 Yep, I'm started. Okay. So what I wanted to talk about was And I'm a bit wary. Oh You can only come in if you tell me your name What's your name sir? Merrick, okay So I'm just a bit wary because often I give a presentation and I realize I talk about the same old stuff Okay, I'll stop with the name now thing because people keep coming in and keep interrupting me. Um, so because it all really boils down to kind of ancient principles and Yeah, the the meaning of life and wisdom and all that sort of thing But if you get to that too soon, it will sounds a bit like motherhood and apple pie So there's really two slides in my presentation. They're both Venn diagrams For no reason whatsoever because they're not actually Venn diagrams. We'll get to that but Essentially, it's what we're hearing from the market in terms of what's going on and all the cool stuff and then what's actually going on from the organizations we speak to and And and how how software is actually being being written and created and so on and largely I think because so I should introduce myself. So my name is John Collins. I'm an industry analyst I spent 15 years of my career In a real job. I started as a programmer I've got a very very very small reference to to myself and my old student house sharing mate Steve Pate in the French Linux manual Written by René Kunyang who later died, which is really sad But so 1989 I first got got involved in in the Linux thing But not in the way that real people, you know that actually get involved in it. Oh my god, you know that kind of Because it's it's it's been such a movement and it's been so powerful But I was running Sun Microsystems environments and so on so I just wanted so much to play with and so so I got to know the Minix user group and Through that the Linux user group. So that's kind of my I can kind of get away with being here a little bit because of that tiny little It was about what does a demon mean by the way, and I can't even remember but it's something something monitor, but the Then I became 99 industry analysts and the great thing about being an industry analyst is you just think about stuff all the time There's two great things actually the first is you just think about stuff all the time Rather than in a real job. You've got no time to think you're running from meeting to meeting to meeting You're finding out that what you thought was right next week is wrong. You're firefighting all the time You've got and I did have you know Sun Microsystems would come and see me say what do you want to buy and I'm like, I don't know It's all too complicated and literally it was all coping strategies But then when I became an industry analyst actually I just got to think about the stuff I've been trying to think about for 15 years which was amazing Oh, and I missed out the bit where I was a soft development consultant. I did some ITIL stuff and I was working at UK for the UK government So I got all my clearances and I so I'm pretty much mr. Dev sec ops if you want to if you want buzzword bingo I'm pretty much there, but Since then I think working for smaller analyst firms you had to cover everything But really the only things that I care about are those those three things because that's my career if you like so Development operations and throwing it throwing a modicum of security So I can't don't ask me anything about analytics because it just goes horribly wrong very quickly, but that's that's pretty much me in background and I Suppose I should say what what we're going to talk about So as I said, there's there's two there's two themes really is like what is the next big thing we can all talk about that And but then what I want to do is bring that down a level and say so what's actually going on behind the scenes and and what does what does that really mean and jumping ahead to bullet point four is slide eight the supply chain hook and I think this is the last presentation I'll ever do by the way as an aside without ukulele because I really feel I'm just lacking something up on stage, but you know, we'll come back to that and and then How do we actually get there because I have worked in this job for 33 years? And we've been talking about this crap for as long as I have been doing it And we're still saying we're gonna solve I literally had a vendor There's that word Call me up and say we've worked out how to get on top of the software supply process And I was like That's so great. I'm so pleased and we understand you're a bit of a skeptic and I'm really not I really don't want to be I want to really believe that you've this time you've solved it. You've really worked it out But I don't think that necessarily just kind of working out a way of looking at processes is the way to do it Which we'll come to But a little bit about gig on we write reports about technology essentially and We've got a lovely graphicist called Scott who does who does pictures like this and it all looks great And it all looks very professional and so on and so forth And if you want to understand kind of what's behind the scenes of gig arms kind of working model Essentially, we write an RFP just like an end user organization would and then we Genericize have a kind of generic What are the questions you should be asking the vendors and then we just press play and see what happens? And that that's kind of it So so we're kind of reflecting what I used to do when I was buying stuff We're just buying stuff and our kind of secret source as an analyst firm is the people that we tend to use Are people like me that actually had a job beforehand outside of being an analyst so? And as a quick anecdote, which is unrelated, but it's got the punchline in it when I was working as a software consultant I was working for an insurance company and we were building great things And it was all gonna be fantastic and I was in there doing UML and use case-driven design