 Good morning, welcome back to theCUBE's day three coverage of Snowflake Summit, live from Caesars Forum in Las Vegas. Lisa Martin here with Dave Vellante, as I mentioned. This is our third day of covering Snowflake, the whole ecosystem of Snowflake, customers, partners, executives, leaders. We're going to be talking with Snowflake and Snowplownex. They're providing together a competitive solution for modeling and analyzing conversion journeys in near real time. Dave, that's something that we as consumers expect to have a seamless conversion journey, but it's hard for organizations to achieve that. Yeah, a lot of snow going on here. Snowplow, Snowflake, a lot of Snowteams, bring it on. Ironically, it's like 100 degrees outside, but we're talking all snow. We have two guests here with us, Mike Robbins, the head of engineering at Snowplow is here and Patrick Crosby, senior manager of technology alliance at that Snowflake. Great to have you guys. Thanks for having us. Great to be here, thank you. The one funny thing is, this whole thing is on brand, this whole expo hall is on brand very, very much the whole snow theme and it's ironically freezing in here, so we don't know it's 100 degrees outside. Mike, talk to us a little bit about Snowplow. I see 1.9 million sites and apps use Snowplow to generate and model first party customer data from across their digital interfaces, capture descriptive customer journeys. Talk to us about Snowplow, what you guys do. Yeah, so a lot of it's really around behavioral data collection, as you mentioned. So looking at users and customers on websites and apps and server side and IoT and all that sort of good stuff and really trying to get a better understanding of what consumer behavior is effectively happening and increasingly now a lot of importance and focus on doing that in a first party way. So a couple of years ago, it didn't really matter as much. You could get away with a lot of third party stuff but particularly now in this emergent market and sort of post all these privacy breaches, there's this increasing consideration of how to do that in your own environment and capture that in a first party way rather than rely on some external vendor. So we've seen a really big uptake of lots of people considering that more and more and as a result we've seen a lot of uptake on a lot of those sites. I'm sure you get this question all the time. I was like, why don't I just use Google Analytics and you kind of addressed it but maybe you could address it more directly. I think it's part of it which is there's some people that just can't use Google Analytics now, right? There's a lot of legal rulings particularly in the EU where they just go like, well, no, you can't go and do that because it's being processed in a way that fundamentally from a European perspective infringes on human rights basically. So some people just can't use it and the people that can are going, well, look, it's a good option but there's a change now. There's a big change between Google Analytics 3 and Google Analytics 4 and it's giving people the chance to re-evaluate and go, well, actually, is this what we want to take into the future for the next 10 years, 15 years? And that's a lot of how long GA's has been around. So people kind of that decision point now of going, like, is this going to be right for us in the future? And a lot of people are going, no, maybe not, maybe not. Maybe we want to consider sort of other different options. So there's compliance and there's functionality. Compliance and functionality I think is a big one and just transparency as well, right? A lot of people are advertising on Google or run ads through Google or buy ads through Google and they go, well, actually, I want to be able to measure the effectiveness of my campaigns. I want to be able to measure that without some player that kind of potentially may have a vested interest in favoring towards themselves over, say, Facebook or another advertising network. So a lot of transparency in going, well, if the data is in my environment, if it's sitting in my Snowflake account, if it's sitting in the data cloud there, I have the ability to kind of audit that and for a lot of people that's a really important thing. It's fundamentally about trust. Yeah, and that transparency one is really important and something that we at Snowflake have seen in a variety of different industries and use cases and particularly in this case, like the ability to be able to have first party data collection to actually as a brand to own all of that data about your customers and your own and operated infrastructure is very powerful and so we're working with a ton of amazing partners like SnowCloud and so they're able to bring all of that in and pull it into that customer Snowflake instance and then they're able to combine that and unify that with a variety of other pieces of information and so the transparency, the breaking down of data silos is really powerful. It's critical for organizations that it's no longer nice to have. It's really critical to drive revenue, to prevent churn, net customer retention rate and these are things that businesses across every industry have to do these days to stay competitive, to stay in business. Yeah, and it's customer expectations that you own that customer data, right? Like why are you going and sending it somewhere else? Why are you going and storing it in these other systems and we've all probably been subject to a lot of emails going, okay, you've had this account leak or this account leak but really I think organizations are starting to understand that sometimes using all those different things can be a liability and really you want to have control over that environment yourself. So we're seeing lots of that, we're seeing lots of increases in sort of data sharing with a lot of this Snowflake