(upbeat music) >> From theCUBE studios in Palo Alto, in Boston bringing you data-driven insights from theCUBE and ETR. This is braking analysis with Dave Vellante. >> Database is the heart of enterprise computing. The market is both exploding and it's evolving. The major force is transforming the space include Cloud and data, of course, but also new workloads, advanced memory and IO capabilities, new processor types, a massive push towards simplicity, new data sharing and governance models, and a spate of venture investment. Snowflake stands out as the gold standard for operational excellence and go to market execution. The company has attracted the attention of customers, investors, and competitors and everyone from entrenched players to upstarts once in the act. Hello everyone and welcome to this week's Wikibon CUBE Insights powered by ETR. In this breaking analysis, we'll share our most current thinking on the database marketplace and dig into Snowflake's execution. Some of its challenges and we'll take a look at how others are making moves to solve customer problems and try to get a piece of the growing database pie. Let's look at some of the factors that are driving market momentum. First, customers want lower license costs. They want simplicity. They want to avoid database sprawl. They want to run anywhere and manage new data types. These needs often are divergent and they pull vendors and technologies in different direction. It's really hard for any one platform to accommodate every customer need. The market is large and it's growing. Gardner has it at around 60 to 65 billion with a CAGR of somewhere around 20% over the next five years. But the market, as we know it is being redefined. Traditionally, databases have served two broad use cases, OLTP or transactions and reporting like data warehouses. But a diversity of workloads and new architectures and innovations have given rise to a number of new types of databases to accommodate all these diverse customer needs. Many billions have been spent over the last several years in venture money and it continues to pour in. Let me just give you some examples. Snowflake prior to its IPO, raised around 1.4 billion. Redis Labs has raised more than 1/2 billion dollars so far, Cockroach Labs, more than 350 million, Couchbase, 250 million, SingleStore formerly MemSQL, 238 million, Yellowbrick Data, 173 million. And if you stretch the definition of database a little bit to including low-code or no-code, Airtable has raised more than 600 million. And that's by no means a complete list. Now, why is all this investment happening? Well, in a large part, it's due to the TAM. The TAM is huge and it's growing and it's being redefined. Just how big is this market? Let's take a look at a chart that we've shown previously. We use this chart to Snowflakes TAM, and it focuses mainly on the analytics piece, but we'll use it here to really underscore the market potential. So the actual database TAM is larger than this, we think. Cloud and Cloud-native technologies have changed the way we think about databases. Virtually 100% of the database players that they're are in the market have pivoted to a Cloud first strategy. And many like Snowflake, they're pretty dogmatic and have a Cloud only strategy. Databases has historically been very difficult to manage, they're really sensitive to latency. So that means they require a lot of tuning. Cloud allows you to throw virtually infinite resources on demand and attack performance problems and scale very quickly, minimizing the complexity and tuning nuances. This idea, this layer of data as a service we think of it as a staple of digital transformation. Is this layer that's forming to support things like data sharing across ecosystems and the ability to build data products or data services. It's a fundamental value proposition of Snowflake and one of the most important aspects of its offering. Snowflake tracks a metric called edges, which are external connections in its data Cloud. And it claims that 15% of its total shared connections are edges and that's growing at 33% quarter on quarter. This notion of data sharing is changing the way people think about data. We use terms like data as an asset. This is the language of the 2010s. We don't share our assets with others, do we? No, we protect them, we secure or them, we even hide them. But we absolutely don't want to share those assets but we do want to share our data. I had a conversation recently with Forrester analyst, Michelle Goetz. And we both agreed we're going to scrub data as an asset from our phrasiology. Increasingly, people are looking at sharing as a way to create, as I said, data products or data services, which can be monetized. This is an underpinning of Zhamak Dehghani's concept of a data mesh, make data discoverable, shareable and securely governed so that we can build data products and data services that can be monetized. This is where the TAM just explodes and the market is redefining. And we think is in the hundreds of billions of dollars. Let's talk a little bit about the diversity of offerings in the marketplace. Again, databases used to be either transactional or analytic. The bottom lines and top lines. And this chart here describe those two but the types of databases, you can see the middle of mushrooms, just looking at this list, blockchain is of course a specialized type of database and it's also finding its way into other database platforms. Oracle is notable here. Document databases that support JSON and graph data stores that assist in visualizing data, inference from multiple different sources. That's is one of the ways in which adtech has taken off and been so effective. Key Value stores, log databases that are purpose-built, machine learning to enhance insights, spatial databases to help build the next generation of