and all that sort of thing And then so we've worked exactly what we're gonna build and then we went to see the ops people and said so We're gonna we're building some new software. We want you to run it and they were like when and he said three weeks time and they went You are joking you you're absolutely joking you're expecting suddenly me to and this is back in the 90s You're expecting me to be able to spin up an entire architecture so you can run your software in three weeks I mean procurement just filling in the forms takes four weeks, you know Which makes another story in my head, but I'm not gonna tell you that but it's about become but Well essentially that was filling in a form in sex duplicate and then throwing away three of the copies Yeah, that back then but He said and he used this expression The people out there have no idea what it's like in here And that expressions really stuck with me right through being a consultant right through being an analyst You just have to have had your feet burned you have to have been shouted at you have to have broken things and think on a Friday evening I thought oh crap or on a Monday morning, which is even worse You have to have let that back up job go and then it says it's gonna take 44 hours to complete and everyone's just coming into work Or you have to have had the pallets of new servers pile up in the car park because you don't know how to plug them in Because the health and safety guy hasn't turned up all of those things are what feed our research so we do all this really nice neat stuff and But this is a bit more how my brain works So I sat down in the cafe once and thought what's actually going on in tech and this came out Kind of by accident, but it kind of and it's starting to be a little bit five years old now But it was just this kind of oh well It's like this and building that and that's kind of how my head works and I'll send you a copy if you like just ask me afterwards but it's essentially What I'm trying to do is marry those two things and I'm trying to marry the very kind of your formalized models and so on and so on with From from time to time by the way the point of the picture is there's all the new and then we're still trying to make it work with the old And it's really interesting now when we're starting to talk about your cloud repatriation or that sort of thing and the multi cloud multi platform on prem Hybrid architecture that it's still very relevant, but that side's starting to push back in again so On to trends which is the whole point of the presentation, so I really need to get to it And how are we doing for time? You will tell me when I'm kind of five minutes before when you the people at the back. You'll start waving madly. I hope so Next big thing Really, we're seeing it in three areas So productivity processes and platforms which I chose that because it all starts with P. I'm not gonna lie But these are the things that we're all talking about so when we speak to the big vendors right now You cannot go into the room without the AI bullet point and it's it's clearly, you know This year it's gone from a a kind of nice to have from a nice to have to a need to have certainly only marketing slides and We're seeing a lot of I mean as analysts a lot of people confuse analysts with journalists So we just get all the PR all the time. Don't we? So many press releases about oh, we've built a new LLM. Oh, we've just applied it to healthcare Oh, we've done this. Oh, we've done that. It's all it's all amazing and it's all fantastic and It's actually quite good So that has anyone played with co-pilot and you know all that stuff it it's A colleague of mine actually tried, you know, just a whatever the the commands would be for you Right me the singularity and that didn't work. Which was a bit disappointing. I have to be frank that it couldn't just build the entire Skynet just there just there with a single command, but it's actually pretty good and it's actually very good at We're seeing it in the security area as you may have seen security tools are now starting to Build in AI every year the way our reports are structured by the way You've got the kind of the things everyone needs to do the table stakes the mandatory requirements from an RFP And then you've got the differentiating Features and then you've got the things that are coming the roadmap stuff and for Three years with AI has always been on the roadmap of every single software category. We're looking at suddenly It's not suddenly it's become a key feature. So that that's really interesting and we're seeing that across security for example, you know sim tools security incident event management tools are now bringing in AI to to Look through all the logs and find the interesting stuff and also preempt by by looking for weird behaviors We're seeing it in UEBA user note entity behavioral analysis. We're seeing it in AI ops and in ops in general in terms of The first job, I mean, it's it's always the boring stuff that happens first But it's always useful boring stuff. So it's noise reduction in AI I used to work as a when I was working for Alcatel doing the Sun stuff. That was a network management so many So many things that you could find out about your your telecoms network and so much of it was pointless So much of it was just rubbish noise. So AI is super useful for those things. Is it gonna break everything that we're doing? That's what the debate is all about right now but I just had I mean jumping around a little bit but When I spoke to a colleague who's just joined us he he's some Darrell he's worked for 37 years in in data management and he said look, it's just more automation. It's always about automation And you kind of go it's one of those things that you there's more to it than that But at the moment he said that I couldn't think of any It's one of those ones. So the debate will continue, but