stuff as well. So rather than sending the data to some other kind of third party vendor, we're seeing more of these relationships directly between brands and companies and brands and vendors and customers which is opening up a lot more sort of collaboration and transparency behind how a lot of those partnerships work. Now you're doing this inside of Snowpark? We do a bit inside of Snowpark. So we do a lot of attribution inside of Snowpark. We'd love to move it over to container services. So that's something that we're looking at at the moment but a lot of the multi-touch attribution we do. So in terms of how people are converting on the platform it could be e-commerce, it could be a media site where you're logging in and paying through a paywall. What are the channels that get you to that conversion? What is the marketing team spending money on that is effective? And that's a really challenging thing to measure particularly in a Google style ecosystem where effectively you've got a black box and they'll tell you what's effective but really doing that in Snowflake and Snowpark means you get to see the innards of that working machine, right? It's not a black box that tells you this number and this number and doesn't explain it. And I think a lot of stuff is around that explainability component at the moment which is why is that channel effective or why should we spend more on this or less on this? And particularly in the economic climate that we're at at the moment is that's a really critical factor to explain the impact of those economic decisions on product and users and acquisition as well. And those features like Snowpark and streaming are really what's unlocking a lot of the value here. So being able to combine the power of Snowflake with Snowplow and drive those analytics and actually get some of that real-time perspective and then giving you the flexibility to do the modeling and the visualization, that's all sort of, that's what's really I think coming to fruition here with a lot of the new capabilities that we're releasing and that Snowplow has released and so it's really amazing to see this and actually give brands a really amazing opportunity as they're being forced to shift over to a new version of Google so they can really sort of step back and evaluate are there other options where I could actually get a more powerful solution? Time to switch. But so the journey started with using Snowflake at the back end, right? And I saw, I'm seeing Snowplow and then now you're increasingly investigating Snowpark container services and actually being native inside. Yeah, and I think it's kind of, we're tracking the evolution of the product, right? It really started as this data warehouse and now as emerging as all these other sort of services as part of that data cloud. So particularly for a lot of our customers is that data warehouse is the core of everything but then you go and add on the compute layer and then you can start to run your personalization or your recommendation engines and that sort of thing. And then increasingly we've got, as is the nature of things, people asking more about AI. So a lot of the LLM stuff and the Nvidia keynote and people sort of interested in all this new stuff they can do. So it's been a really massive problem with data historically as your data is sitting in one place and your compute and all the interesting things have sat in another and we're kind of increasingly finding this convergence of well, your data shouldn't be over here and your compute over here. It really makes sense for them both to be in the same place and I think a lot of the data cloud announcements are kind of continuing with the idea of well, everything is in the same place and there's this concept of sort of data gravity which is the more data you're adding into something it pulls in everything else and that's both from a data perspective but also an organizational perspective as well. And it unlocks a ton of flexibility as well and I think it's another dynamic when you're using a single monolithic solution, right? Like you are, you're waiting on innovation and sort of the pace and the choices and decisions that they have to make but as an organization is a brand if your customer data, it is a critical asset to many of these businesses. It is their differentiating asset is the reason that they're able to compete and grow and be successful in these increasingly competitive markets and so you want to own that asset and you want to have the flexibility to leverage that asset in a variety of ways and so this gives them that ability so they're able to unify that, they're able to leverage amazing solutions that are there but they're also able to do things in-house, to do modeling in-house to extend those capabilities in any way that they'd like and so that's another I think critical part of this. Mike, you must have a favorite customer story that you think really articulates the value prop of what Snowplow and Snowflake are doing together that helps maybe even like a, where you came in and replaced Google Analytics? Yeah, I mean we've done that a lot of times. I think one of the big customers for us was folks called Digital Virgo who run a mobile payments platform so effectively they've got something like two billion users worldwide and effectively it's paying for something on your mobile so you could be in a sports stadium buying a ticket or a drink or you could be in a train station getting a promotion or a discount or something like that but effectively pay on your phone which is a massive, massive thing it's kind of prevalent everywhere and one of the big issues they were on Google and one of the big issues they had was just this latency between when they were getting the data and then when they could actually make a decision on it and you think about it, so they were big