there you rarely is post uh uh event learning, uh post-mortems in that sense.
[00:46:31] Anyway, so there's your unless something, of course, goes dreadfully wrong. There's the phase one, and then there's several other phases. Don't read it now. They're all creative commons. I just highlighted at the bottom there's one called design for constant evolution. Uh this is, um, basically a system known as, uh, Pioneer Settler Town Planner. So it's a cell-based structure which actually has a hierarchical form on top of it, which is designed to cope with constant change in a system. It actually, uh, uses a a mechanism known as theft to achieve this with highly autonomous cells. Um, but it's quite complex and you can only do it if you've done all the rest of the doctrine. Um, it's quite funny because people often hear about it and say, oh, we, we need to, uh, to to become Pioneer Setler Town Plan. And the answer I have to tell them, no. Don't. Uh, because unless you get your doctrine sorted,
[00:47:24] um, it it won't work. Uh you can read more about that in something like boiling frogs, a similar system, so intelligent services GCHQ, uh, it's a freely available document, go read yourself, you'll you'll see the system being described in there. Um, but yeah, as I say, people like to they like to mess around. It's like they're on the Titanic and and people managers love to mess around with the deck chairs. They're all going, oh, it's a quick reorg and it will solve everything. Uh and the answer is no. Uh you need to deal with a great big hole in the side of the ship first. Uh and that basically is the doctrine. Uh and most companies are pretty appalling at this stuff.
[00:48:05] So, I can get on to sovereignty. Now I'm looking at the time, uh, because we had a bit of a, um, uh, a uh, delay, a technical hitch.
[00:48:17] Um, would you like me to dive into sovereignty right now or do or do we have some burning questions?
[00:48:30] Hell, why would you want to boil the frogs? It's an old saying, boil the frogs. Uh, otherwise they will jump out of the water. There we are. Oh, go ahead into sovereignty. All right. This is going to get a little bit complex, a little bit messy, um, because, uh, we're going to go into the issue of culture.
[00:48:52] All right. So physical sovereignty is really easy.
[00:48:57] Well, easy-ish. Uh what you've got is you've got a map, you can draw our border and you say, this is our collective, these are our behaviors, these are our values, if we get ever get around to defining what our values are. many collectives are pretty poor at that, and pretty poor at defining behaviors. Uh collective, by the way, can be family, it could be a church, it could be a nation state, um, just a group of people, but anyway, we can at least say this is the borders, this is the land we occupy. So, so, um, that's that's easy enough.
[00:49:33] Problem is, people want to do that in the digital world.
[00:49:37] Okay. Now, in order to do that, we have to use maps. So, let me start with the HS2 map, high-speed rail. When you look at the map, I told you about there was a principle, use appropriate methods. So doctrine is a collection of universally useful principles. Focus on user needs. Um, understand the value chain, etcetera. One of them, use appropriate methods.
[00:50:04] But when you look at that, you've actually got two beliefs.
[00:50:09] One is a belief of people over process, and one is a belief of process over people. So very much the extreme programming world is a belief in people over process. That's why you've got a couple of very lightweight things. And where in the Six Sigma world, it's much more strongly tied to the process itself, follow the process.
[00:50:31] And that's because one thing can't be defined and one thing can be defined.
[00:50:36] And each of these methods are basically built up to be good in a specific context. But here we have a system being built with multiple methods. So both beliefs, which are somewhat polar opposite, can happily coexist at the same time within the same system. Now beliefs, um, and when we talk about values, that's what we're talking about, things we believe. Um, it means that sometimes you're going to get, uh, beliefs which are polar opposite can't coexist. Uh, sometimes you can get beliefs which do exist. But this touching on the issue of beliefs brings us into culture.
[00:51:17] So this is Prova. And one of my favorite statements is this, despite a century of efforts to define culture adequately, there is no agreement among anthropologists regarding its nature.
