 I think the biggest surprise or the biggest sort of thing to circle and focus in on really is just around customers seeing multi-cloud. It's almost like a split vote at this point right now. It's half say, multi-cloud is the way of the future. I love it. It's the best way for me to do the most with the resources and the teams that I have. The other half is saying it's complex, it's hard, it takes me a lot of time. Why would I go double down efforts on different places when I could just be doing it in one? This is your host, Nibhantia and welcome to another episode of TFR. Let's talk. And today we have with us Jim Murphy, strategic sales executive at Akamai Gym. It's great to have you on the show. Thank you, Swapna. Great to be here. And today we are going to primarily discuss the kind of report that you folks came out with Forrester, the great cloud reset. Talk a bit before we go deeper into this specific report. Just tell us what this report was all about. Also, do you folks do this report? Of course you folks do a lot of reports, but this report and this was like 2023 edition or you do this report a couple of times. So I just want to understand the goal of the report, the frequency of the report before we talk about what we learned from it. In terms of the frequency of the report, I think it's whenever Akamai sees an opportunity for us to understand a bit more around what the trends are in computing, what the customers are interested in and just sort of the things that we should be understanding about the market in terms of the dynamics that are emerging. In terms of the commissioning of this report, the real focus was on just, again, evolutions and multi-cloud, evolutions in cloud computing trends. That's why the title of this is the great cloud reset. Some of the main salient data points that came out of that were just a lot of interest from customers and reevaluated their understanding of their cloud infrastructure and cloud spend in a way that we haven't seen really since the early days of cloud. What are the major trends you are seeing in the cloud space where you look at one side, you're like, hey, no, this is the natural progression of cloud computing. And then other side, you said, hey, we were not expecting these kinds of changes, but the market, the users are driving these kinds of trends. Yeah, no, that's a great question. So the few trends that both the report outline as well as experience of working with some of our biggest customers out there, I think the first is reevaluation of the way in which their applications and services are deployed. You've been in the industry long enough to remember when microservices and service-oriented architectures took over and folks were interested in the ways in which they rebuilt their applications. I think as a byproduct of that now, customers are really starting to evaluate those in which they host and run those, right? So not necessarily just with what provider they're working with, but also, is this something that we should be pushing to the client side? Is this something we should be pulling to the server side? Is this something we should be running at the edge? Does it benefit from more resources sitting at sort of a core site? So it's been interesting to see the way in which customers are reevaluating the sort of deployment targets of these applications and services. So that's the one on the technical side. I'd say outside of that, the requirements around where these applications and services are hosted are starting to take over more about how decisions are being made. So we've heard data sovereignty and sort of regulations in the EU, but a lot of that comes to where is the data store and where is the data process and who has access to that data? And those are real-world implications that are changing the ways in which these services and applications are being built and deployed. So those sort of outside effects coming in, not necessarily just about efficiency or power or speed or all the things that we like to get excited about, there's also real-world implications that are starting to change us. I think those are two really big ones. And the third to your point around cost, right? There's a lot of downward pressure right now on reducing budgets, making sure that people are doing the most with what they have. I think through that lens of sort of reduced resources compared to where we were at three, four years ago, I think there's a big focus on efficiency. So I think the idea around, where can I do the most with the investments that I make? It's starting to push customers to think about not just wholesale partnerships with big providers, it's where can I eke out the best advantages for each of the different things I'm trying to do. And a lot of the cloud technologies, especially in the open source world, are allowing that flexibility. And I think that's where that emergent theme of multi-cloud comes from. Now let's just look at the report, because I want to see the trends and the findings from the report. So let's talk about the reports from the findings, key finding and one thing, and I want to look at it from the perspective of, hey, we were expecting these kinds of outcomes and hey, we were not even expecting, oh, this is the wow, aha moment there. Yeah, so in terms of the outcomes of what we saw, I think the biggest surprise or the biggest sort of thing to circle and focus in on really is just around customers seeing multi-cloud. It's almost like a split vote at this point right now. It's half say, multi-cloud is the way of the future. I love it. It's the best way for me to do the most with the resources and the teams that I have. The other half saying, it's complex, it's hard, it takes me a lot of time. Why would I go double down efforts on different places when I could just be doing it in one? So I think that was kind of the big thing to circle and look at and try to understand a bit more around is sort of where those inefficiencies or where those concerns are coming from. To understand, is this challenges with the technologies, the challenges with the information or the education or is it challenges in just the different ways in which customers are working with their applications and services. So I think those are the big circles outside of that. The thing that popped up that was interesting that we could have expected, but there was the percentage of respondents saying