 Hi everybody, welcome to this CUBE Conversation. I'm really pleased to announce a collaboration with Rob Strecce, a guest CUBE analyst, and we'll be working together to extract the signal from the noise. Rob is a longtime product pro working at a number of firms including AWS, HP, HPE, NetApp, Snowplow. Did a stint as an analyst at Enterprise Strategy Group. Rob, good to see you. Thanks for coming into our Malboro studios. Well, thank you for having me. It's always great to be here. I'm really excited about working with you. We've known each other for a long time. You've been on the CUBE a bunch, you know, you're in between gigs and I think we can have a lot of fun together, covering events, covering trends. So let's get into it. What's happening out there? We're sort of exited the isolation economy, things were booming, now everybody's tapping the brakes. From your standpoint, what are you seeing out there? Yeah, I'm seeing that people are really looking how to get more out of their data, how they're bringing things together, how they're looking at the costs of cloud and understanding how are they building out their SaaS applications. And understanding that when they go in and actually start to use cloud, it's not only just using the base services anymore, they're looking at how do I use these platforms as a service? Some are easier than others and they're trying to understand how do I get more value out of that relationship with the cloud? They're also consolidating the number of clouds that they have. I would say to try to better optimize their spend and getting better pricing for that. Are you seeing people unhook clouds or just reduce maybe certain cloud activities and going maybe instead of 60, 40 going 90, 10? Correct, it's more like the 90, 10 type of rule where they're starting to say, hey, I'm not going to get rid of Azure or AWS or Google, I'm going to move a portion of this over that I was using on this one service, maybe I got a great two-year contract to start with on this platform as a service or a database as a service. I'm going to unhook from that and maybe go with an independent, maybe with something like a snowflake or a Databricks on top of another cloud so that I can consolidate down, but it also gives them more flexibility as well. In our last breaking analysis, Rob, we identified six factors that were reducing cloud consumption. There were factors and customer tactics and I want to get your take on this. So some of the factors were you got fewer mortgage originations, FinTech, obviously big cloud user crypto, not as much activity there, lower ad spending means less cloud. And then one of them, which you kind of disagreed with was less analytics, fewer, less frequency of calculations. I'll come back to that, but then optimizing compute using Graviton or AMD instances, moving to cheaper storage tiers, that of course makes sense and then optimize pricing plans, maybe going from on demand to a set of pay by the drink, buy in volume, okay. So first of all, do those make sense to you? With the exception, we'll come back and talk about the analytics piece. Is that what you're seeing customers? Yeah, I think so. I think that was pretty much dead on with what I'm seeing from customers and the ones that I go out and talk to. A lot of times they're trying to really monetize, understand how their business utilizes these clouds and where their spend is going in those clouds. Can they use lower tiers of storage? Do they really need the best processors? Do they need to be using Intel or can they get away with AMD or Graviton two or three? Or do they need to move in? And I think when you look at all of these clouds, they always have pricing curves that are arcs from the newest to the oldest stuff and you can play games with that and understanding how you can actually lower your costs by looking at maybe some of the older generation. Maybe your application was written 10 years ago. You don't necessarily have to be on the best newest processor for that application per se. So last, I want to come back to this whole analytics piece. Last June, I think it was June, Dev at Tachiria, who's the, I call him Dev. Spell Dev, pronounce Dave. Same pronunciation, different spelling. Dave at Tachiria, CEO of Mongo on the earnings call. He was getting hit. Things were starting to get a little less visible in terms of the outlook and people were pushing them. Like because you're in the cloud, is it easier to dial down? And he said, because we're the document database, we support transaction applications. We're less discretionary than say analytics. Well, on the Snowflake earnings call that same month of the month after they were all over Slutman and Scarpelli. Oh, Mongo CEO said that they're less discretionary than analytics and Snowflake was interesting comment. They basically said, look, we're the cloud. You can dial it up. You can dial it down, but the area under the curve over a period of time is going to be the same because they get their customers to commit. What do you say? You disagreed with the notion that people are running their calculations less frequently. Is that because they're trying to do a better job targeting customers in near real time? What are you seeing out there? Yeah, I think where they're moving away