State of Cloud Adoption Report (ESG)

[UPBEAT MUSIC] Hi, Scott. How are you? Thanks for joining. Much appreciated. And I'm very, very excited about what we're going to talk to today. It's been a long time coming. I've seen snippets of the research. I'm very excited by it. And I can't wait to broadcast this to the wider audience. For everyone that's listening that doesn't know you, Scott, could you give a quick introduction? Yeah, sure, Mike. So my name is Scott Sinclair. I'm the practice director at the Enterprise Strategy Group. For those of you who may not know who the Enterprise Strategy Group is, we're an IT analyst firm with a heavy focus on both research and strategy. I'm the practice director. I head up the Cloud Infrastructure and DevOps team. I have a strong group of analysts and experts on my team. And I'm really excited to share a lot of the research that we have today and get into the conversation. Great, great. And I got to say, this took a little bit of working out. And we'll explore that in a few minutes. So all I can say is your team, your analysts, they know their space. So when I came with this idea, and it took a bit of stretching, they were there. They were all over it. And they understood where I was coming from. So a big thank you to you and your team. Thanks. So now, what is this? So this came from, I suppose, a long time where, cloud adoption strategies, I wanted to get a sense of, well, where is it now? There's a couple of research surveys out there that give some good data points there. We got a sense that things were changing from our customers, that the cloud adoption was becoming-- strategy was becoming a little bit more refined, more mature. We couldn't get some real, hard evidence for that. It was all anecdotal. In parallel, I was having a lot of conversations around managing legacy systems, managing-- I was going to say legacy relational databases. But they have their place. So we're having a lot of these conversations around managing legacy systems and modernizing. And when you look at modernizing, cloud adoption is really a subset of that modernization strategy. So that was the nature of when I came to your group about commissioning this study, is, is it possible to get a cross-section of cloud adoption and modernization? And I have to say, you guys did a superb job at gathering the evidence and making sense of the results. So without further ado, could we get a quick introduction from you as to where you want to start and how we want to articulate the results? Yeah, absolutely, Mike. So this was a quantitative research study. We were focused on IT decision makers that had a direct influence over their cloud migration and modernization strategy. About 70% were enterprise organizations. And, Mike, to your point, you brought up maturity, which I think is a key aspect here, is we've been talking about cloud adoption for years, over a decade really. But right now, we're seeing a transition into organizations becoming more mature in their insight, their understanding of the cloud, the benefits of the cloud, and really what they want from their overall IT ecosystem, both within the cloud as well as on premises. And one of the things we do as part of the study, which I'll break out, is try to not just see where people are in their cloud journeys, but really also look at the differences between, how do more mature cloud organizations perceive their overall IT and application ecosystem versus those that are maybe a little less mature or are still starting their cloud journey? Right. And what I saw from the results-- and I've had snippets-- was, for those that are beginning that journey, there's a lot of lessons to be learned there and almost some leading indicators as to what you can expect when you get to a more mature position. Exactly. And I'm glad you brought that up because one of the reasons we look at maturity is not as a way to identify, hey, look, these people are good, and these people are bad. That's not how it is at all. Cloud adoption is a journey for every organization, even the most mature organizations. But what you can do is, by looking at the lessons of the more mature organizations, you can embrace those lessons and take those learnings and apply them to where you are in your own journey and say, look, as we get a little more mature, as we increase our adoption, we're more likely to run into these issues. So how do we prep for those? How do we think about them differently? I have a slide up on the screen that really sets the stage for the sample size and the study overall. And what you see is 52% of organizations say, look, we are mature in our cloud adoption strategies. And so we've had this modernization strategy for more than two years. 41% we call more nascent. They're not quite mature, but they do have cloud adoption. And then only 6% are still on that newly emerging. So if we take a step back, what does this say? Well, it says that, look, we've seen a huge uptick in cloud adoption over the past couple of years. Pretty much every organization has some mix of cloud providers, multiple cloud providers, as well as on premises. But there still is a difference. And even for the more mature organizations, as I mentioned, it is still a journey. This is not a destination. There are still learnings for the mature organizations but also for the nascent organizations, as well, as well as the 6% newly emerging. Right. And for me, that's fantastic. To get that cross-section is super, super important. Could you tell me a little bit about what the research shows about the state of cloud adoption? Or is there anything that you want to touch on on this? Really what I