The Problem-Solving Power of the Cloud

Fast forward to the 2020s and I can’t remember the last time hardware was even mentioned in a conversation or contract. Rather, it’s now assumed that all the data is going to live in the cloud where there’s an unlimited elasticity with data storage and computation. The upfront data storage cost and investment are way less. Data security, connectivity, interoperability are way better and easier. And new cloud native capabilities and platforms like private brand identity graphs and CDPs and clean rooms are being leveraged to help brands adapt to some of these massive disruptions happening in MarTech today. Kyle, what are your thoughts?

Kyle Hollaway:

Well, first, thanks for carbon dating me for our listeners.

Dustin Raney:

Sorry about that.

Kyle Hollaway:

Yes. No, that’s all right. Yes. No. 30 years, around there in the industry. And certainly a lot of change, just like you talked about. We’re not just talking about hardware. And there’s not just one cloud provider that’s now taking place of that and being a key player. There’s really this epic battle going on for cloud dominance. It’s taking place between Amazon AWS, Google’s GCP, Microsoft Azure, so on, so forth, which has presented a challenge of its own. For instance, if my enterprise is using multiple clouds, what do I do? Or what if I want to share my data that’s in AWS with a brand that’s using GCP or Azure, right?

So, this cross-cloud conversation. And don’t forget the brands are still looking for a scalable solution for onboarding and activation when cookies go away. So, if an advertiser and publisher has data that’s going to go live in different clouds, how will the matching and the identity part work? So, all of those are emerging questions that not only we and our clients are wrestling with, but really the entire industry. And so with that in mind, this is a perfect opportunity to introduce our guest today to the podcast. It’s Rob Saenz, senior sales engineer at Snowflake. So, Rob, welcome to the show.

Rob Saenz:

Thanks. Thanks for having me.

Kyle Hollaway:

Yeah. So, hey, why don’t you start off by giving our listeners just a snapshot of your background and the role that you have today at Snowflake?

Rob Saenz:

Okay. Yeah. So, again, my name’s Rob Saenz. I, of course, began my career at AT&T being an Acxiom client, like I’m sure every hopefully great person does. They start out as an Acxiom client.

Dustin Raney:

There you go.

Rob Saenz:

So, I grew up in data warehousing and did some web application development. And then I found my way into analytics, which led me into Snowflake. And so I’m a senior sales engineer, like you said. I think I have one of the greatest jobs in the world because I get to work with world-class organizations and help them imagine what the future can look like. Because Snowflake has really developed this candy that didn’t used to exist, just like you alluded to. So, it’s fun to say, “Hey, what can we put together with this?” And that was actually my first conversation with Dustin was like, “Hey, what could we build? What could we put together with the stuff that Snowflake is producing and coming out with?”

Kyle Hollaway:

Why don’t you just take a second for our listeners to just define Snowflake? You guys have made a really good market presence and it’s like you hear Snowflake everywhere. But just give a sense of the scope of what Snowflake is offering and then move into Snowflake and clean room, the word clean room and how that’s evolved and the correlation between the two.

Rob Saenz:

Sure, absolutely. So, when we show up at trade shows, it’s really funny, especially in the south where we’re all based. People walk up and they look at a sticker and they laugh at us. They’re like, “What is Snowflake?” and they turn sideways and look at us. So, in a sense, it’s very remarkable, but we like to call ourselves the data cloud. So, we solve a lot of the same issues that you mentioned, some of our predecessors like Oracle or Microsoft or their SQL server, or even some of the nascent cloud data providers like Azure or Amazon Web Services. We would say a first generation data provider, Snowflake is reimagining what would it look like if we all had the ability, not just to move the things that we’re doing, maybe building our first party graphs? But what if we were able to collaborate at a very granular level with all of our friends and compatriots and build this beautiful ecosystem together to accomplish all of our goals.