[00:51:28] They spent a hundred years, more than a hundred years, trying to define culture and can't agree. So I love it when people say to me, oh, the problem is culture, because I just go, define it for me. Um, and they're very capable people. And the reason why there's a struggle, I suspect, is this. One of my favorite anthropologists is Margaret Mead. And Margaret Mead pointed out that language itself is a discipline of cultural behavior. And why does that matter? Well, that matters because of this. Guru's incompleteness theorem. No model can be true within itself, complete, true and complete within itself. What does that mean? It means if language is part of culture, you will not be able to use language to adequately define culture.
[00:52:28] Which is probably why they spent 100 years trying to do so. So I thought to myself, well, if that's true,
[00:52:36] why can't we use a map?
[00:52:40] We'll break out the boundary a little bit.
[00:52:44] So I, you know, think about maps, when I look at, uh, the, um, labels at the bottom. Uh and I say they're labels. Genesis custom-built product commodity. They actually, they those actually come from this table here. This is a piece of work you'll find, as I say, it's all creative commons, this is something I did back in 2005, 2006, when I did all the data analysis, 9,223 publications, and basically it was four different stages of evolution came out of that work and these are just labels. So when we talk about physical activities, Genesis custom product commodity, if we were talking about practices, we can use novel emerging good and best. Data, modeled divergent, convergent, modeled, knowledge, concept, hypothesis, theory accepted.
[00:53:30] Um, technically, it should be stage one, two, three and four. Uh, but if I start drawing maps which went stage one, two, three or four, it wouldn't mean anything to anyone. That's why I put the labels on there.
[00:53:44] And so when I showed you the map of doctrine, Genesis custom, does that really make sense? Not quite sure. Or we could just use the practices labels.
[00:53:54] Exactly the same.
[00:53:57] Novel emerging good and best. Yeah, that makes a little bit more sense.
[00:54:01] You know, know your users. That's heading that's good practice. Is it best practice? You you you're getting there.
[00:54:09] Surprisingly, it's taking a long time.
[00:54:13] Okay. So I can mix and match. So I can go, I'll take concept, emerging, convergent, accepted. Put that at the bottom. I do that when I'm mapping out ethical values. So I start off with a collective. And it might have values like universal basic income, paid holiday, unionization, and that might be based upon other things like anti-discrimination laws. Which will be based on things like civil rights and worker's rights, which could be based on things like abolition of slavery, which could be based upon other things like reciprocity and fairness. And all of these components are actually evolving values. Ethical values.
[00:54:54] Now, normally, what I do, because there's usually a a lot of values, is I take a section of them, the more visible ones. And I crunch them down into what I call a pipeline.
[00:55:08] So basically, uh, collectives, you know, nation-state, church, football club, whatever, we like to succeed, and by that I mean spreading our values. And, um, there is, uh, you know, our sort of values. It doesn't mean everybody, uh, well, the ones that are accepted are generally pretty uniform, there'll be a lot of disagreement maybe on the emerging ones, um, and more agreement on the converging ones.
[00:55:35] All right, so I've got collective, succeeding values.
[00:55:39] But I normally just show it as a square, and if I expand out from that, I get something like this.
[00:55:49] So this is a map of culture. It is imperfect, and it's also wrong. Yeah, but it turns out it's quite useful in places. So we've got concepts of power and belonging and safety and behavior and enablement systems, and that can be like the town hall to propaganda, doctrines in there, memory, symbols, rituals, heroes, your landscape, game play. These are all components, they're all interconnected.
[00:56:15] And the total entirety of them is is basically culture.
[00:56:24] Several things. One. Collectives aren't singular. And we belong to many. So you belong to your family, your church, maybe your company, your nation, whatever it happens to be. Each of those will have potentially different values, there may be some common values between them. Uh, we have uh, Oliver Eskarri's seven universal values, things like uh, respect for law, respect for property, uh, family, things like uh, bravery, things like that. So it's not singular.