want to get into is, we've kind of set the stage in terms of where people are, but really dive in and look at, OK, what are the big drivers? Why are people moving to the cloud? Because, look, we all think about, hey, look, cloud is good. It has tremendous benefits. So maybe many of us take this for granted. It might be, hey, look, we're moving to the cloud because our executive team told us to. But it is actually more detailed than that. So if we look at this next slide that I'm showing up here, I think one of the things we asked was a very simple question, is, hey, what are the biggest drivers of your modernization, your migration to the cloud strategy? I think one of the big takeaways out of this big laundry list of benefits is it's not all about cost. Right now, you see long-- long-term, excuse me-- long-term cost savings, 40%, right there in about the middle. So there's far more benefits of cloud adoption than just cost. I think that's a key take. The other thing, too, when I think about it-- because we do a lot of research both on cloud as well as on-premises technology-- is the public cloud has some core inherent benefits to it. And we see that at the very top, that 56%, benefits like scalability to flexibility, that ability to leverage infrastructure that is ready and waiting, that gives you the flexibility to spin that up very quickly and scale it very quickly. Now, the others in that box that I have at the top, things like better performance, better compliance, better resilience, all these-- the ability to provide a better experience-- the cloud delivers benefits in all that regard. But often, on-premises technologies continue to evolve, as well, and add benefits. A lot of that, what we see-- and this is relative to everything in the cloud; and it gets back into that maturity conversation-- is so much depends on your own implementation, your own environment. So many organizations are dealing with technical debt or legacy operations or legacy infrastructure. And so understanding which benefits that you're going to get in terms of your cloud adoption depends not just on the app and the cloud, but also your relative experience of the infrastructure you're coming from. And I think that's a key element here. So again, a lot of organizations looking to move to the cloud for a variety of benefits. But often, the mileage, so to speak, is going to vary based on what your experience was and where you're coming from. And a lot of the conversations I have are, let's say, legacy systems. And some of the legacy systems are fine where they are. It's just that when they're mission critical, that's when you need scalability and performance. And that's where the modernization piece comes in. And that's where cloud adoption comes in. And that's what we're seeing. We're seeing these drivers of scalability, flexibility, workload performance. Regulatory compliance is in there, which I'm surprised with. But of course, I'm not surprised with business continuity and resilience. Yeah, and, Mike, the compliance one surprised me, as well. And it may surprise some people. Security in there, as well, at 37%. Because I think there's a narrative that persists, that on premises is better for compliance and security. And cloud is better for other things. That's not really the case. At the end of the day, cloud does a lot of things really well, security and compliance. But again, it goes into that, where you are relatively. If you have lower confidence in your own organization's ability to better understand, to better deploy cyber resilience and cyber security, if you have lower confidence in your organization's ability to better understand the compliance and regulatory requirements, you may see the cloud as providing advantages there. And it gets back to that relative benefit. Something else, though, that I did want to show is-- so this was more the technical benefits. If we go over and look at the business benefits, this also really highlights the opportunity of cloud adoption. I think something that we've been talking about for, gosh, a decade now but I think is important to always hit home in these conversations is technology, whether it's on premises or in the cloud, is not a cost center for your organization. It is a profit center. We have seen-- we've been talking about the digital era of business for a while. But at the end of the day, your ability to deploy digital initiatives, to leverage those initiatives to drive new business opportunities and create new revenue opportunities, your ability to do that effectively has a competitive advantage on your business overall. And so when we look at why people adopt the cloud, again, it's not about cost right there. I mean, there are cost benefits. But it's not just about cost. We're seeing, hey, look, we need better access to innovation. We need to improve our risk management. We need better time to value. We need-- and again, regulatory compliance popping in there. Yeah, I know. That's a big surprise for me because a lot of the talking points are cloud costs, reducing your cloud costs, reducing your storage cloud costs, reducing your compute cloud costs. And I'm looking at there, bottom of the list is OpEx-based cost model. And the other one is-- third from the bottom-- is reduced total cost of ownership. That's a surprise. I think what we're seeing is the last one, OpEx-based cost model is just, hey, can I get out of this CapEx idea to more of an OpEx standpoint? Right. To me, I think that-- [CHUCKLES] and that delivers tremendous benefits. And as I've explored it with organizations, organizations that think about overall budgeting processes and their ability to move the