So, to talk about clean room, clean room’s the latest manifestation of that. Clean room with Snowflake, we hope gives our customers the ability… And we have several dozen, I guess I would say the number’s always changing, companies trying to take advantage of this. But it gives two companies the ability to, without moving their data, have a conversation about maybe how they could go to market together, maybe how they could, like you’re talking about activate or onboard without having to say, I’m going to ship you all my data, or you’re going to ship me all your connections. We can have this completely online data collaboration conversation where everybody retains the security and privacy, which is so important for all of us, of that data.

Dustin Raney:

So, you mentioned some of the onboarding stuff and how clean rooms are. That’s certainly a use case that’s going to try to help mitigate as cookies go away. What are some other common use cases that you’re seeing brands leverage? Or why would they want to use Snowflake other than their on-prem solutions? What else are we enabling in Snowflake?

Rob Saenz:

Yeah. And I know we’re going to get to walled gardens later, but there is a vision of the future that has brands having more and more flexibility to build their own first party graph and take control of who they market to, how they market to them, how accurate they can be, like you guys do and talk about all the time with measurement and attribution and building these campaigns.

So, a lot of our clients, especially in the advertising, media, entertainment vertical are really trying to make the most of their marketing dollar to figure out which platforms which they can advertise on using the data that they already have. And then perhaps they’re using someone in the middle to resolve that identity, maybe even enhance that identity. I know that companies like Acxiom do all of that. And so we want to produce the pipes and a lot of customers are taking advantage of the pipes to just say, “You know what? Let me connect you to somebody who does that.” I think what Snowflake customers want to do is not reinvent the wheel. Stick with their core competency, offer the best products that they can and collaborate with people who have already gone before them who can already solve a lot of the problems that are already… They’ve been out for years and solved for years.

Kyle Hollaway:

A, Snowflake while you’re founded in 2012. So, in terms of technology companies, you’re on the newer side. What are some of the trends and changes in the industry that you’ve seen are really opening the door for Snowflake with clients, or actually having clients come to you and saying, “Hey, we want to look at this. We want to move into that.” What are some of the key drivers there?

Rob Saenz:

I heard Jim Collins, business author, talk one time about how what companies want to do is they want to fire bullets and then cannon balls. They want to take a shot at something, figure out what does that experiment look like? That pilot look like? And then they bring the heavy artillery and they march down that path. And I think Snowflake is part of that generation of technology companies that says, we want to make it easy for you to try this out. Maybe you bring in a single audience. Maybe you go and you run it through this data clean room. Maybe you have the ability to onboard. And there’s a small investment for that. In a similar way, even when Snowflake customers come to us for the first time and build a capacity contract, they don’t have to decide, to Dustin’s point earlier, how much hardware am I going to use? How much data do I have that I need to store? How long am I going to persist it? Set all those retention policies?

What they can do is just lay down a credit card and get started testing some of their hypotheses, loading some of their data, collaborating with some of the people that they know are already here. And then eventually create data products that other people can come to Snowflake and test out that data product. So, we like to create this virtuous cycle. And I think that’s really what the new industry trend is. I want to go solve problems, but I don’t want to have to make a giant capital or operational expense and scope it out from end to end and make a seven year plan. They’re thinking in much shorter terms, this is something that we could attack in the next three to six months. And so they’ll come to a company that has that agility.

Dustin Raney:

I know you’re working with both advertisers and publishers, because you have to build that connectivity across both. What kind of adoption are you seeing, especially as advertisers? What size of publishers are you seeing adopt Snowflake or actually be willing to put their data, their audiences there to be matched against, to buy inventory?

Rob Saenz:

Yeah, no, that’s a great question. When I joined the company, they had us all read this book, “The Rise of the Data Cloud.” And what our CEO was writing in that book was the history of how Snowflake was founded. And I was very shocked because I’ve worked with other technology companies before. I was very shocked to see that one of the first groups of people stepping forward to test out Snowflake were some of these big publishers that you’re talking about. And that has been a big chunk of how we are going to market the people that we’re trying to serve. We have a whole vertical that we’re always hiring more and more people to serve because some of the biggest ones that have these digital reach and even offline reach are using Snowflake to connect to potential advertisers as well as connect to people who can enhance their graph, help them sell their inventory.