from using people and more expensive marketing or they're trying to figure out what's my Google ad spend? What's my meta ad spend? And what they're trying to do is optimize that spend. So what is the return on advertising or the ROAS as they would say? And what they're looking to do is understand, okay, I have to collect these analytics to better understand where are these people coming from? How do they get to my site, to my store, to my whatever? And when they're using it, how do they better move through that? What you're also seeing is that analytics is not only just for kind of the retail or financial services or things like that, but then they're also using that to make offers in those categories. When you move back to more, take other companies that are building products and SaaS delivered products, they may actually go and use this analytics for making the product better. And one of the big reasons for that is maybe they're dialing back how many product managers they have and they're looking to be more data driven about how they actually go and build the product out or enhance the product. So maybe they're an online video service and they want to understand why people are either using or not using the whiteboard inside the product and they're collecting a lot of that product analytics in a big way so that they can go through that and they're doing it in a constant manner. This first party type tracking within applications is growing rapidly by customers. So let's talk about who wins in that. So obviously the cloud guys, AWS, Google and Azure, I want to come back and unpack that a little bit. Databricks and Snowflake, we reported our last breaking analysis. Kind of on a collision course, a couple of years ago, we were thinking, okay, AWS, Snowflake and Databricks, like perfect sandwich, and then of course they started to become more competitive, my sense is they still compliment each other in the field, right? But publicly they've got bigger aspirations, they've got big TAMs that they're going after. But it's interesting, the data shows that so Snowflake was off the charts in terms of spending momentum and our ETR surveys, our partner down in New York, they kind of came into line. They're both growing in terms of market presence. Databricks couldn't get to IPO, so we don't have as much visibility on their financials. Snowflake obviously highly transparent because they're a public company. And then you got AWS, Google and Azure. And it seems like AWS appears to be more partner friendly. Microsoft, you know, depends on what market you're in and Google wants to sell BigQuery. So what are you seeing in the public cloud from a data platform perspective? Yeah, I think that was pretty astute what you were talking about there because I think of the three, Google is definitely, I think a little bit behind in how they go to market with their partners. Azure's done a fantastic job of partnering with these companies to understand and even though they may have Synapse as their go-to and where they want people to go to do AI and ML, what they're looking at is, hey, we're going to also be friendly with Snowflake. We're also gonna be friendly with Databricks. And I think that Amazon has always been there because that's where the market has been for these developers. So many Databricks and the Snowflakes have gone there first because Databricks' case, they built out on top of S3 first and going and using somebody's object layer other than AWS was not as simple as you would think it would be moving between those. So one of the financial meetups, I said meetup, but it was either the CEO or the CFO was either Slutman or Scarpelli talking at, I don't know, Merrill Lynch or one of the other financial conferences said, I think it was probably their Q3 call. Snowflake said 80% of our business goes through Amazon. And he said to this audience, the next day we got a call from Microsoft, hey, we got to do more. And we know just from reading the financial statements that Snowflake is getting concessions from Amazon, they're buying in volume, they're renegotiating their contracts. Amazon gets it, lower the price, people buy more long-term, we're all going to make more money. Microsoft obviously wants to get into that game with Snowflake, they understand the momentum. They said Google, not so much. And I've had customers tell me that they wanted to use Google's AI with Snowflake, but they can't, they got to go to BigQuery. So honestly, I haven't vetted that, but I think it's true. But nonetheless, it seems like Google's a little less friendly with the data platform providers. What do you think? Yeah, I would say so. I think this is a place that Google looks and wants to own is that now are they doing the right things long-term? I mean, again, you look at Google Analytics being basically outlawed in five countries in the EU because of GDPR concerns and compliance and governance of data. And I think people are looking at Google and BigQuery in general and saying, is it the best place for me to go? Is it gonna be in the right places where I need it? Still, it's still one of the largest used databases out there just because it underpins a number of the Google services. So you almost get, like you were saying, forced into BigQuery sometimes if you want