products, the next automobile, streaming databases to manage real time data flows and time series databases. We might've missed a few, let us know if you think we have, but this is a kind of pretty comprehensive list that is somewhat mind boggling when you think about it. And these unique requirements, they've spawned tons of innovation and companies. Here's a small subset on this logo slide. And this is by no means an exhaustive list, but you have these companies here which have been around forever like Oracle and IBM and Teradata and Microsoft, these are the kind of the tier one relational databases that have matured over the years. And they've got properties like atomicity, consistency, isolation, durability, what's known as ACID properties, ACID compliance. Some others that you may or may not be familiar with, Yellowbrick Data, we talked about them earlier. It's going after the best price, performance and analytics and optimizing to take advantage of both hybrid installations and the latest hardware innovations. SingleStore, as I said, formerly known as MemSQL is a very high end analytics and transaction database, supports mixed workloads, extremely high speeds. We're talking about trillions of rows per second that could be ingested in query. Couchbase with hybrid transactions and analytics, Redis Labs, open source, no SQL doing very well, as is Cockroach with distributed SQL, MariaDB with its managed MySQL, Mongo and document database has a lot of momentum, EDB, which supports open source Postgres. And if you stretch the definition a bit, Splunk, for log database, why not? ChaosSearch, really interesting startup that leaves data in S-3 and is going after simplifying the ELK stack, New Relic, they have a purpose-built database for application performance management and we probably could have even put Workday in the mix as it developed a specialized database for its apps. Of course, we can't forget about SAP with how not trying to pry customers off of Oracle. And then the big three Cloud players, AWS, Microsoft and Google with extremely large portfolios of database offerings. The spectrum of products in this space is very wide, with you've got AWS, which I think we're up to like 16 database offerings, all the way to Oracle, which has like one database to do everything not withstanding MySQL because it owns MySQL got that through the Sun Acquisition. And it recently, it made some innovations there around the heat wave announcement. But essentially Oracle is investing to make its database, Oracle database run any workload. While AWS takes the approach of the right tool for the right job and really focuses on the primitives for each database. A lot of ways to skin a cat in this enormous and strategic market. So let's take a look at the spending data for the names that make it into the ETR survey. Not everybody we just mentioned will be represented because they may not have quite the market presence of the ends in the survey, but ETR that capture a pretty nice mix of players. So this chart here, it's one of the favorite views that we like to share quite often. It shows the database players across the 1500 respondents in the ETR survey this past quarter and it measures their net score. That's spending momentum and is shown on the vertical axis and market share, which is the pervasiveness in the data set is on the horizontal axis. The Snowflake is notable because it's been hovering around 80% net score since the survey started picking them up. Anything above 40%, that red line there, is considered by us to be elevated. Microsoft and AWS, they also stand out because they have both market presence and they have spending velocity with their platforms. Oracle is very large but it doesn't have the spending momentum in the survey because nearly 30% of Oracle installations are spending less, whereas only 22% are spending more. Now as a caution, this survey doesn't measure dollar spent and Oracle will be skewed toward the big customers with big budgets. So you got to consider that caveat when evaluating this data. IBM is in a similar position although its market share is not keeping up with Oracle's. Google, they've got great tech especially with BigQuery and it has elevated momentum. So not a bad spot to be in although I'm sure it would like to be closer to AWS and Microsoft on the horizontal axis, so it's got some work to do there. And some of the others we mentioned earlier, like MemSQL, Couchbase. As shown MemSQL here, they're now SingleStore. Couchbase, Reddis, Mongo, MariaDB, all very solid scores on the vertical axis. Cloudera just announced that it was selling to private equity and that will hopefully give it some time to invest in this platform and get off the quarterly shot clock. MapR was acquired by HPE and it's part of HPE's Ezmeral platform, their data platform which doesn't yet have the market presence in the survey. Now, something that is interesting in looking at in Snowflakes earnings last quarter, is this laser focused on large customers. This is a hallmark of Frank Slootman and Mike Scarpelli who I know they don't have a playbook but they certainly know how to go whale hunting. So this chart isolates the data that we just showed you to the global 1000. Note that both AWS and Snowflake go up higher on the X-axis meaning large customers are spending at a faster rate for these two companies. The previous chart had an end of 161 for Snowflake, and a 77% net score. This chart shows the global 1000, in the end there for Snowflake is 48 accounts and the net score jumps to 85%. We're not going to show it here but when you isolate the ETR data, nice you can just cut it, when you isolate it on the fortune 1000, the end for Snowflake goes to 59 accounts in the data set and Snowflake jumps another 100 basis points in net score. When you cut the data by the fortune 500, the Snowflake N goes to 40 accounts and the net score jumps another 200 basis points to 88%. And when you isolate on the fortune 100 accounts is only 18 there but it's still 18, their net score jumps to 89%, almost 90%. So it's very strong confirmation that there's a proportional relationship between larger accounts and spending momentum in the ETR data set. So Snowflakes large account strategy appears to be working. And because we think Snowflake is sticky, this probably is a good sign for the future. Now we've been talking about net score, it's a key measure in the ETR data set, so we'd like to just quickly remind you what that is and use Snowflake as an example. This wheel chart shows the components of net score, that lime green is new adoptions. 