that this was more top of mind was that data sovereignty, which was, hey, we're getting a lot of pressure to move our data centers into the physical locations within the boundaries of different countries or consortiums where, hey, we have to be storing and processing this data in certain places. And that's, again, forcing their hand a bit more. We knew it was important. We've been hearing a lot about it, but the numbers are very staggering to hear that this is a requirement upon them and it's taken a lot of focus. I think those were the big things that came out from the report. If you just look at multi-cloud, multi-cloud does not really mean that, hey, you are like, it's not, they don't say it's like not putting same application everywhere else for depending on the workload, offering from a specific cloud provider. There are so many reasons. Can you talk about one based on the report? Second is based on your own customer basis or the trend that you were saying that what is driving this kind of need or demand for multi-cloud? At the same time, sometime a lot of folks, I mean, we have seen this, hey, we should be on the cloud, we should be using Kubernetes. It doesn't start with solving our problem and this is the right solution. We're looking at the technology shiny object move there. So when it comes to multi-cloud, what is the, so I just want to understand what you're seeing there at the same time. This is why multi-cloud strategy is needed. So in terms of what the report put out, definitely a lot of focus on multi-cloud and I think a lot of the questions that were asked were oriented in a way for folks to be able to understand what's driving this interest and also how do you feel about it? And I think the thing that's driving it from the research and the report, one continued pressure in order to make applications and services as real-time and responsive as possible, right? Like, hey, we've got a lot of focus on making these things really hum. We've got to be able to deploy this in a variety of different locations in terms of either resiliency, locations, right? Some providers are available in different places and we need to be able to satisfy our customer requirements there. Or just, hey, we have development teams who have this familiarity with these different platforms. We want to satisfy all of the abilities of our developers. So let's make sure that we're able to use those different platforms in order to give them flexibility and choice. I think that was some of the stuff that came out of the report. In terms of the experience of having with customers, I think the big interest in multi-cloud is, one, as a means in order to affect better pricing, right? If you're able to work with multiple providers, if there's an opportunity in order to get better pricing because you're able to say, hey, I'm not totally dependent on one. That's just one that I deal with regularly. Number two, and then again, to that point of sort of downward pressure on budgets right now and that efficacy in which they need to use their resources. I think there's a bigger focus on, what's the right platform for the right use case, right? There are wholesale hyperscalers out there who would have general sense that they could do pretty much everything and man, they can do a lot, right? Also, some of them specialize in certain areas as well as Akamai. And I think as customers are starting to think through, hey, how do I reap the most from the things that I'm doing right now? It's which partner is going to do the best in this space, right? So Akamai, we focus very heavily on edge infrastructure, fast delivery and sort of lower costs on the egress because we do have that established delivery network. And resiliency as well as protection for whatever was being transmitted, so you can call it security as well. I think those things combined together make us really strong. So when customers are thinking about workloads in that space, they're saying, oh, I should go talk to Akamai, they do really well in this. And that is really reaping that benefit again of, hey, I need to make smarter decisions. I don't want to be all in on one. I need to pick the ones that are going to do the best for the application and service that I'm trying to run. Also then make sure that I can negotiate good pricing on all of those different platforms as well. That's my experience at least. That brings me also to the point we talked about earlier was distributed application workloads, architecture, environment. Talk about number one is that either what is driving the adoption of distributed cloud platforms. And second is that how prepared cloud providers, organizations are for distributed cloud platforms. And once again, we can talk over hyperskillers. We can talk about, I mean, we can go all the way talk to open stack as well where you can learn your own cloud in your own data center. Yeah, no, so it's a great question. So the sort of two things I heard there is one, talk a bit about the emerging trends you see in distributed and what's the pressure from the customers then too. What do you see from the ecosystem at large that are starting to think through this thing? For the customer side where I've heard from them is there's so much more pressure to make applications and again, application service, you're gonna be saying that a lot but just like the ability to satisfy their customer requests very quickly with no downtime, with really new and interesting ways in order to interact with those apps and services, right? Whether that be gaming or banking or media, you name it. All of them are, hey, we need to make this really responsive. We need to make this really interesting because there's so many different entrants to the market that have access to the same resources. We have to come up with something really innovative in order to capture and retain our customers and make them excited to continue to work with us. I think that's leading towards, our customers thinking through, hey, the Edge offers a promise that we can push more of these requests, closer to these end users in a way that one satisfies that demand for faster, better to also for again some of these regulations and requirements that are coming around to say, hey, this needs to be in a certain place. And if I have access to Edge infrastructure where I can do basically the same thing from these sort of core sites, that's the place I