it's it's great and it's it can really help People get to develop things faster. I think it causes problems as well in software. I think it immediately Because the trouble with tech is it doesn't know when to stop and I have this ongoing debate with the hyperscalers hyperscalers Yeah, AWS, GCP, etc. That they deliberately Caused people to spend more money than they should do and now we're talking about yeah cloud cost management and And all that that kind of thing they didn't they just didn't stop people from doing it And would you if someone said oh, we're just gonna we're just gonna do more and more and more with the clouds And then be out of control with it. You know carry on great I'll do as a I remember my my Oracle sales rep when I was when I was back in the job he When we realized that our Oracle spend was out of control and the way that Oracle licensing worked at the time was to have any any unit of Licensing whichever was the bigger one was it users was it processes was it instances whichever was the bigger one that's the one they charge you for and And so we we got completely stung on processes because we had everything as a microprocess Long before Containerization and we got completely stung and all he did was just put on this big grin and said I'm going to the Bahamas I'm going to the but you know So you can't you can't blame blame vendors for just kind of going. Thank you very much, but that's that's really what I digress a little bit, but something that AI can do pulling it back in again is is is help get on top of those costs as well Processes I write my last report was on value streamer. It's anyone here familiar with value stream management I'm going to mention it a bit in a minute So let me tell you about value stream management It is is anyone here familiar with business process management and the whole kind of Drawing out business processes and then looking for whether where you can make the more efficient kind of it's that all right So when I was speaking to I mean that's a bit of a misnomer I went to a cloud bees event cloud bees brought one of their best customers along. It was a big bank it was HSBC and Publicly and on stage they said we've got five thousand different software development processes because we've got five thousand applications and Each one has a completely individualized software element process And I thought that I mean it was kind of thank you But it's a bit like saying you are completely 100% zero competence in your organization around standards or anything else So it's a bit of a it's a bit of a tell But but that's that's the reality. So value stream management is essentially the discipline of Writing down all your processes Working out where the commonalities are and then working out how to get value out of your processes and Just while I finish the thread on value stream management It's it's about two things one is becoming more efficient. It's looking for bottlenecks So, you know, it was developing stuff with glee and throwing it at testing And there's only three people in the test team and they're saying well I'll get round to it in three weeks time and then suddenly you get towards the actual deployment and Everyone goes oh crap. We've we've got like 2,300 feature points that still haven't been tested not even individually It never mind against each other and everything falls in the heap So it addresses that sort of thing for efficiency and within the process But then it also addresses it in terms of effectiveness and if we go back to the original Phoenix project kind of stuff That was all about if I deploy a feature Am I going to see an increase in sales if I deploy a feature our customers going to like me more if I deploy a feature does my net promoter score go up and it so very much the Spotify model they've applied a lot of that too and value stream management enables you to kind of stick some metrics on onto your stuff and Then make better prioritizations and so on so it's great. Most people Dora metrics familiar with that Dora Don't ask for it DevOps It bifurcated didn't it went to go remember more about the politics than the Than the than what the thing is it's metrics about Software delivery, so so it's you know time to resolution It's a number of feature points you can deliver it within a certain period of time There's a lot of noise over that as well really interesting and the tools that are helping deliver on that So kind of the cut-down version of value stream management, which is the value stream analytics stuff So it's just literally having a dashboard onto your software processes all very good all very useful Vendors are talking a lot about it So so we're saying that and then finally platforms, which is kind of I mean was anyone here Did anyone go to cubecon in Amsterdam One I mean what a what a buzz what a what a what an amazing event? It's just literally it's kind of taking in my top three big events where you're just gonna see everyone and Everything's happening and so on and so forth to an aws reinvent and and cubecon They're now my top two it fantastic event Everyone's talking about Kubernetes. I could stick some charts up. Yeah, that's what analysts normally do and they then they say You know, this is where it was. This is where it's going to be etc Three years ago when we were asking about Kubernetes everyone said yeah, we're kind of piloting that So if you asked are you doing it you'd probably get a kind of blanket Yes, three four years ago, but actually you'd find it was the librarian Had built something in his kind of you know, and then when you then when you do the survey if you're actually just only ask the librarian It looks like the whole of Morgan Stanley Bank is doing it because there's there's one little pocket. So so we but we were seeing a lot of pilot studies Literally 18 months later everything flipped and it became the de facto way of Thinking