to use the tech on top. You do strategy, right? You do strategy, you do messaging. Is it the right call by Google? I mean, it's not it. I criticize Google sometimes, but I'm not sure it's the wrong call to say, hey, this is our ace in the hole. We got to get people into BigQuery because first of all, BigQuery is a solid product. I mean, it's cloud-native and it gets high marks. So why give the competition an advantage? Let's try to force people essentially into what is, we think, a great product. And it is a great product. The flip side of that is they're giving up some potential partner, Tam, and not treating the ecosystem as well as one of their major competitors. What do you do if you're in that position? Yeah, I think that that's a fantastic question. And the question I pose back to the companies I've worked with and worked for is are you really looking to have vendor lock-in as your key differentiator to your service? And I think when you start to look at these companies that are moving away from BigQuery, moving to even Databricks on top of GCS in Google, they're looking to say, okay, I can go there if I have to evacuate from GCP and go to another cloud, I can stay on Databricks as a platform, for instance. So I think it's people are looking at what platform as a service, databases as a service, they go and use because from a strategic perspective, they don't want that vendor lock-in. Yes, or super cloud becomes interesting, right? Because if I can run on Snowflake or Databricks across clouds, even Oracle, they're getting into business with Microsoft. Let's talk about some of the cloud players. So all the big three have reported. We saw AWS's cloud growth decelerated down to 20%, which is I think the lowest growth rate since they started to disclose public numbers. And they said they exited, sorry, they said that January, they grew at 15% year on year. Now they had some pretty tough compares, but nonetheless, 15%, wow. Azure, kind of mid 30s, and then Google, we had kind of low 30s, but well behind in terms of size and Google's losing probably almost $3 billion annually. But that's not necessarily a bad thing. I've advocated them investing. What's happening with the cloud? Is AWS just running into the law large numbers? Do you think we can actually see a re-acceleration like we have in the past with AWS cloud? Azure we predicted is going to be 75% of AWS IaaS revenues. We try to estimate IaaS, even though they don't share that with us. That's a huge milestone. You think there's some people that I think Bob Evans predicted a while ago that Microsoft would surpass AWS in terms of size. You know, what do you think? Yeah, I think that Azure is going to keep growing at a pretty good clip. I think that for Azure, they still have really great account control, even though people like to hate Microsoft. The Microsoft sellers that are out there making those companies successful day after day have really done a good job of being in those accounts and helping people. I was recently over in the UK, and the UK market between AWS and Azure is pretty amazing how much Azure there is. And it's growing within Europe in general. In the States, I think it's growing well. I think it's still growing, probably not as fast as it is outside the US, but you go down to some place like Australia, it's also Azure. You hear about Azure all the time. Is that just because of the Microsoft software estate? Is this so convenient? I think it has to do with, and you can go with the reasoning, they don't break out Office 365 and all that out of their numbers is because they're in all of these accounts because the Office suite is so pervasive in there. So they always have reasons to go back in and, oh, by the way, you're on these old SQL licenses. Let us move you up here and we'll support you on the old version with security and all of these things and be able to move you forward. So they have a lot of, I guess you could say, levers to stay in those accounts and be interesting at least as part of the cloud estate. I think Amazon is hitting the large number, laws of large numbers, but I think that they're also going through, and I think this was seen in the layoffs that they were making, that they're looking to understand and have profitability and more of those services that they have, over 350-odd services that they have and as somebody who went there and helped to start yet a new one while I was there and finally it went into beta back in September, you start to look at the fact that that number of services people, their own sellers don't even know all of their services. It's impossible to comprehend and sell that many things. So I think what they're going through is really looking to rationalize a lot of what they're doing from a services perspective going forward. They're looking to focus on more profitable services and bringing those in because right now it's built like a layer cake where you have S3, EBS and EC2 on the bottom of the layer cake and then maybe you're using IAM, the authorization and authentication in there and you have all these different services and then they call it EMR on top. And so EMR has to pay for that entire layer cake just to go and compete against somebody like Mongo or something like that. So you start to unwind the