[00:56:55] Secondly, you just can't copy values. You can't go, oh, look at the values of those people. Let's copy them and we'll be the same as them because there's many, many other components involved.
[00:57:07] You can adopt things. Um, you can look at say competitors and you see there's some web giant, really good doctrine, they understand their users, their user needs, they do all that sort of stuff. You can say we're going to adopt those principles. That's fine. Uh of course, what these principles mean will depend upon your landscape. So if I adopt the principle of focusing on user needs, or focusing on users and their user needs, well, of course, that's going to be different if I'm a health care company and from say military, there there are different users and different needs.
[00:57:42] Uh there's also feedback loops.
[00:57:45] So when we get when we're successful in spreading our values, yay, go go democracy, etcetera. Um, and our behaviors are supportive of this, we feel safe in our sort of collective, we're doing really well, great sense of belonging, etcetera, etcetera. Uh, but those can rapidly become negative as well.
[00:58:05] So we start going, oh, look, uh, somebody else is being more successful who doesn't have our sort of values, and so we feel less safe and less sense of belonging. So a lot of these you get positive and negative feedback loops as well. Um, a little bit of this going on negative sides, going on in the UK, uh, with a loss of public trust. Uh, partly kicked off by um, uh, particular individual going for an eye test. Wasn't wasn't the greatest way. All right, when we're talking about sovereignty. So we've got those ideas of collective behaviors, values and landscape, they're all interconnected. But how do we really show this? At this point people go normally, oh this is all complex and complicated. Does it really matter in a digital world?
[00:58:51] Well it does. I'll show you why.
[00:58:55] This is a map of the automotive industry. Uh, we put this together in 2014 and in the DVLA, that's the licensing authority within UK, uh, and rolled the automotive industry forward to 2025. And the reason why it matters is because, um,
[00:59:11] uh UK government makes quite a significant sum of money from, um, people getting driving licenses. Uh, so we wanted to see potentially how things would change. So we started off with a user, what are they needed to be, route management, affordability, car. And the components were much more spread out, but we rolled it forward to where we think it might get to in 2025, which is a world increasingly where we're starting to see more self-driving cars, etcetera, starting to appear. Okay.
[00:59:44] Now, from a nation state competition point of view, the interesting thing is that China is superbly good at identifying components which are heading towards that industrialised state, building special economic zones, creating startups who play a game of last man standing in an internal market to industrialise a space before they get pushed to the world stage. They are exceptional at this. Uh, there's one corporation who's also exceptional at this, which is Amazon. So China basically just moves up the value chain very much on the right-hand side and has done a remarkably good job of this.
[01:00:20] And you're thinking, well, from a car manufacturer's point of view, um, whether it's 2025, 2030, it doesn't really matter, it's going to be start getting a bit miserable because you're not going to be able to differentiate much. But you'll try and differentiate on status. So what we thought is that somebody would come up with a digital subscription model. Um, and, uh, so you don't own the car, you have a digital subscription model, you know, platinum, silver, whatever. And so it's more utility cars, uh, but you get a better experience if you're a platinum member. Sure enough, about four years later, BMW, 2018, I think it was, started talking digital subscription models. Okay.
[01:01:03] Well that's all right. In this scenario, you get into a car, you're a platinum member, it's a wonderful experience. If you're a silver member, you get into a car, it's a wall-to-wall adverts. Okay, fair enough. But from a government perspective, there's a bigger problem, and that's this. When you're in your car being driven along and you're a platinum member, and there's a whole bunch of silver members in front of you. People will go, oh, well, silver members need to move out of the way, and of course, self-driving cars don't do that automatically.