about software development. I'm not say and and we can have a long debate about By the way, I as I said at the beginning I go back to very very old principles So the very very old principle is 1975 Who are the guys? Tom DiMarco and Timothy Lister when they write the paper on structured design it then turned into a book It was all about cohesion and coupling it was about modularization of software because otherwise you can't manage it And then you go into OO and you've got that what shape that those things should be in but ultimately It's about getting the size of the box right and then thinking of the second step is thinking about getting the boxes right so you can actually Have more maintainability but you know containerization all for it fantastic get the size of the box is right and And you are literally fulfilling the dreams of people in the mid-70s of how to write software correctly You get the sizes of the box wrong all hell breaks loose get them too small and you got swarms get them too big And you got a mainframe, so it's not about I don't know why I keep looking at you It's because that light that light. It's shining me. I'll look at you instead. Sorry but it's it's running maybe clearly because it makes me do this it it's Creating as many problems as it causes Not because it's the wrong answer But because people don't know how to build software still properly So that's fine as long as we understand that but when I see things written about our cubanities It's not the thing that it said it. Yes. It's still just doing a thing. It's still just a bit of tech and Similarly, you know, it's same as serverless serverless Started in a fantastic way as well. You won't have to even think about programming You'll just have to low codes the same you won't even have to think about programming because All the hobbits taken taken away and then you actually start to do things and all the old stuff like configuration management like kind of you know version control like The fact that suddenly you're doing three things that you weren't expecting and it's not architected for that and etc so those are what we're seeing as the as the kind of themes if you like in software, so I now feel I fulfilled the What I said I'd do in the presentation and now I can talk about what I like So The actuals are much more simple and more profound as I've said here talked a bit about costs costs are completely out of control What changed in November December January? I'm not quite sure when it changed at the end of last year beginning of this year was money started to cost money So money was free. You could actually have you know Such low interest rates on money that you could borrow for virtually nothing so you could invest without any overhead and That changed with a whole bunch of things happening in the markets with the war with with everything else and suddenly money cost Half a percentage point more than 1% more and then one and a half percent more than it did And if you're investing a million quid suddenly that these start to be very big numbers or if you're investing a billion quid It's not to be very big numbers indeed We're not yet when it comes to software There's an issue with software. It's that no one knows how much it costs at all and what I hear from CIOs is We've been found out. So when money was free. It was can I have some more developers? We're building this thing We're doing digital transformation. It's all great. We're gonna change the world. So can I have lots of money and When I was earlier in my career, I didn't fully realize that some of that was just so I could the people could get stuff on their CVs It's yeah, it's like kind of we're gonna build the most exciting project ever. I'm gonna win I'm gonna look so great. It's gonna look fantastic on LinkedIn But no one was really thinking about the costs or how to manage them and the Valley stream management stuff starts to kick in at That point if they were thinking about it properly first Maybe wouldn't be in this whole but what changed was CIO then goes back or the ahead of development or whatever then goes back to The people with the money on the finance side and says I need more and they say Why he says I don't know and Then it all falls apart because there's no conversation to be had the it cannot show Where the money is going it can only show that it's spending far more than it was going to on stuff that isn't provably delivering value and So we're in a bit of a juncture where Until it's a bit like who threw that in the classroom until you show me how you're spending that money and showing That you are actually delivering value to my organization. I'm not going to give you any more money and we're hearing that again and again and so no surprises I did actually write a blog about this because it kind of Both tickled me and made me a bit annoyed That when aws at a couple of years ago, they stood up and said we never said that it was about saving money And I was like, yes, you absolutely did So I then did a lot of research into and actually they were quite canny. They never did they got their customers to say it I know I know So, uh, so I did that and then I wrote a blog about that but um, uh, the That's why we're seeing so much interest in in cloud cost management in finops I've got a little footnote in the finops book as well. By the way, which I'm very proud of Just because I can say it on stage like this and it makes it me look like I know what I'm talking about but the Costs are a massive thing right now efficiency is a massive thing The juncture is real, but we're not quite over the hump yet. So I know what's coming next What's coming next is computer companies are going to come out with fantastic ways of Saving money. It's all about and what actually I am seeing it. I was on a call with uh, um I don't think this was handy. I was on a call with Cisco. They said Our next