cost of growth out in the future quarters, some organizations get it. But I think, relative to-- but I think, as a narrative, often it isn't as well perceived as some of the other benefits. Reduced total cost of ownership, I think, is really fascinating in terms of-- I don't want to say how much lower it is, because 45% is still pretty strong. Yeah, it's close to half. Yeah, part of that is deemed from-- look, again, you're going to hear this a lot, it's all relative-- when we think about cloud 10 years ago, where on-premises technology took a long time to deploy, often organizations had more of a ticket-based system, where I had to submit for new infrastructure, it may take six months to provision, those sorts of things, there was a huge operational cost to on-premises technology. Also, we also saw on premises also having typically very high costs in terms of their cost per compute, cost per capacity. Pressure, particularly competitive pressure, from the cloud has driven much of the on-premises organizations to innovate a little bit better. So they've gotten a little bit better on that front. We've also seen pressure on driving down the cost of hardware, both within compute as well as capacity. So it's not to say the cloud doesn't still provide TCO benefits. It does absolutely, as identified by 45% of organizations. But that cost benefit is not so drastic. If you're truly comparing the best of on premises versus the best of cloud, what the key part is-- and this goes back to what we were talking about before-- is most people don't have the best of on premises, because they're running legacy infrastructure that has a lot of technical debt, a lot of things that they purchased four, five, six years ago that they're still trying to maintain and maintenance. So as you think about cloud adoption, it goes back to what we talked about before. It truly is not just a migration story, it's a modernization story. How do I modernize these environments? So when I look at this chart, there's two things that pop into my mind. One is, for anyone that's watching this and you're trying to build a business case, yes, total cost of ownership; OpEx, yes; additional soft benefits; better access to innovation feature functionality, cloud native; time to value, cloud native; improved risk management, that scalability and redundancy, you could even argue cloud native there, as well. So for me, this is some fantastic set of data points to include in a business case if you're trying to build a business case for a cloud migration for a legacy system, for example. Second thing that I was trying to get across and just get a sense, a cross-section of the data points was that [? geo ?] distribution. I had a hypothesis that more and more applications were going to be globally, or definitely regionally, distributed. And we didn't see that in the set of results. It wasn't a big driver. And you can see that here. So that, for me, super, super set of insights here. Mike, you bring up a good point. We do see that. I think part of it is driven a little bit by sample size. Even though 70% of these organizations were enterprise organizations, the global distribution of apps does happen. Often, though, it tends to be skewed towards different industries, or more likely to do that, especially much larger organizations. So it may be part of that. But to your point, that can create a set of acute challenges for those organizations that experience it, versus this chart is more, OK, have you experienced these benefits? Are you doing these things? And very natural question-- we talked about drivers, benefits. A very natural question for me to ask is, have you discovered the challenges that organizations, hopefully mature organizations, have come across that less mature organizations, in terms of cloud-adoption strategy, may be able to learn from? Absolutely, absolutely. Let me hit on this next slide here. So again, it looks kind of like an eye chart. But I tried to simplify it for everybody. So when we look at, what are the top cloud-modernization challenges, if we look at that top group, the lighter blue, they tend to fall in really three categories, one of which is, hey, I've got to get the performance right of these apps. If I'm deploying it on the cloud, the performance of an app is not just its performance, its ability to meet its own SLA. But also, it drives customer experience, which can often drive revenue opportunities. It drives your employee experience in often, which can influence operational productivity. So performance is key. And you want to ensure that when you're moving from that apples to oranges environment, off of an on-premises environment, to which you maybe have a more understanded idea of what the architecture is or the infrastructure is, to cloud, you need to ensure that you still get those benefits to performance. We see, in a similar thing, around just migration evaluation. Hey, look, migrations are complex. These are not easy activities. And I don't want to say-- I almost started to say "doing it wrong," which isn't really the case. I think more mature organizations are realizing, look, migrations are not simple. And they depend on a variety of factors that may not be entirely clear at the onset. So for example-- 100%. Yeah. Workloads-- you may believe a particular workload has a certain performance profile, it has a certain priority among your organization, it has a certain type of data set. But we find that organizations consistently get that wrong. And that's where it comes into, is, hey, we thought it only needed low-end performance. It actually needed a much more higher performance because an entire team was using