costs of that whereas Azure went and they build basically ground-up services for the most part and Google kind of falls somewhere in between and how they build their sort of layer cake type of fact but not as many layers I guess. I feel like Amazon's trying to be a platform for the ecosystem. Yes, they have their own products and they're going to sell and that's going to drive their profitability because they don't have to split the pie. But they're taking a piece of, they're spinning the meters, Zia Scaravalla likes to say on, every time Snowflake or Databricks or Mongo or Atlas is running on their system they take a piece of the action. Now Microsoft does that as well but you look at Microsoft and security, head-to-head competitors for example with a crowd-strike or an octa in identity whereas it seems like at least for now AWS is a more friendly place for the ecosystem. At the same time, you do a lot of business in Microsoft. Yeah, and I think that a lot of companies have always feared that Amazon would just throw bodies at it. And I think that people have come to the realization that a two-pizza team as Amazon would call it is eight people. I think that's two slices per person. I'm a little bit fat, so I don't know if that's enough but you start to look at it and go, okay, if they're going to start out with eight engineers if I'm a startup and they're part of my ecosystem, do I really fear them or should I really embrace them and try to partner closer with them? And I think the smart people and the smart companies are partnering with them because they're realizing Amazon, unless they can see it to $100 million, $500 million market, they're not going to throw eight to 16 people at a problem. I think when you could say, you could look at the elastic with OpenSearch and what they did there and the licensing terms and the battle they went through but they knew that elastic had a huge market. Also, you had a number of ecosystem companies building on top of now OpenSearch that are now domain on top of Amazon as well. So I think Amazon's being pretty strategic in how they're doing it. I think some of it will be interesting. I think this year is a payout year for the cuts that they're making to some of the services internally to kind of, how do we take the fat off some of those services that you look at Alexa. I don't know how much revenue Alexa really generates for them, but it's a means to an end for a number of different other services and partners. How do you make this chat GPT? I mean, Microsoft obviously is playing that card. You want to, you want chat GPT in the cloud come to Azure. Seems like AWS has to respond. We know Google is sharpening its knives to come up with its response. Yeah. I mean, Google just went and talked about Bard for the first time this week and they're in private preview or I guess they call it beta, but right at the moment to select AI users, which I have no idea what that means, but that's a very interesting way that they're marketing it out there. But I think that Amazon will have to respond. I think they'll be more measured than say what Google's doing with Bard and just throwing it out there that, hey, we're going into beta now. I think they'll look at it and see where do we go and how do we actually integrate this in because they do have a lot of components of AI and ML underneath the hood that other services use. And I think that they've learned from that and I think that they've already done a good job, especially for media and entertainment when you start to look at some of the ways that they use it for helping do graphics and helping to do drones. I think part of the buy of iRobot was the fact that iRobot was a big user of RoboMaker, which is using different models to train those robots to go around objects and things like that. Quick touch on Kubernetes, the whole DevOps world, we just covered the cloud native foundation security, CNCF, the security conference up in Seattle last week. First time they spun that out, kind of like reinforce. AWS spins out reinforce from reinvent. Amsterdam's coming up soon. The KubeCon, what should we expect? What's hot in KubeLand? Yeah, I think, you know, Kube's, you're gonna be looking at how my OpenShift keeps growing. And I think to that respect, you get to see the momentum with people like Red Hat. You see others coming up and realizing how OpenShift has gone to market as being, like you were saying, partnering with those clouds and really making it simple. I think the simplicity and the manageability of Kubernetes is going to be at the forefront. I think a lot of the investment is still going into how do I bring observability and DevOps and AIOps and MLOps all together. And I think that's going to be a big place where people are going to be looking to see what comes out of KubeCon in the Amsterdam. I think it's that manageability ease of use. Well, Rob, I look forward to working with you on behalf of the whole Kube team. We're going to do more of these and go out to some shows, extract the signal from the noise. I really appreciate you coming to our studio. Well, thank you for having me on. Really appreciate it. You're really welcome. All right, keep it right there. Thanks for watching. This is Dave Vellante from theCUBE and we'll see you next time.