goal is simplification of your architecture I was with a storage company. What we're doing next is simplification. It's all about Making things simpler making things cheaper making things easier, which brings me on to complexity because um, no one ever thought about, uh, you know, letting the genie out of the bottle No one ever thought that the genies were going to procreate and we've got lots of little genies running around They're all turning into big genies and they're going into little. Sorry. That's a bit of analogy I'll I'll come back on that one But the long and the short is and the the slide I've used is the the magic sorcerer making lots of brooms Um That's just a reality right now and you know, it's a reality Even in small organizations The more that you build it's a reality in security With the cvs and the provenance of software open source has got challenges there as you know, um about Which library you're actually using so then security companies are coming out with tools So that you can find out which library you're using so you can save money see how this works But getting on top of complexity is a big thing. We're just not there yet in terms of providing the responses to that My top tip is Start simple and it's going to get complex less quickly complexity I've probably talked about in the three decades of work in tech as the number one theme It's the one thing we never get past. We never get past complexity Things always get more complicated than we were thinking they were going to because software is ephemeral And everything in tech is about software until you get down to a transistor level and even those are designed by software I my first job was as as a cad programmer. So You know everything Is made up. It's all imaginary virtualization. It's all imaginary top to bottom and therefore complexity is going to be a thing and then finally the big theme which aws are still not acknowledging And why would they? But they will is is multi cloud On premise never went away someone once likened it to me as the death star. That's not aws. That's your average large company However, much you tinker on the surface There's a heck of a lot going on underneath all the way through right down to the very core That you will never be able to deal with and what big companies have decided And probably small companies and definitely mid-sized companies is because of the first two things There's this so um, I talked to an organization. They said we had a scorched earth policy about cloud Literally, let's get rid of everything move everything to the cloud And then remember what I said about the cfo saying why When the cfo said why and they said because it's better. Yeah, will it save me money? Yeah, how much not very much So why are you focusing on that as this massive and it's turning into you know Spending millions and millions on trying to shift stuff in order to save thousands And it just doesn't make any sense So how about someone said we just leave it where it is Because it's not gonna. Yeah, exactly because it it's it's safe there It's it was already sellotape and string and anyone that's ever seen my programming Actually, my programming was really boring. You know full of comments nicely spaced didn't do much, but it looked very tidy but um So we're seeing the cloud repatriation thing coming. I just saw Four signals. I think I said they just saved millions on on moving stuff out of the cloud But we're seeing it largely In large companies saying how about we just don't That job that we were halfway through shift lifting and shifting workloads into the cloud. How about we just stop that? because it's not A priority for us anymore to have this full on all in on the cloud strategy So those are the things that we're seeing Which is kind of Again, as I say that that's kind of what I came here to talk to you about and And how am I doing for time? Does anyone got Time on 30. I've got another 10 minutes. I think haven't I? We'll take another 10 minutes. It'll be fine It'll be fine. So what I then wanted to talk about was just The the kind of so what which is more it's more slides than just those two But the but it's it's kind of it's more of the wrap-up. So the first thing Is let's face it We're still at the starting blocks Everything that we've seen we've been working in this industry. I mean First computers. I mean, I'm not quite going back to Ada Lovelace that the first computers 50 years 60 years etc in terms of programming and so on It's not been long enough given the the rate of morse law and so on and so forth Generally computer technology has advanced faster than we've been able to mess it up So we've always been behind the curve and therefore we've always felt like we were advancing But morse law is already starting to slow down And I think what we'll start to realize is just what a mess we've made of it And so We'll we'll get better on top of it. Hopefully over the next 50 years. We shall see It is about automation as I say, but I think the thing And this is David where I said your light slide eight because it's about the supply chain The thing that we haven't fully Hooked on to Is that software is a supply chain So we see this in open source with libraries and so on largely But we stop at the tech What we don't do is we take we don't take that supply chain Into the organizations that we're serving into the customers that we're serving And the reason i'm talking about that is because digital transformation We seem to have stopped talking about that one and there's a reason it's because it hasn't been happening And everyone got a bit bored of pretending that they knew what it meant And no one ever really knew what it meant And so we better just kind of put that one to bed But what it actually means Is changing your business models Using technology If you're just building stuff If all it means is if all you do with tech is