into a lot of inventory and campaigns as part of FIFA but also the NBA finals recently as well is the performance of that campaign is happening in real time, right? Like people are on their phones it's not a you're buying a car and you go and buy it in three months time they wanna know how that spend is influencing and how users are interacting with that and then basically personalize effectively in real time so they really connect those telecommunications providers with the brands directly as a proxy but they couldn't do that with something like Google Analytics they just didn't have the quality of the data and the latency was too high so a lot of stuff in Snowflake where we can get that data in there in seconds and increasingly things like Snowpipe streaming they can then go and make those decisions go back to brands and go okay we're going to optimize this in real time look at this acquisition we're going to stop running this campaign we're going to switch campaigns we're going to increase the impressions so for them that was a really powerful thing because it's effectively an audience of two billion people, right? Is even a tiny optimized changes has a massive, massive impact. Yeah and I think what's really powerful about the partnership that we have here is that on the front end of that like we want to make sure that customers are having all that critical data within Snowflake but the collection of it is a really challenging problem to be able to actually know how to get the data from the point of activation if it's your phone, if it's a site and then being able to do the validation and all of the important some cases you actually need to retry those events so there's a lot of actually nuance and complexity to achieving the landing of that data accurately and completely in Snowflake and Snowplow is a really amazing solution on that side and that's what's able to give customers the granularity and the accuracy that they need to have these sorts of campaigns. Explain Snowpipe streaming, what does that do for you? So basically, so historically you've had to kind of batch load data into Snowflake which basically means you're not going to insert one event at a time or one user at a time you're going to group them all together and you're going to do it as one big batch which has historically been the fastest way to do it and generally you can do that every couple of minutes which is pretty good for a lot of folks is generally they're not reacting much faster than that but Snowpipe streaming gives us this new capability of essentially having a fire hose where we can put in as much or as little as we want and it goes into Snowflake effectively immediately so within a couple of seconds. So particularly for real-time use cases where you want to look at what's happening now compared to six months ago or three months ago historically that could have been quite difficult to do because you would get it at minute intervals but you wouldn't be able to sort of predict beyond that whereas this sort of second to sub-second latency really unlocks a lot of those real-time use cases that previously would have needed a lot of engineering effort to build whereas now it all runs basically effectively within the same cloud. So you're essentially working on live data. You're working on very close to live data. Very close to live data and then when there's a change do you use some kind of like change data capture or something similar to that? Yeah, so a lot of people doing CDC with a lot of event stuff it's kind of effectively a change in state or an action for a lot of customers and that just opens the door to very quick personalization which is if someone is abandoning a heart, sending them an email 30 minutes later an hour later they've forgotten they're watching TV they're not going to tune back in whereas if you can do that in far, far closer to real-time and that's kind of what people expect is you don't want to go and open up your Netflix or whatever your video player is and see recommendations for stuff that really isn't relevant to you because actually you only cared about that thing at a point in time. You've just watched something, here's something that's going to be far more relevant to you. So, and that's just kind of becoming what everyone is expecting and if you don't get that experience you're kind of a little bit befuddled as to like, okay, why, why? It doesn't really make a huge amount of sense. Why don't we just add dollars? Yeah, yeah. Next best action is to your point, we all expect it. We expect it to be personalized and relevant like we were talking yesterday. You know, if you go online and you buy a tent, you don't want to, or your analogy with dishwasher. You don't want to be served more ads after that transaction is complete for more dishwashers. You want, what would go with that? Or if I want to buy a tent, if I bought a tent, I don't want to be having more ads for tents. I want things that would go with it. We just have that expectation that wherever we're doing a transaction, that next best action is going to be relevant. Absolutely. And we've just seen an explosion of customers and partners that are actually using Snowflake to drive these types of marketing and advertising outcomes and a lot of them are built on these types of real time and near real time use cases. So you really need to incorporate that. And I think Snowflake Streaming is unlocking that for a lot of our partners and it's allowing them to really drive those outcomes. You got the user experience, which you're describing as it's annoying. But the other end of it is you're just wasting money as the advertiser, throwing it away to an individual that has no longer has intent to buy. Yeah, absolutely zero intent. It is zero intent. Yeah, I've been told that Amazon does it because a lot of people will then go and