this workload that we didn't actually know about. Hey, we didn't think there was sensitive data as part of that workload. We were wrong. There's actually a whole bunch of customer data in there. We didn't think that needed some sort of compliance environment. All these sorts of things can shift. And even if you get it right, which a lot of people don't, as we see in our study, things can change. The performance environment-- you may move a workload or a data set that may have a lower-end performance profile. But six months from now, it may change. Now, the cloud gives you some flexibility around that. But if it does change, that can change your cost profile, which is that third thing, is hitting into budget constraints. But also, that 31% accuracy of forecasting costs becomes a huge issue because, often, cloud migration or modernization activities are undertaken with a certain TCO or ROI model in mind. And look, if things change, if you have to put it on a different tier, you have to change things, often that can throw that original TCO model out the window. So these are all things that organizations have started to figure out. And I think if we take it as a whole, really, if you're doing a migration, you need to be thinking about performance and benchmarking performance. Hit on that. You need to think about availability not just of the app within the cloud once it's deployed, but availability of that app all the way through the migration process, because migration isn't an instant thing. You're going to have the app obviously running on premises. You're going to move that over. You need to make sure that you're ensuring availability of that app all through the migration itself. Then there's also understanding the complexity of the migration. Those are all things, as well, in addition to things like cost forecasting and data sovereignty, which we hit on, as well. You touched on some emotional trigger points for me, I have to say, data migration specifically. I had a great conversation last week. We were trying to explore risks for data migrations. And I wanted to kind of get a good, long list of risks associated with data migrations. And we didn't even get that far. We got to about five risks, five very, very obvious risks for standard database size for a mission-critical system. And what we learned very, very quickly was for something that size to move, for migration, for five mistakes-- in terms of risks-- something happened, OK, that was a risk, it happened, restart; five-- that's 60 days. You lose 60 days just with those risks occurring. And that surprised me. That's what took me so long. And that's where, anyone that's watching this, spend time in design and testing because that time will pay you back tenfold when it comes to the actual migration. That is such a brilliant point because when we do these studies, we often just identify risks and say, oh, we've identified risks. But in these things, you actually hit those risks. And those risks actually have impact. They have costs associated with it, absolutely. Yeah, yeah. And just to finish-- I did pre-warn it was an emotional trigger point-- so some of our advice is to, if it's mission critical, availability-- I mean, that was a challenge here that's called out in that survey-- if availability is important, then certainly take a look at blue-green deployments for a mission-critical system. That can help reduce the risk and save you time if some of these issues occur. Absolutely. And something else, Mike, that I want to show is the whole time we've been talking about maturity and, how does it impact? And what lessons do we learn? If we look at-- I got another slide here. Showing this on the screen, we broke it out by, the dark blue are mature organizations. And the light blue are those more nascent organizations, so not the emerging but the ones that said, hey, look, we've started our process in terms of cloud modernization. But it's relatively new. What you see is, for most of these things, they're kind of in line. But there's a couple big differences, one of which is that one in the box, regulatory and data sovereignty. Mature organizations are running into this far more often than nascent organizations. Now, part of it is-- my expectation for this is just, as you continue your cloud modernization activity, the workloads you tend to identify for the early migrations tend to not even-- they're ones that you're more certain are not going to run into these issues. And as you modernize, as you get further along in your cloud modernization strategy, you start to explore these workloads that may have regulatory or data sovereignty issues. You know what? That absolutely makes sense to me because you might be-- it's almost like opening Pandora's box, where you've got all these new toys, increased scalability, increased flexibility. OK, let's do more. And the first step is, let's do more, especially around your data management because that's where a lot of the pain points live. Let's try and do more for our data so we can make better decisions and have faster data processing. So that's what we're seeing, is as you jump in with two feet, you need to be very cautious around data sovereignty. That's what I'm hearing from this. Yeah, absolutely. It reminds me of the old joke that the reward for good work is more work. The reward for successful cloud migrations is more cloud migrations. Right. And as you expand, as you start to push the envelope, you're going to run into these data sovereignty and regulatory concerns. And that hits. The other thing, too, is, on the flip side, security is really interesting, as that is more prevalent in the nascent