you buy a new tool and someone actually said to me This was the pretty senior person pretty senior technical. It wasn't a cio, but they said They thought the digital transformation meant buying a new tool Literally, you know get snowflake get you know get a data warehouse get get a data lake get something and then you're digital Get some iot And then you digital but then everything else has stayed non-digital analog in the analog world The digit it's not about digital technology driving business It's about the digital business needing technology and that's the difference So if the digital business needs technology the digital business thrives on A set of requirements that it's trying to fulfill And that then pulls from its suppliers and you end up with a pretty reasonable situation You say hey tech people I need this and that's the way supply chain is working. That's why I say it's digital supply chain So how to navigate this personally? I think the very simple thing is ask why so why am I building this? And if you can't and and there's always more why is you can ask Why am I building this? Yes, but what what is the reason for that reason and what's the reason for that reason and if it doesn't land in the business reason Then it's pointless Ultimately and we should all be doing that all the time and if our bosses aren't doing that Then we should get new bosses. I know that's quite a straightforward thing to say in a room full of green chairs, but um it's literally What the business should be doing and and one of the issues we've got there as you know Is the business still thinks that technology just kind of works So as long as that's true and the technology people aren't telling them that it doesn't Then we've got a problem hold my stomach in while the camera's being out but there you go and uh, so um, That's what we can do personally as a company There are essentially three business models that we're applying. We're either Making ourselves more efficient We're improving our products or we're going out to market better. That's it That's literally it So if we're not able to say that technology is helping us on one of those three things Right from the outset and then as a software vendor By the way, if we're not able to say what those three things are for our customers Then we're still doing it wrong and we see this all the time as analysts Where we're building technology in order to satisfy a need that isn't a business need And if so it's a kind of easy thing to say, but it's also a very good litmus test If your technology that you're building does not deliver on digital transformation Then ask yourself why are you doing it? It's kind of as simple as that and i'm not saying the ephemeral one I'm saying is it changing one of those three business models? Is it helping customers become more efficient? Is it helping them deliver better products? Is it helping them go to market better? That's that's kind of it And we apply that across strategy management and delivery As I say here, so we see it again and again that organizations don't have a strategy for this So it's kind of It's not rocket science. It's like, you know And it's a bit like, you know, Christmas is coming soon and we've all done this It's a very human trait to not have a strategy I will leave my Christmas shopping until far too late and then I will rush into town in the rain and snow and cold Hoping that the things that I'm going to buy will magically appear in the shops and almost be just handed to me And what do you know? It's much much much easier to sit in front of the fire And just with a cup of tea and just write down what auntie more it's going to get this year Oh, she's going to get jam. That's fine. I'll just go to the jam shop. That's all strategy needs to be It doesn't need to be very complicated so Wrapping up a little bit When we're looking at just DevOps, uh, there are I had a big bit of imposter syndrome when I got back into into writing just about development because I thought I used to know it all but these days everything's You know, I'd go to cube con and everyone would be rushing around talking with new language that I didn't really think I understood So I literally spent a year trying to work out How to address my imposter syndrome By talking to a lot of people about what was actually going on and I found out Very handily it divided into 15 things across Five areas which makes for a beautiful picture and But having you know joking apart that that's literally what happened But joking apart we're still looking to address the these areas. We're still looking to Get on top of uh, how we automate to play everything is code We're still looking to work with legacy system We're still and and all of these things have to have to be addressed in some way So if we're looking to get our own house in order as as as the DevOps world Then largely it it's not about carrying on building things and a common theme. I hear is yeah, but people just want to build stuff And that's fine But as a as a business if you're letting people just build stuff and you're not addressing the everything else then You're just costing yourself money. So um We need to get over ourselves a bit. I talked a little bit about value stream management I said it's about dashboards. I said it's about process improvement and so on So I don't want to bang on about that too much These are the areas that we we decided were important and I think that at the one point I will pick up on there is end to end. So A lot of these things ball down to very traditional disciplines for for me One of the requirements management configuration management risk management very very old stuff if you if you don't know how to define a requirement It's worth learning test test if you don't know how to design a test It's worth learning because these are all the things that go around Expertise and then you you build better