wanna go. Outside of that, where I'm seeing a lot of the ecosystem play with this is one, and to your point around open stack or Kubernetes, right? It's like, what are the deployment pipelines to get those things to the Edge, right? We're so familiarized with this idea, especially with these sort of core mega data centers, either through what people have been managing themselves or offload into the hyperscalers where everything's kind of contained in one, not even one site, but in sort of one provider. But the idea of, hey, I need to be able to push all these deployments or these updates or just like new requests out to the Edge. How do I manage that in a way where I can do that fast and I can do that seamlessly and it replicates the experience that I'm used to in terms of these deployment and development pipelines? I think that's one that I'm seeing the ecosystem is more around like, how do we invest in this to make it work really well? And obviously Kubernetes and things like OpenStack do a really good job with, hey, bringing consistency of what's being deployed, but also like creating that layer above, which is like, hey, how do we pick and choose where we put it, not even in terms of just the Edge, but also the different providers. Another thing that we're seeing in terms of the ecosystem is too, it's just the number of sites and it's like, it's a balance between the number of sites and how much capacity you put at each of these Edge sites. And it's always a hard one to strike because it's like, if you put too much, then you've got to centralize in one area. If you don't put enough then are you gonna be able to satisfy the time and latency requirements that customers have? So I think right now you're seeing a lot of balancing between the amount of capacity and the number of locations. Again, I think that's where Akamai shines really, really well is the fact that we're finding a good strike between the two of, hey, we can get really close into these distributed Edge sites, but also still provide enough capacity to satisfy their demands, while sort of building up more and more and more so that over time you'll see it available as many places as you need it to be. When it comes to, of course, moving data closer to users, of course, you mentioned GDPR, more and more countries, especially in this changing political situation, they want the data locally. Of course, there are laws which also drive a lot of these technologies whether it's come to a distributed platform. Talk about, once again, what you're seeing beyond just Europe because Europe is also in California, they have laws, but what are you saying globally and what would it mean for cloud? What it would mean for data? And of course, that would mean that the technologies, the consumption will also change. And just so we're all clear, I'm not a legal expert in data sovereignty requirements. Yeah, I can say from my experience working with customers, some of the requests that they have, and the trends in the sort of direction which I've seen it going based on those experiences so far. I'd say first is one better metadata or just better insight into where information is going and being sort of to your point. Images and videos are being captured by all of these different devices around the world. I think some folks are challenged right now to really know exactly where those go outside of the folks who are building and running those pipelines. But if an agency were to come along and say, hey, I wanna know where this is going and I want you to be able to show me in a report or an audit that says this is where it's being stored, this is how it's getting there, this is what's protecting it, this is who has access to this information. I think the tools to be able to provide that visibility and reporting are becoming more and more important. And I think they're gonna continue to be more and more critical as folks look at the partners that they work with in order to be able to do that. And I think second over time, the trend that I think will be more important is sort of sectioning out the data and storing it in different ways. You're seeing that right now in the different sort of services that are available, object storage, I've got block storage, I've got non-structured information, right? Like all these different types of information, I do think you're gonna see finer segmentation around the ways in which those things are being handled for things that, again, fall under PII, but they might also be like images and videos that might contain people, right? And that's gonna be even more of a focus over time if people start looking at in terms of privacy, is like, could we put those in places that have even more strict requirements in terms of access, more strict requirements in terms of auditability? Seeing that as a bit of a trend as well. Outside of that, more folks come online, more countries start getting massive populations coming and joining in on the internet. I think you're gonna start seeing, again, more and more regulations that come in to protect their people. And I think you'll see things like GDPR and the stuff in California. I think you'll see that all over the globe too. Since the title is the whole cloud reset, I also wanna talk about some of the new workloads that are emerging, one of the biggest and hottest is Genetic, I know, chat, GPT and all those things, which are also kind of creating a new challenge. Just from the pricing perspective, because it depends on, you know, when thinking how much you're consuming, it's not. Talk about how do you folks look at these, the emergence of this new workload and what does it mean for cloud providers from the perspective of consumption, the whole pricing model may also change, consumption model may also change? Yeah, no, it's a great question. You know, in terms of what we've been seeing for new workloads, there is a whole lot of interest in generative AI or the models that underlie it. It's also coming out on the other side of that too, is sort of a reimagining in where, you know, logic is running, again, you heard me say it earlier, but is it happening on the client side? Is it happening on the server side? And there's so much happening right now on the client side. I think there's an interest to employing some of that off and running it on the server side. And the reason being is I don't wanna over-tax devices that their end customers