And I think for marketers, there is a sense, especially as you move down market, that they can come to Snowflake to do some of the basic audience building or plugging into an identity platform. And they’re looking for different things. Maybe they’ve already got a platform that they want to publish with, but they want to use a company to help deliver a SaaS CDP for them to manage what they’re doing or to plug into their CRM, to enhance what they’re doing. Typically, people are using Snowflake on the marketing side to just help them assemble and build that graph often without having to persist it. If I can connect to someone, they can tell me what I want. I can just slap those attributes onto what I already know. Or you guys have offered, if I can do some hygiene, what I have. And then that becomes a viable platform for me to take it to a publisher or to an onboarder. Or perhaps they want to do the whole experience. And I know that we’ve been collaborating on a product that would help them go a little bit from end to end.

Kyle Hollaway:

There’s an interesting challenge here, where as an industry, or even just broader than that, I guess just as a whole economy, the sense of privacy and constraints around data and cloud. And so this pertains to all cloud providers and with Snowflake being the data cloud, it’s in one way, it’s a bit of oxymoron because it’s like, oh, we’ve got an issue. And then now there’s this data cloud, which means I’m going to have data in it and stuff. How are you guys approaching that? And how do you see Snowflake as being a better option as it relates to consumer privacy and the overall trend within the industry of managing data better?

Rob Saenz:

Yeah, that’s a great question too. The data cloud messaging, when we first heard it as an account management team, we scratched our head too. What does that mean? But I think a lot of what that means is rather than operating at an infrastructure level, I’m going to spin up a certain amount of architecture, compute resources, storage in order to address a problem. What Snowflake is proposing is let us worry about all that stuff. You are just going to load your data. And what’s different about Snowflake than other platforms that are maybe even on the native cloud providers. So, you talked about, you have folks that they’re committed to an AWS. Other folks are committed to an Azure. Snowflake is an interconnected software as a service running on all three platforms. So, I’ve seen clients say, “I’ve got data on this platform and it lives in my Snowflake. How do I work with someone who’s on a different platform?”

And so that was one of the things that we solved pretty early on was if you have data that you want to live on both platforms, we facilitate that without you having to leave it from your account. So, to your questions about privacy and how do we approach that differently? Snowflake architecture is unique in what I’ve seen in 20 years of working with databases and data platforms in the sense that it’s purpose built for the cloud. So, if you look at the most atomic measure of a data file or something like that, we store everything that’s ingested into Snowflake and encrypt it and proprietary colonized for instant access. Even though that it lives inside of one tenant, one Snowflake tenant, it is protected by measure after measure of architectural controls to say, you know what, today maybe Kyle can access this data, but he’s going to only be able to do it through Snowflake.

And even if you were able to snip some of those packets coming along the way, chances are any keys and credentials and all the stuff that is quite honestly, even over my head has changed because from the clean sheet of paper innovation, that Snowflake was, they said, no one’s going to trust their data to be in the cloud if it’s not secure. And so we have made it so that almost no one can see your data, not even you, unless you go through a very specific process through this service. And that’s why Snowflake is certified for PCI, for HIPAA compliance, for even Fed rant moderate. Understanding completely that this all sounds great and good and it produces the ability to go as fast as you want and test out as many experiments. But if it’s not secure, man, it’s not going to fly and it’s never going to gain traction.

Dustin Raney:

And if you extend that conversation to even the global scale, a lot of brands they’re collecting data and have their data located in different regions around the world and every region, it’s not just that you have multiple clouds, you have multiple regulations per region where the data’s stored in those clouds. So, I would think that what you all are doing at Snowflake is actually helping those brands manage all that, even the geolocation, right?