because you know what the other people are expecting Just by the way, that's what our picture ended up looking like With um, there's no top right by the way This is this is the fit the constant conversation we have about where's your top right? We don't have one Um Because we we we do have a top right which is the kind of older more broad players And then we have a bottom left, which is the the kind of faster smarter more innovative more feature players Which is how our pictures work Um, but but there you go and and do follow up if um, if you want any further information on that But putting this stuff in place to finish then Uh, that old quote by Peter Drucker Ultimately, it does all boil down to the the the culture of organizations because um And I I think the point that I just made about Uh Just letting people do what they like you you can get so far with that but it's costing you money but if you're Assuming that people will just do the right thing then Then that's costing you money too. So So so that's really um the only point i'm trying i'm suddenly realizing i'm i'm saying the same thing in three different ways I'm going to move on very quickly and so it's all about mountains and um, the one thing I would leave on really Is so what have we seen we've seen there's all these new things that we can do with tech We've seen, you know, there's the AI AI stuff We've seen that there's there's the the process improvement stuff, etc. And uh, and actually We're walking in the mud. We're walking in deep mud right now and we're trying to get on top of it So the biggest thing we can do at this point Is start thinking strategically and What i've i've just written a blog about this because what i've realized Is that even in everything that we're doing we're still thinking tactically we're still thinking Oh, well, I just need a little bit. I don't know a kubernetes software security management system And then everything will be okay and all my problems will go away and I can get on with just building stuff again It's it's no it's no longer true and back to the conversation It was never true back to the conversation. I had with that vendor where they thought that they worked out the perfect process stuff They don't know how to get to the top of the mountain. Do you know Killian Jornay, have you heard of him? he's um He's the fastest person up Mont Blanc. I think he's actually now the fastest person off everest He went up and down he ran I mean, don't worry about oxygen deprivation if you just leg it, you know, it's it's fantastic But what what we see again and again with anyone trying to solve for this is they get halfway up the mountain And they run out of steam and I think for software companies. It's often They're just hoping they get bought before they run out of steam Uh, but the bigger companies um Have ideas in order to keep their customers going for another few years So no one's ever really tried to get to the top of the mountain, which is really sad So the point here was we're all on different routes, but beware full summits Because you may think that you're aiming for the top, but actually Uh, the real summit is is eluding you and the reason it's eluding you is back to the The software supply chain thing. It's because we're not addressing the business needs from the start So that's kind of that's my wrap up if you like and if we've got any time for questions I'd happily take them Yeah, okay, let's jump back to that Okay, so, um Right, so what I talked about earlier was The next one Yeah, okay, so I jumped around a bit and kind of held it all in my head So what I talked about much earlier was the way we build research, which is we do it as an rfp process And we have the mandatory requirements, which is essentially Does your software deploy provide dashboards onto your software development process in in this instance? And then we've got the key features which are Can it actually come back to what I what I was talking about just now, which is does it do the business level stuff? Can you do root course analysis? Does it actually give AI driven sometimes linear bees are a great little product for that? For for improvement insights and can you manage stuff as a portfolio? um, which links into project portfolio management practice So all the agile ppm tooling and so on Are largely giving strategists the tools that they need to say here's our Five key initiatives for the organization for the next two years. How do we measure them? Etc. So what value stream management done well does is it plugs the development into the top down? uh strategy and portfolio driven stuff um Yes, kind of so if you look at the icd generally it's Following a reasonable set of steps that you know, you're going to be designing something you're going to be coding it You're going to be testing it et cetera that that's reasonably well mapped But it no one's ever wanted to put too much restriction on those things when it was back in waterfall days It was very much requirements specification design um Development a unit testing user acceptance. It was like a very strict strict strict structured and rigid easy for me to say um Whereas when our child came along it was all just about what can you do this week? Let's have a scrum. Let's have a you know, let's have some uh Some function points. Let's have some user stories. Let's have some epic And it's very much kind of um, and I was a pre agile consultant. I was a dsdm consultant so it was all about time boxing just kind of uh Just just get some stuff done but the the gsd has lost touch with uh, actually those bigger slower ways of doing things so All the tools that you've got you see icd tools they enable you to go through the process But what they don't do is they they don't try to give you any structure particularly So what value stream management does is it acts as an overlay So you can see your process And then