organizations, which suggests that there's some learnings that organizations get as they start moving workloads, particularly around cyber resiliency and cybersecurity, that once they capture those learnings and apply them to future migrations, I mean, it doesn't go away. But it diminishes a little bit. And that's an area where, working with companies that have done these before, thinking about it from a, hey, how can you leverage the right tools and technologies to get a better sense of these environments, can deliver a tremendous amount of benefit. Have we gone full circle here? Has cloud adoption paid off? Or do you think it still needs a little bit more time to push the envelope before it starts paying off? It has. So I'm throwing out a slide here. We ask organizations, across these factors, how much have they improved? I think what's really interesting is we've seen benefits-- in general-- benefits across the board. So 60% of organizations have said, look, we've seen improvements on every single one. Or it's not 60% said yes to all of these. But for each one, 60% or more said yes to each one, if that makes sense. So improved uptime, reduced risk, better compliance, better alignment with what we want to do as an industry. We've even reduced vendor lock-in to some extent. That being said, what you also see, which I think is important to also look at, the takeaway is cloud is beneficial. It is paying off. But I think it's irresponsible if I don't bring up that, if you put the gray and the dark blue boxes on the edge together, you have about 10% to 15% of folks saying, yeah, we got worse than this in this regard, which-- And I've heard of that. I have anecdotal evidence of that, where they didn't get the performance they were expecting. And they repatriated. So this is good-- this is firm evidence of that. Yeah, and I'm glad you brought that up, Mike, because you brought up repatriation. Repatriation, to me, kind of has a negative connotation because it kind of says cloud is bad, which is not the case at all. But what happens is, typically when you do large migration projects or you're migrating a large number of applications, it is difficult to do all the right due diligence upfront. We talked through a lot of that. And often, when you do that, you find out, oh, hey, our diligence was wrong. Or we didn't do enough analysis. Or we really didn't explore the performance characteristics of this particular workload the right way or the proper way that we should have. Or maybe we didn't truly understand the performance characteristics of the cloud service we were moving to. We thought we could get away with a lower tier. Things like that, [AUDIO OUT] all that can play a role. We overprovisioned. And I've heard this multiple times, where we give a specification of a system to them for the cloud, and then they double it just in case. They double it just in case. It's like, no, you don't need to double it. You've just overprovisioned. You're overpaying. Stop overprovisioning just in case. You have the spec. And look, if I'm honest, do some benchmarking. Collect your telemetry over a month, over three months. Collect that telemetry. And there's your baseline for your requirements for the cloud, or specifications for the cloud. Absolutely, absolutely. Go on, Mike, what were you saying? Yeah, no, I was just going to say, so we touched on repatriation. And I have an itch-- [CHUCKLES] --that I need to scratch here, is that mode of lift and shift is, a few years ago, we're moving everything to the cloud. And that repatriation is where some workloads just are not suited in the cloud, because where the data lives and the application lives, you're paying for that data transfer, et cetera, et cetera. So I'm wondering, has that cloud-adoption strategy or methodology changed in any way shape or form? Has the data's-- have you seen anything in the data in that sense? Absolutely. I want to say, hopefully we've all learned something in the 10-plus years that cloud has been around. And I think we're getting smarter. That being said, as the data shows, we're not perfect. We haven't figured this out yet. One of the things we did in the study was we looked at-- is the 6R framework when we think about app modernization. So I highly recommend, if you're not familiar with the 6R framework, to definitely go research that and take a look at it. But it's essentially, when you look at your apps, are you going to retain these? Or are you going to retire them, refactor them, repurchase, replatform, or rehost? And we're all different. Scott, someone in marketing had a field day with those six R's. Oh, yeah, yeah. I wish I came up with that, but no, I'm not that smart. But yeah, it was very much a, there is a marketing take on this that is far above my pay grade. What's interesting, though, is we looked at-- and we as we investigated now, there's a ton of data behind this chart. But I like the chart because of how simple it is. And we looked at, OK, what are the drivers behind each one of these? And again, it's an eye chart, if you're watching this. I'm going to give a quick summary on it. But if we think about why people retain workloads, honestly it's, hey, look, our on-premises environment, either we've modernized that, or it's good enough. So it's things like, hey, look, we've made investments in that to where it's working fine. We don't need to move that anywhere. That tends to be the narrative behind the top drivers. If we think about why we retire apps, it's basically, hey, look, either the infrastructure has gotten too old, or the app itself is just too old. The app has either scalability or performance issues. It can't get us to