Rob Saenz:

Yeah. So, you mentioned it, like GDPR is really huge. And we’re very fortunate to have a lot of smart people, our founders at the top of that list that understood not only the problems with these legacy data platforms, but they were solution architects that had to deal with really complicated requirements from other customers. So, they knew, “Hey, we’re going to have to build a lot of traceability, a lot of audit ability into this platform so that we can have our customers be able to with a very straight face tell, look, I know exactly where your data’s been. You can see exactly where it’s been. You’ve got the ability to see it and trace it and set alerts if something goes sideways that you don’t expect.” A lot of those basic use cases that were not a part of some of the earlier innovations of data platforms were absolutely at the front of the mind of the people that were founding this company.

Dustin Raney:

You alluded to wall gardens, let’s go there. So, Facebook, Google, others. Do you ever see them…Because honestly, what I see is you guys building this marketplace where all the fish are, as many fish as possible. And brands want to go make sure that they’re interoperable and they can connect their data to as many advertisers and publishers as possible. So, Facebook and Google, two big ones. Do you ever see them playing well with Snowflake?

Rob Saenz:

That’s a great question. What I can say is that their customers do, and their customers have to. So you guys are always preaching the value of taking control of your data. We talk about cookies that you don’t eat. Those guys eventually going away. Lots of theory, I’m sure you guys have had about whether those things should have even played as big of a role as they did. But here we are and now I liken it to music publishers got to sell cassettes for a low price for a while. And then they got to sell ridiculously high price CDs for a brief moment in time. And then all of a sudden you buy tracks by the single 99 cent purchase. And it’s super easy, but the whole game has changed. And a lot of these music publishers are like, “Wait a minute. I want to sell CDs again, because that was way more lucrative.”

So, as the market changes, you just have to move with it and be flexible. And I think our platform was when we knew that we could deliver the ability to take all those data warehouse, data lake workloads. And then we could also deliver the workload of enabling people to share their data through very sophisticated governance, but effectively not having to ship data and rehydrate it, we realized we had the underpinnings of a more sophisticated sharing mechanism that could include even include monetization. But Snowflake is agnostic in the sense that we want as many people as possible to take the first party data that they want to activate or onboard or enrich in a very secure way, share it with people who are willing to provide those data services to enrich it or activate it, or onboard it. And so we want you to work together and need each other.

One of the things that I usually mention is that one of our chief technology VPs came from YouTube and he sees the data cloud as we provide the platform and you guys provide, as our customers, you provide the rich content that makes the platform more valuable. So, if Apple and Google are the experience that marketers want, which is like, “Hey, I trust you. I trust you to run the right kinds of campaigns and to do the right attribution and to deliver me the ROI.” Man, go at it, go for it. But if you want the ability to have a little bit more control, work with who you want to work with, activate onboard the way that you want to and measure how that’s going the way that you want to with a sophisticated staff and sophisticated partners, that’s the door that we’re trying to open for people to use.

Kyle Hollaway:

Yeah, it’s really interesting. And it is going to be interesting to see the evolution of that even over the next, probably 12 to 24 months, as you see more of the retail media networks. Take Disney and Disney coming on board to Snowflake and offering it as an option for that audience. So, it’ll be interesting to see as there are more of those retail media networks that have a very significant reach that can begin to compete with, say, the Meta’s of the world. As we’ve seen recently in Meta’s most recent announcement, they actually had some slippage in their market share and in the advertising. Now we’ve seen obviously some acceleration on the Apple side, but their new capabilities and the emergence of probably a DSP soon of their own. But at some point, it is a transaction. There is a point where there’s an audience, that’s at the first party on the brand side that they’re saying I want to reach.

And then they’re going to have optionality across these various, the big walled gardens and then more of these retail media networks and then other intermediaries that they can share with. So, I think it’s going to be an interesting area to watch, to see how with privacy-centric capabilities like Snowflake as a sharing mechanism, do some of the wall gardens become easier to play with because that’s where the audiences can reside and transact. And now you’re seeing, as the competition heats up, they may recognize that and be like, “Hey, we need to at least have some channel here that facilitates that.” So, that would be interesting.