[01:01:33] So platinum member gets to their journey, uh, the journey end faster. So we've embedded social inequality into the transportation system. And that's okay, we may or may not think that's a good idea. But the problem is, when we have a flood, that means all the people with platinum membership get out, and all the silver members, well, maybe bronze members, might be waiting and no way of getting out. So we end up with a lot of people dying because of class. And the next day, well, that generally leads to
[01:02:07] um, pitchforks. It's not a good place to be. So embedding social inequality into transportation systems was a something we were talking about in 2014 as a potential danger threat coming towards us. Mainly because somebody in the market would think digital subscriptions are a good idea. Because they don't think about anything other than money. Now, there's another problem.
[01:02:32] Those intelligent agents are building simulation models. And those simulation models are where values are embedded by the cultures or collectives, sorry, the collectives that create them.
[01:02:47] And you're seeing this already in a little bit. And people go, well, what do you mean? What what values?
[01:02:56] Well, it's the trolley problem. Trolley coming along, there's uh you can kill five people, one person, who do you kill? Well, if you're in a more sort of, uh, uh, say, uh, we community focus society, you might say, well, it's tough luck for the one person. But, you know, what if the five people are unemployed and the one person is a really, really wealthy investment, investor? Well, if you're in a society which values maybe money, uh, it might be tough luck for the five people.
[01:03:28] Um, whichever way it's going to be, those values are going to be embedded into the training models which those AIs are trained on.
[01:03:36] And we're already starting to see this sort of thing. Uh Beijing AI principles, you know, very much about inclusion, um, benefit has got to be for all. Uh, very much a Confucian sort of ideal.
[01:03:50] That may not be the same values that'll get embedded in AI by other collectives and other nations. And of course, when we're talking about sovereignty, it really is about the imposition of values across and across our borders and across other borders. So what we have to do really with digital sovereignty is once you understand what the landscape is, is make those decisions about where our borders are in this landscape. Where our collective, our behaviors, our values need to operate. Which of course, what we do. Except we don't. Instead, what we do is a whole bunch of storytelling. Blah, blah, blah, digital sovereignty, it's all about data, it's all about blah. No map to be seen anywhere, and uh, a lot of people getting very upset if you challenge them. On, well, what do you mean by digital sovereignty?
[01:04:48] So summary. I started off with the issue of strategy. We talked about maps, uh so I went through the whole strategy cycle, the importance of understanding the landscape. Uh maps, uh why maps matter because they're a representation of that landscape. Talked about being trapped by context, which is pretty common. Um, it's not because people are daft, it's because they, you know, it's like Vikings, they weren't daft to use stories, they just couldn't see their environment until they got sunstones and somebody actually drew a map. Um, the, um, lots of people do things which once you map it out, don't make sense. Um, but it's never because they're daft, it's because they're trapped. Um, then we talked a little bit about patterns. I showed you about five economic patterns, there's 30 of those. There's about 40 doctrine patterns, there's about 110 forms of game play, manipulation of the market. So we didn't go through any of that. Um, I talked a little bit about how to go about mapping. And then took a wonder into the whole area of digital sovereignty.
[01:05:54] My tips.
[01:05:56] Well, my tips are just say no to strategy, culture, structure, vision, sovereignty. Lots of people want to talk about those subjects. It's like wonderful. Uh, but if you don't have any awareness of your environment and you don't have those basic doctrine in place, it is pointless. It's just a blah, blah, blah conversation, and, um, it's not worth spending time on, um, until you actually have a basic understanding of the landscape. you're operating in. I suppose that's it. It's all creative comments, help yourself. There's an entire community of mappers out there. Uh something called map camp where people come from all over the world. I mean, I I get sent books all the time. I got one from uh recently the Puerto Rico government got a book which is all about their use of mapping. Really amazing stuff goes on. Uh people using it for combating uh things like human trafficking.
[01:06:46] So it's all great, it's lovely. Uh obviously that's horrible, but it's lovely people combat it, um, with uh, with things like mapping. I mean, um, so, so that's it, I suppose.
[01:07:00] Right, so the last thing, oh, somebody said thanks a lot, Simon. Thank you. So

Transcript