see where the problems are within the process. But again, it's not trying to restrict the process I think actually The order I saw it was it blinked first Uh, but they blinked too late for the cloud and then the business blinked and it's kind of Embarously saying we're trying to sort it out, but it's they should they should never have got themselves into that place in the first place so uh, I think the The and the it blinked first definitely about cloud. It was um And the way I see it is very simple. We're moving from The cloud to a cloud. It just becomes part of it It's just a change of terminology. It's super duper important And some organizations will be able to do everything on a cloud But bigger businesses won't So so that was it's blink, but the the cost Moment was the cfo definitely And that was Happening last year, but then it just became excited. It became kind of set in stone if you like over christmas last year But we're still There were a lot of layoffs as you know in tech companies that was that was a kind of tactical move And I think that was largely if if i'm not mistaken that was largely The big players had hired people so that their competitors couldn't get them And they weren't doing anything with them And they suddenly realized how dumb that was So so they then just kind of stopped doing that stupid thing and it looked like they were making a lot of layoffs But actually they were just not holding on to people for no reason Yeah No nit nit pick away. Um, so, um The Hang on just uh, so The whole if you take the expression journey to the cloud it There's a lot to unpack there in that very short expression It says a you're on a journey And it says to wards obviously i'm just getting rid of that word But then to the cloud so it says there is a place That you are going where everything will just work And that's not actually what's happening. There isn't a place where you can go and everything just works and um, you know all of the cloud providers have got different strengths and weaknesses as yours is is kind of um Uh, it's done typical microsoft second mover advantage, but it it doesn't care where your stuff runs It could be on-prem it could be in the cloud as long as you stay in microsoft shop. So so they're happy aws, it's just a bunch of bricks Um, and uh, you manage them all individually Uh, so it it then creates, you know, um, and they're working on that. I know which is great. Um, but the The a cloud thing is The multi so multi cloud every large organization is using at least two hyperscalers now So you can't say we're going to one place because they have completely different places It's not like they're all the same place. It's just different flavors That they're just they're they're just completely different places and the second thing is they're not the only places So you can choose multiple uh cloud providers within An overall architecture. I don't like the term cloud at all because it messes with your head I like the term multi platform architecture But no one's going to adopt that but because but ultimately there's multiple places you can run stuff and you're making a a decision based on The advantages of doing it in one place, you know, I'm a great fan of um, I'm a great fan of kubernetes for that reason because in principle You can run anywhere within, you know, obvious technical Reason the issue reasons And similarly I was a was a great fan of virtual machines because again in principle you can lift them and shift them We're with relatively less uh less effort But keeping in mind that the whole hyperscaler model was to get you in and not let you out You know, hence egress fees hence, you know talking about wall gardens, uh, et cetera, et cetera um, which was great great business model I can't can't fault it um so Yeah, it's a state of mind Ultimately, so so it's a good. It's a good nitpick, but it's getting away from that kind of Ah wonderful place that we can arrive at and it's everything's just going to work Hey, what david What did I miss? Oh, it's gay Yeah, yeah, that that's pretty much it and and then we've got to think well, how are we going to build to that? and so That as I said earlier that everything's ephemeral in tech So we've got two choices either. We let that ephemerality get out of control and we end up with too much complexity that we don't know what to deal with Uh, or we use that ephemerality to our advantage Uh, so i'm very interested in web assembly, for example and what we can do with that. Um Because it's just different places to put the lines Yeah Yep Yep Absolutely 100% Um very quickly because I know we need to we need to wrap up so so The answer is all of these things are problems and they're all driving each other. So the tensions you identify are um The mud that we're running through if you like multi cloud engenders complexity and cost um The the change that we've got to make Is accepting that that's the only thing it's literally If we're gonna if we're literally no one is building for multi cloud We're all still building as though There's only one target. No one's building as though um We're not sure where things are going to be next week And certainly sorry no vendor is providing the tooling. So something like harness Discovered a magical thing when they with their deployment tooling That built in cost management and they said people are using us to see how expensive something is on aws Check the cut pricing on azure and then shift it over if they can save a hundred thousand dollars That was by accident We need that stuff to be on purpose Not not sorry when I say no one's building builder. Yeah developers are Creating great things But we don't have the environments within which this stuff is taken into account It's coming it's coming at the moment a lot of fin-op stuff is just a spreadsheet and a consultant. It's coming Thank you. Thank you so much for um being here