Rob Saenz:

Yeah, absolutely. I remember when I heard about this concept of the long tail. If you took the best sellers on Amazon, the total volume of those books sold doesn’t match the very long tail of books that sold 5, 10, 15, 100 copies. And I think you’re right. As customers demand that optionality, as customers demand a little bit more interoperability with the platforms that they use, I imagine you’d be foolish if you didn’t for instance, open up some of your partners to see your graph or open up your partners to make activation a little bit easier or easier to measure coming out of there. Everybody wants to find the right experience for their customers so they can keep those customers. They can keep that money flowing in.

Kyle Hollaway:

That’s great. So, as we’re talking about the sharing of data, obviously a key component of sharing is matching and the ability to take two disparate sets of data and bring them together. That’s obviously a key component to Snowflake and Snowflake’s clean room concept. Recently, we announced together the release of Acxiom’s match multiplier as a native app within Snowflake’s at marketplace. Before we go into maybe some on the match multiplier itself, talk to us a little bit about the marketplace. What are you guys trying to achieve? And what do you see as the future of the value prop of having a marketplace on top of Snowflake or within Snowflake?

Rob Saenz:

So, if we’re going to talk about what’s in it for us, very frankly, we want more people to use Snowflake and we want people to turn on… We sell our software as a utility. There’s no licensing fee. So, the more you use it, the more you turn on computing resources to query data, whether it’s your data or it’s data that you have been granted through governance, through our data collaboration, the ability to see, man, we’re happy. So, we see ourselves as the matchmaker in this process where even outside of your vertical, we have oil and gas, energy data that oil and gas companies in order to function, they need to consume it. We want them to consume it through Snowflake. So, we enable providers to do that. And in the process we save man hours and energy and all these things because, man, you don’t have to take these giant data sets and transfer them over the internet compressed and then uncompress and load them up into the customer’s environment.

And then as soon as the customer has done it, man, it’s out of date with whatever was sent across those pipes. So, the first principle of the data marketplace is realtime. Can you make it so that if I’m using Acxiom as a mechanism to enhance an audience or all of my addressable audiences, man, I want to see what’s the latest and greatest that Acxiom has. I don’t want to wait for them to ship it over and wait for my operations team to hydrate it into my environment. I just want it to be there for me to query and even run a campaign or run an onboarding with an audience that I’ve selected. So, the first principle is let’s just make it easy for people to just work together. And so a lot of our marketplace, probably the gross majority of our marketplace, I don’t know the percentage, is just customers taking advantage of the marketplace to use free data sets, whether maybe some foot traffic, to query someone else’s account.

And to incentivize that behavior, what Snowflake does is we give a rebate, kind of like YouTube gives payouts so people watch your videos. We give a rebate. So if you go and query someone else’s data to that someone else, that provider is incentivized to make more data available for everyone to use. And then maybe the second pillar of that is, let’s make it even more attractive by reducing more complexity.

I think you’ll sense a theme with us, which is one of the things that drew me to the company is we really try to make it super easy for our customers. We even try to make it easy for customers to do the TNC process and even collect payment information and pass that on to the provider. So, part of when we talk about match multiplier, one of the central tenants of that thing is it will allow you as a marketer to come and use the application, pay for use of the application for just what you get. And then the parties that participated the publisher and the provider of the app will get that revenue. And we’re only going to charge a nominal fee for hosting and the administration of the app. So, the vision of the marketplace is to just be a destination for people to transact with each other and to make it really simple.

Kyle Hollaway:

So, Dustin-

Dustin Raney:

Do you see… Oh, Sorry.

Kyle Hollaway:

No, I was just going to say, you’re the chief evangelist for match multiplier.

Dustin Raney:

Yeah.