are using in a way that the experience is poor, so are there things that we can do to pull off pieces of certain rendering parts, like putting that on some sort of server side where the edge comes in and people are interested in looking at doing that. The second in terms of, you know, these new workloads and what are they doing in terms of pricing? You know, pricing is always a very hard challenge to figure out. It's always like trying to draw a box around a bubble that's always kind of moving and trying to figure out what's the best way. One, to provide something that's sort of clear and understood by your customers. But two, you know, you monetize the investment that you're making in these different things. You know, what I've seen is, you know, you get into a business for long enough, like you're down to like microsecond pricing on some of the compute platforms at this point. And it's like, it's good because you continue to narrow that band down. But sometimes you're like, I don't want to deal with picoseconds in terms of numbers, how long this thing is running. So I think that's been interesting is to see the finer grain sort of time scales in which the compute and the applications are running. And again, you go look at things like, you know, functions as a service, you're seeing those very, very minimal amounts of pricing attached to what's being compute. You know, I think you're also seeing a lot of interest in what we would typically call like spot so prepaid time on compute platforms have been very powerful for a long time, especially in one forecasting for the providers to understand capacity and requirements, but two for customers to build a budget and articulate to their CFO, like, hey, this is what we should be expecting in terms of spend over the next three years to support this thing. I do, I've seen a lot more requests around like, hey, we're not sure what's gonna happen with this. We're testing this out. We're trying this. We're not sure where the requests are gonna come in or training models, how are we gonna be doing all this? So it's like, hey, what flexibility do I have in terms of being able to use your platform in a way in which that I can sort of allocate the budget but not also be bound to doing something. I think you're seeing a lot of that. And then number two, you know, less long-term agreements. I think a lot of customers right now are kind of curious where things are going in the economy. And again, with budgets over, you know, what typically had been three and five year purchasing cycles, seeing a lot more interest right now to one to three year agreements, which is like, hey, can we reduce this timeline? Cause we're not exactly sure what's gonna happen in the next few years. Things are changing so fast to your point. We might have to rethink the way in which we're building and running these services. So, you know, can we reduce the time that we make commitments on? So I think those are some of the trends that I've been seeing. Yeah. When you look at the whole cloud, when you look at the Kubernetes, it's not that easy actually anymore. If you just look at the pricing unit to be a data scientist or a mathematician to just calculate the thing. And then when we talk about new technologies, everybody wants chat, GPD, everybody wants Kubernetes. What advice do you have when you look at these reports, when you look at your customers? As you're also saying, you know, that, hey, you know, this is how you should approach it, look at what problem you're trying to solve. So when we look at whether it's distributed, when we look at multi-cloud, when we look at data sovereignty, how organizations, who are still, we can call them, you know, green-filled deployments, they should approach cloud, they should approach, you know, some of these practices. Yeah. No, that's a great question. And I would say, you know, when talking with customers and, you know, they're either bringing new ideas or it's a new business that's trying to come through with something unique, you know, my adage is what's the problem you're trying to solve and what's the tool that's gonna help you solve that the best, right? We get a lot of interest and, you know, you hear the shiny objects syndrome, like, hey, I heard this is really cool. You know, Edge Compute sounds like it's gonna solve all the world's problems. I wanna go do that, talk to them and get a sense of what they're trying to accomplish. You're like, you know, this could help. I don't know, this is the perfect way in which in order to do what you're trying to accomplish. So my advice is, is tending to be, you know, understand the problem you're trying to solve and figure out the right tool is gonna solve that problem for you in the terms of the cloud infrastructure and all the different trends that are emerging and those different things, you know, is that going to satisfy the needs of your customers the best? So when we talk about distributed compute, is that, is latency or requirements the most important thing? Yes, okay, then let's start looking at where those edge deployments might look like. You know, again, data sovereignty, does the information need to be stored in certain places and, you know, process in certain places? Yes, okay, we should start to consider distributed locations to make sure that those satisfy, those requirements are being satisfied. You know, understanding those requirements and those constraints are the things that I think lead towards making those decisions. You know, developers want to have fun. They want to try new things. Folks want to get out there and feel like they should be excited. So by no means is it like, get very focused on just the problem to be solved. You know, great innovation comes from people playing with new things, right? Really interesting stuff has come about by doing so. So I think also giving flexibility for your teams to be able to try new things and explore is always a good idea. I'd say sometimes though, you know, folks get overly focused on, you know, trying the new thing rather than, hey, what are the pain points my customers are experiencing? What's the best way to solve this? That tends to be where I focus. Jim, thank you so much for joining me today. Talk about the support, great insights there and of course, great advice. Thanks for all those insights and I would love to chat with you again. Thank you. Yeah, me too. Thanks, Bob.