29% of the customers in the ETR dataset that are new to Snowflake. That's pretty impressive. 50% of the customers are spending more, that's the forest green, 20% are flat, that's the gray, and only 1%, the pink, are spending less. And 0% zero or replacing Snowflake, no defections. What you do here to get net scores, you subtract the red from the green and you get a net score of 78%. Which is pretty sick and has been sick as in good sick and has been steady for many, many quarters. So that's how the net score methodology works. And remember, it typically takes Snowflake customers many months like six to nine months to start consuming it's services at the contracted rate. So those 29% new adoptions, they're not going to kick into high gear until next year, so that bodes well for future revenue. Now, it's worth taking a quick snapshot at Snowflakes most recent quarter, there's plenty of stuff out there that you can you can google and get a summary but let's just do a quick rundown. The company's product revenue run rate is now at 856 million they'll surpass $1 billion on a run rate basis this year. The growth is off the charts very high net revenue retention. We've explained that before with Snowflakes consumption pricing model, they have to account for retention differently than what a SaaS company. Snowflake added 27 net new $1 million accounts in the quarter and claims to have more than a hundred now. It also is just getting its act together overseas. Slootman says he's personally going to spend more time in Europe, given his belief, that the market is huge and they can disrupt it and of course he's from the continent. He was born there and lived there and gross margins expanded, do in a large part to renegotiation of its Cloud costs. Welcome back to that in a moment. Snowflake it's also moving from a product led growth company to one that's more focused on core industries. Interestingly media and entertainment is one of the largest along with financial services and it's several others. To me, this is really interesting because Disney's example that Snowflake often puts in front of its customers as a reference. And it seems to me to be a perfect example of using data and analytics to both target customers and also build so-called data products through data sharing. Snowflake has to grow its ecosystem to live up to its lofty expectations and indications are that large SIS are leaning in big time. Deloitte cross the $100 million in deal flow in the quarter. And the balance sheet's looking good. Thank you very much with $5 billion in cash. The snarks are going to focus on the losses, but this is all about growth. This is a growth story. It's about customer acquisition, it's about adoption, it's about loyalty and it's about lifetime value. Now, as I said at the IPO, and I always say this to young people, don't buy a stock at the IPO. There's probably almost always going to be better buying opportunities ahead. I'm not always right about that, but I often am. Here's a chart of Snowflake's performance since IPO. And I have to say, it's held up pretty well. It's trading above its first day close and as predicted there were better opportunities than day one but if you have to make a call from here. I mean, don't take my stock advice, do your research. Snowflake they're priced to perfection. So any disappointment is going to be met with selling. You saw that the day after they beat their earnings last quarter because their guidance in revenue growth,. Wasn't in the triple digits, it sort of moderated down to the 80% range. And they pointed, they pointed to a new storage compression feature that will lower customer costs and consequently, it's going to lower their revenue. I swear, I think that that before earnings calls, Scarpelli sits back he's okay, what kind of creative way can I introduce the dampen enthusiasm for the guidance. Now I'm not saying lower storage costs will translate into lower revenue for a period of time. But look at dropping storage prices, customers are always going to buy more, that's the way the storage market works. And stuff like did allude to that in all fairness. Let me introduce something that people in Silicon Valley are talking about, and that is the Cloud paradox for SaaS companies. And what is that? I was a clubhouse room with Martin Casado of Andreessen when I first heard about this. He wrote an article with Sarah Wang, calling it to question the merits of SaaS companies sticking with Cloud at scale. Now the basic premise is that for startups in early stages of growth, the Cloud is a no brainer for SaaS companies, but at scale, the cost of Cloud, the Cloud bill approaches 50% of the cost of revenue, it becomes an albatross that stifles operating leverage. Their conclusion ended up saying that as much as perhaps as much as the back of the napkin, they admitted that, but perhaps as much as 1/2 a trillion dollars in market cap is being vacuumed away by the hyperscalers that could go to the SaaS providers as cost savings from repatriation. And that Cloud repatriation is an inevitable path for large SaaS companies