where we need to go. So this is not essentially the comprehensive playbook of all things to think about with apps, by the way. I should have started that way. But these are kind of rules of thumb, things to think about, or just some ideas of things to think about. That's fantastic. So the top two drivers, deciding drivers, for this particular mode of modernization. Exactly. Retire, is that a mode of modernization? You're putting it to bed. But that speaks to me about legacy systems, where it's like, look, we're going to buy new. Or we're going to retire it. It's just not fitting the business need anymore. Very, very interesting set of drivers there. Yeah, absolutely. And just to round out the rest of the list is when we think about refactoring, these practices were heavily tied to modernization activities. Makes tremendous sense. Hey, look, we're modernizing. We want to refactor our apps and take advantage of more modern architectures. Repurchases is, hey-- it's kind of related to retirement, honestly. Hey, the cloud app is just better. We found a better version of our software either offered as a service or something else. So that made tremendous sense. Replatforming tends to be more cost-related. That tends to be, hey, look, we just looked at an A/B analysis. And the cloud offered better advantages. And rehosting, that decision tends to be more related to scalability and security. But essentially, again, it's not-- just play it back-- this isn't the comprehensive list of all the things to think about. But they give a sense of the top narratives that tend to direct organizations to choose one approach versus another. And speaking the same language as what you've called out earlier, I can see two groupings here, very clear groupings, of refactor and replatform. And then the second one is repurchase. And I think it's repurchase or rehost. There's some very interesting similarities there. I think it's retired and repurchase, actually. But definitely replatform and refactor, where you're looking for increased scalability, increase flexibility. Change of infrastructures necessitating the need to replatform, cloud native, there's a lot of similarities to what you've spoken about before in this one chart. Very, very interesting. Yeah, absolutely, absolutely. And so-- So have you got a handle on repatriation, the scale of repatriation from the data? Yeah, absolutely, absolutely. Let me jump to that real fast because I do have that slide, because I know we talked about it a little bit already. And I think what we see out of this is the top one, or really the top two, are all related to, hey, look, costs or performance don't align with our original expectations, which makes sense. It hits on all those things we-- it hits on the top challenges beforehand. And the bottom line is-- and this is the way I like to think about repatriation-- repatriation happens when the challenges you embrace are so painful that you basically have to acknowledge, hey, look, we made a mistake. And we have to do something different. We have to incur a cost to change. Everything else-- challenges are challenges until they force you to actually pay more money, which is what repatriation is. So again-- and often, it's the result of not doing the upfront due diligence. And then, next, in these buckets, the one in the box so to speak, is around data sovereignty. And similar to what we talked about before in terms of the challenges organizations run into, mature organizations are running into this, as well. So repatriation, again, aligns very nicely with that challenges chart of, hey, look, if you run into these challenges and they're painful enough, this is what's going to drive you to move back-- performance, not hitting all the costs, your cost target, which honestly is very often related to performance. But also data sovereignty, as well, is going to be something that drives back. Other things tend to be, hey, look, there's different audits that happen. Developer preference we still see, people saying, hey, look, no, I like this. There's a number of other impacts there. I'm not going to read through the rest of the chart. No, no, don't. I mean, the data sovereignty, performance-- you're not getting the performance that you expected-- costs-- you're over-spending. Data sovereignty surprised me. I got to say, data sovereignty has surprised me. And that would be a key learning for anyone, is to look into where your data is. And if you don't know where it is, that's a clue to start figuring out where your data is. But further down the list, there's some really interesting-- I saw one earlier where you had access to talent, resource, knowhow, skill. That's very, very interesting, as well. Yeah, I mean, it's still hard to hire people. But something else, too-- and I said this before, but I feel like I need to hit on it-- when you look at a chart of why people are repatriating, it is not, cloud is bad at these things, and on premises is good at these things, or even vice versa. It is, these are the big pain points that tend to cause your organization to say, look, our app or our data, we have to move it. So when you're doing the analysis, when you're leveraging the right tools to better understand it, things like understanding your performance and latency requirements in detail, understanding data sovereignty requirements, understanding how all of that is going to play into costs, these are the big things where people tend to trip up. And again, it's not cloud versus on-premises. It tends to do with understanding your own applications, your own workloads, your own data sets, and what it means for your own organization. Is it-- That is a key