return those things and actually it works quite well. So they'll buy a vacuum cleaner and then maybe it won't work out. So they'll still recommend vacuum cleaners, but it feels like a diminishing return sort of thing. Which is like, it doesn't make sense, like particularly dishwasher, like, do I really need another? But I mean, maybe they know best. It's like selling crappy products and they know it's going to break. I think it got back to our earlier point. But you experienced that actually with these days. You say, do I want to buy that on Amazon? Do I trust that from Amazon or is that? Yeah, but I think I get back to our earlier point about transparency and having the data in Snowflake and having it explainable. Which is like, if you are going and serving, you bought a dishwasher and you're serving inventory of dishwasher to that user again, what is the data? What is the behavior that you're capturing on those users? And how do you go and explain that sort of thing? And in a lot of these black box systems, you don't get that explainability. You're trusting the algorithm. So a lot of this stuff on data cloud is you're building it yourself. It's within your environment. You can explain it to your stakeholders, but you can also explain it to your customers as well. Or you could actually do the analytics to design that campaign such that you could ship the initial vacuum cleaner and then have the ads for the next one that you're going to have to ship after the return ready. So what does AI do for all this? Because now we're entering this new era. We're in the before, we're kind of in the middle. What's the after going to look like? I think what this is doing is it's really building that data foundation that allows you to unlock those AI capabilities. So we've talked a lot about how we're getting that real time data landing directly in snowflake. We're getting that to be valid. We have transparency. You've got all of that data unified. So now you actually have the critical data asset that's needed to run those use cases. And you can unlock that in a variety of ways. Like you could leverage capabilities that you are building in-house. You may have a data team that's doing that. You can leverage some of the amazing solutions that are driving that and building directly on snowflake. So this is building the data foundation that is unified, it's secured, it's easy to collaborate, and it's easy to make that extensible across all of those use cases. So that's what's I think also really powerful about this partnership. Mike, the name, Snowplow Snowflake Coincidence? Yeah, it is. You guys are about 10 years old, right? Yeah, about 10 years old. 10 or 11 years old, I think. And you guys are probably similar because there was a couple of years in stealth mode. Yeah, I think if I remember correctly, Snowplow is six months or a year older. Okay. But I think there's similar origins of the name. You know, they are both related to the cloud and things like that. I think the partnership was meant to be. Yeah, it kind of writes itself. There's a little bit of confusion, but there's also a lot of complimentary stuff there. We're going, oh, that makes sense. Like it's a solution. Yeah, and you still every once in a while have to tell a snowflake employee that Snowplow is in fact not a feature, but another organization. Well, that first thought that I thought it was. And I'm like, oh, they're just, they're so good with the naming. So Mike, you talked about what you guys are doing in terms of helping customers with campaign performance, conversion, next best action. What does the future hold for Snowplow, especially in terms of your Snowflake partnership? Look, I think taking more advantage of a lot of these new services is, you know, the more and more it grows, you know, we started very focused on the data warehouse because that's what it was early on. And we had a lot of early customers going, hey, look, we're looking to migrate off these legacy platforms to something new. Let's try this. And that's kind of where our partnership started. And I think now is looking at these new services at it. So now for a lot of customers we're getting asked about LLMs and, you know, what can you do with GPUs in Snowflake and how do we do ML and AI? So a lot of it's these more advanced use cases, a lot of it's infrastructure as well. So now, you know, with containers, how do we go and move these workloads that we have potentially in other areas closer to the data? And that's both from a security perspective but also a cost perspective as well, right? You're not shipping all this compute and data around between completely different regions. You're really being able to centralize a lot of that. So I think a lot of it for us is figuring out what are those services as sort of complimentary and how can we kind of enhance the Snowflake experience from a user perspective and a customer perspective, make it cheaper, make it easier to run, make it more holistic. So take advantage of those services and migrate some of those things into sort of what is increasingly become this larger platform. Yeah, absolutely. Guys, thank you so much for joining Dave and me on the program talking about Snowplow on Snowflake, what you're able to enable customers to do. I had to really think about that to make sure I got it right. We appreciate your time and we wish you continued success. Thank you. Thank you so much for having us. Our pleasure. For our guest and for Dave Vellante, I'm Lisa Martin. Up next, do you want to be a query boss for your company? Well, we're going to be talking to the CEO and co-founder of Sundack who's going to tell your data engineers how to achieve just that. Miss any of our content? Thecube.net is where you go, saleclenangle.com for analysis and editorial. You're watching theCUBE, the leader in live tech coverage.