at scale. I was particularly interested in this as I had recently put on a post on the Cloud repatriation myth. I think in this instance, there's some merit to their conclusions. But I don't think it necessarily bleeds into traditional enterprise settings. But for SaaS companies, maybe service now has it right running their own data centers or maybe a hybrid approach to hedge bets and save money down the road is prudent. What caught my attention in reading through some of the Snowflake docs, like the S-1 in its most recent 10-K were comments regarding long-term purchase commitments and non-cancelable contracts with Cloud companies. And the companies S-1, for example, there was disclosure of $247 million in purchase commitments over a five plus year period. And the company's latest 10-K report, that same line item jumped to 1.8 billion. Now Snowflake is clearly managing these costs as it alluded to when its earnings call. But one has to wonder, at some point, will Snowflake follow the example of say Dropbox which Andreessen used in his blog and start managing its own IT? Or will it stick with the Cloud and negotiate hard? Snowflake certainly has the leverage. It has to be one of Amazon's best partners and customers even though it competes aggressively with Redshift but on the earnings call, CFO Scarpelli said, that Snowflake was working on a new chip technology to dramatically increase performance. What the heck does that mean? Is this Snowflake is not becoming a hardware company? So I going to have to dig into that a little bit and find out what that it means. I'm guessing, it means that it's taking advantage of ARM-based processes like graviton, which many ISVs ar allowing their software to run on that lower cost platform. Or maybe there's some deep dark in the weeds secret going on inside Snowflake, but I doubt it. We're going to leave all that for there for now and keep following this trend. So it's clear just in summary that Snowflake they're the pace setter in this new exciting world of data but there's plenty of room for others. And they still have a lot to prove. For instance, one customer in ETR, CTO round table express skepticism that Snowflake will live up to its hype because its success is going to lead to more competition from well-established established players. This is a common theme you hear it all the time. It's pretty easy to reach that conclusion. But my guess is this the exact type of narrative that fuels Slootman and sucked him back into this game of Thrones. That's it for now, everybody. Remember, these episodes they're all available as podcasts, wherever you listen. All you got to do is search braking analysis podcast and please subscribe to series. Check out ETR his website at etr.plus. We also publish a full report every week on wikinbon.com and siliconangle.com. You can get in touch with me, Email is David.vellante@siliconangle.com. You can DM me at DVelante on Twitter or comment on our LinkedIn posts. This is Dave Vellante for theCUBE Insights powered by ETR. Have a great week everybody, be well and we'll see you next time. (upbeat music)

>>Okay. We're back live here. The velocity conference is Santa Clara live. This is the cube Silicon angle's flagship program. We go out to the events, restrict the signal from the noise. I'm John furry, the founder of Silicon angle. And our next guest is CEO of guarantee a data, his here in the cube. Welcome to the cube. Thank you. It's great to be here. We are at the velocity conference, which is really the intersection of infrastructure and application development, kind of in a holistic way, full stack new technologies. Um, so first tell us a little about your company and what you guys are doing here at velocity. >>Well, we deal with a new type of database, which took the developers community by storm. This is a no sequel in memory database called Radis where this is very, very fast. You know, it, it processes hundreds of thousand transactions per second at sub milliseconds. And this is all about performance. So velocity is the right right place to be when you deal with Radis. >>So why, why red is, first of all, is taking everyone by storm. We use it, um, great technology. Um, why, why, why is it so popular? >>Well is, has many attractive datatypes and commands, which are very useful in many, many use cases today for almost any application. So that's why, you know, developers really love it, >>The in-memory database. So we cover a lot of storage, SSDs and infrastructure. Um, SSDs had brought up, uh, with flash, a whole nother level of caching on the level for storage area networks really exploded open source scale-out. Um, but people still need the real fast, low latency data, no doubt. And that's where in memory, but developers don't need to be storage gurus to do that. So is that an area that you guys are? >>Yes, definitely. The basic idea is to provide developers what they need in terms of database needs, without all the hassle of, you know, operating those databases. So with our products, which with our product, which is called the Radice cloud right now, this product is provided as a fully managed hosting service over various clouds and platforms as a service. So with this product, the user does not need to do anything, simply send your data and forget about it. We take care of scalability, high availability, stabilizing performance, and all the ops. >>So one of the things about the web that's really challenging it's asynchronous, right? So persistence is a really big thing. How do you guys look at that channel? >>Okay. We have built a whole suite of high availability provisions for Radis. First of all, you can with a click of a button with a checkbox, you can replicate your data set within the same data center, uh, and when a node fails, and this is something which happens in the cloud almost