point. Yeah? Is it increasing or decreasing? I would expect it's increasing. You mean number of repatriation? Based on what you said. The repatriations. One of the things that we have seen is it's still prevalent. I actually think it's coming down a little bit because organizations have gotten smarter and are leveraging more tools before they do-- and they are starting to do more analysis. Now, granted, I want to say, if I compare this to eight years ago, people were just lifting and shifting everything to the cloud. So there were all sorts of challenges. So we're not seeing a world where repatriation has been near eliminated. No, it's still very prevalent. But it used to be much more wider spread because, frankly, everyone was moving everything to the cloud and doing very little analysis beforehand. OK, so we've reached peak repatriation. But it's tapering off a little bit. And it's still-- Yeah, it kind of went up. And then, now, it's kind of come down a little bit like that, if we're going to draw a curve. OK, interesting, interesting. But you bring up an excellent point because, actually on the next slide here, one of the questions we ask-- if I can get it to change-- is, do you expect to see more repatriations in the future? And I love this as a chart-- well, not because I love repatriations, [CHUCKLES] but because it shows the difference between mature organizations and the less-mature organizations. It's, overall, 55% saying, yeah, actually, this is going to happen. This is going to continue to happen, which, first off, from a psychological standpoint, I love because it's basically people saying, look, I made a mistake. Are you going to make more mistakes? Oh, yeah, we're totally going to make more mistakes moving forward. Basically, they're acknowledging the complexity of these activities. But essentially, that's what they're saying. But what's fascinating is-- There isn't a recipe book. There isn't a rule book. There isn't a book that you can-- people are discovering this. They're trying new things. And they're learning as they go. And that's what we're seeing here. Yeah, not only is there not a recipe book, but the ingredients continue to change. And that's what we're seeing. And with mature organizations, because they've done this enough, they're far more likely to say, you know what? I understand this data. I understand how complex this is, how the stuff keeps changing. So yeah, we're going to continue to see more repatriations in the future, versus the nascent organizations that, they acknowledge it. But a good portion of them may be in the state of, you know what? I feel pretty confident. We learned some things. We think we're going to get it right the next time, when, in fact, the more mature, the more experienced organizations go, you know what? We tried to get it right. The next time, we still got it wrong sometimes. Things just change. That tells me the nascent group just don't know yet. That's what that sounds to me. Yeah, exactly, exactly. That's where it is. And this is part of the experience that comes with it, is understanding the complexity and the difficulty of this. I love that, the recipe book analogy, because there isn't a recipe book. But again, the ingredients continue to change. And-- Trial and error. --the meal that you're trying to prepare continues to change. Right, trial and error. And it's going, oh, yeah, we could tweak something. Yeah. I suppose, what are your final takeaways from the body of research that you've done and any advice that you would offer the audience? My top advice, which I've hit on a few times, is do the analysis. Do the work. Put in the work. The things that tend to trip people up, things like data sovereignty, performance, availability, and the impact that those have on cost tend to be where people get tripped up the most. But we saw, in that repatriation slide, even things like understanding individual skill sets. When you move a workload from point A to point B, you're going to be using a different interface. What does that have on your own team? What is their expertise? Because their own expertise actually influences the results you have, such as with cybersecurity. So do the analysis. The more work you can do up front to understand the performance characteristics of your workload, as well as understanding availability throughout the workload, understanding that is big. Observability is a key part here. We really haven't mentioned it. But as we think about this rise of cloud native development and application development as a whole and the importance of understanding what your app is doing and the characteristics and the experience it's delivering everywhere now, observability solutions can deliver a huge benefit in that regard, especially once you deploy these things on the cloud. And again, remember there's no recipe book. There isn't a perfect solution that's going to be applicable in every situation. Be smart. Do the analysis. As you said that about the do the design, do the analysis up front, I wrote it down. That's tough because there's a demand for speed. I think there's a patience, a certain degree of patience required, knowing that we do need to get this right, that we do need to design. But there's also a balance or a trade-off between speed. You can't spend 18 months in design. So I think that's where organizations should consider, should we get in some experts to help speed up the learnings, to speed up the process, and spend more time in design, as well, at the same time? Fascinating, Scott. Absolutely. And I'm looking forward to digging into the report. Absolutely fascinating. Thank you so much, Scott. Thank you. [UPBEAT MUSIC]