everyday, we immediately, uh, switch your data to the, to the replica and, uh, you are up and running without any, any problem whatsoever. So this is one thing we recently last week, we announced another layer, which is multi a Z replication, which means that you can with a click of a button, replicate your data set to another data center. So if the entire data center fails, we immediately use the replica in the other data center, the backup replica. And again, you're up and running without any interruption. >>This really is a value proposition. That's as a dream scenario for developers with dealing with the cloud. I mean, because your alternative is to provision bare metal, exact load Linux systems >>Administrator. This is crazy. I mean, >>Oh, and cuing too, is another another issue. I mean, how do you know? So if I'm going to manage large volumes of data set to say that, um, my side becomes popular, my application becomes popular because, uh, someone shitted, virally, I want to have that queuing and that persistence that's really, really important. I might not have the time to provision a new server, a new database. So what you're saying is if I get this right with Reddis cloud, I can spin up in dynamically handle that those kinds of replication and persistence >>Over, you know, basic red. Is that right? Absolutely. Absolutely. You know, our native red is the open source is basically, uh, limited in scalability. You cannot grow beyond the single master server. Now the community is working for a while and something which is called Reddis cluster, which is supposed to solve all that. However, this is taking for a very long time with our Reddis cloud, you can grow your dataset from megabytes to Jigga bytes, to terabytes and even more, and all that is done in a fully automated manner without you do not need to deal with nodes, clustering, scaling, stuff like that. And while supporting all the data types and commands of Radis, which is really, really unique. >>Yeah. I mean, I got to say, you know, one of the challenges with the cloud is orchestration, right? And so that's one element. So automation has been a big problem for folks on premise on large enterprises and application developers. The other challenge has been real time. So a lot of apps need to have real time, like no JS or things of that nature. So how does a developer, I'm a developer and I'm, I want real time. I want persistence. And I want to have the flexibility to, to, to just push code and everything take care of itself. How does Reddis help me there? >>Well, red is, as I said, is the fastest data store available today, much faster than anything else. Like, you know, people talk to them about HANA SAP HANA, uh, red is, is, uh, 10 X, you know, in terms of speed, we are talking about hundreds of thousands transactions at sub-millisecond latencies. Whenever you want performance, whenever you need performance, the best database for that is rarely snow. >>Okay. So I got to ask you the question, first of all, big fan, really glad you're here in the cube. So we like, we like what you're doing, um, for the folks that don't understand what you guys are doing or are red or new to Retis. Why is it so good? Why is it so popular and what, what benefits does it provide the developer and say a business that wants to use that? >>I would say use cases, use cases, use cases whenever, whenever you, whenever you need a job management, for example, you know, signaling inside your, your, your application. So platforms such as sidekick, sidekicks, you know, et cetera, use Radice whenever you need, uh, stuff like, uh, Twitter type functionality, you know, followers, et cetera. You have a built in clone within radius for that whenever you need, uh, you know, uh, fast analytics, there is nothing better than red is caching, you know, already since replacing Memcached totally today, new apps, uh, page ranks, post ranks, you know, stuff like that. All these are great use cases for remedies. And if you, you know, in any one of those various is the best for that. Yeah. >>Well, congratulations, really like what you guys are doing. Um, and you're at the show here. What are you showing here at velocity? Again? Congratulations on your success. Well-deserved reticence is really becoming the standard. What, what are you guys doing here at velocity and what are you guys showing? >>We demonstrate, uh, first of all, the service we demonstrate the performance. You can, you know, if you have a minute drop over to our booth next door here, and we show the great performance, you know, we are showing hundred thousands of transactions, you know, with large databases in sub-millisecond latencies. This is, you know, this is real life and we are demonstrating our high availability with multi a Z replication and instant out of fail over. >>Okay, well, we are here with Ofer B gal with the system guarantee, a system data, um, Reddis cloud, great product, congratulations on your success. Thanks for coming inside the cube. This is the velocity conference. This is the kind of technology folks that velocity is about the loss of these, the intersection between a, almost a systems view of user experience, user design with cloud and infrastructure or dev ops, whatever you want to call it, we'll figure out a word for it, but it's really kind of coming together. I guess we call it velocity conference. This is the modern infrastructure that a lot of the web-scale companies or hyperscale companies are using and developers, developers who are small-scale today. We'll be, we'll be big scale. We'll use things like redness. This is what it's all about. This is the Silicon ankles flagship program. Go to youtube.com/looking angled to watch the videos go to siliconangle.com to get, to see the blog posts and coverage. We'll be right back with our next guest. After the short break.