Kyle Hollaway:

Why don’t you give listeners a quick rundown of what that is and what we’ve launched there with Snowflake that is easier, right?

Dustin Raney:

Yeah, absolutely. Easy is the word and that’s what we were going for. And that’s why we were so drawn to the app marketplace positioning. And the way that if you think about how match multiplier works in the most basic terms, let’s say an advertiser has dustin@yahoo as an email, but the publisher might know me as dustin@gmail. So, if I were to try to match those two signals, it’s obviously not going to match, but both of them are me. They’re emails that represent me. So, Acxiom, our match multiplier product sits in the middle. So, we have both dustin@yahoo and dustin@gmail.

So, we’re able to help enhance and make that match and to be able to set ourselves in a position to increase, provide lift on those match rates allows brands, if you think about in whatever use case that is whether it’s activation or you’re wanting to do just oversight analysis, maybe it’s two brands that are wanting to see what their overlap is in their best customers. So, they’re just trying to really analyze that. Increasing the lift in those match rates using direct first party data that’s secure is a big deal. That’s the value we’re trying to add.

And honestly, it leads me to my last question before we wrap it up, Rob, which is the type of user that is in Snowflake. You guys started in 2012 and I’m sure that’s changing and evolving and is going to be rapidly evolving over the next few years. But would you say today, the end user is more technical, like a developer type? I think of a marketer, if I’m working for an agency trying to run a campaign, is it for those people as well? Tell our audience who you see using Snowflake the most.

Rob Saenz:

Yeah, absolutely. So, I think the last count that I saw, we have around 6,000 customers. And I would say if you trended it from, like Kyle was saying from when we were founded a decade ago, it was technical users. They were spending a bunch of money on a data warehouse appliance and they wanted to replace it. They were infrastructure people, they spoke SQL, they spoke maybe some infrastructure networking, et cetera. And we were trying to sell to those people initially. In a lot of ways, we were very successful in getting those types of personas involved in the platform. But increasingly what we’ve seen is the trend of people trying to solve real business problems. In general, right cost savings and revenue growth. But the latest chapter with Snowflake is enabling maybe even, not the consumer, not somebody who’s going to log to a platform and start clicking and all of a sudden realizing value, but enabling people like you guys to design products, deliver products that can live in the marketplace.

So, right now a user of match multiplier could be that guy that you mentioned. It’s a guy that’s used to building lists maybe in Excel and he’s used to clicking some buttons and launching a campaign. And so this experience is in his wheelhouse, it’s going to be, except that it’s completely online. He would go in, he would select that audience, he’d see a preview of that audience. And then he would choose his destination and then Acxiom would just magically jump in the middle and deliver and an output that’s ready for a campaign. So, for him, maybe the experience is very non technical, it’s very familiar, but that’s based on what you have delivered.

So, a lot of the shift has come from people that were down in the data, organizing, writing the best routines and architectures to make that happen to an audience that is like, “Okay, now I see the power of what these Lego blocks are able to do. I need some higher, more abstract tools.” So, Snowflake starts building that. We acquire Streamlit, a UI interface for you to start building out, not only the building blocks under the covers, but the visual building blocks to create that experience that is easier to grasp, it’s stickier for an end user that doesn’t have the technical chops of what a Snowflake buyer had 10 or 12 years ago.

Dustin Raney:

Awesome. Rob, we’re going to wrap up with our last question. So, it’s 2035, what does consumer engagement look like? Do you see a scarier or whaling type future, or do you see brighter days ahead?

Rob Saenz:

I’ve heard you guys ask this to so many people and when I knew it was coming, I thought about it, but I figured I was probably just going to take it as the mood struck me. I really think that at the end of the day, the very best that we have is the future in a sense that we design. And it’s also limited on the other side, by the future that we tolerate, we are willing to trade off convenience to get security or security to get convenience. So, I think what I like about being a part of an organization like Snowflake and working with a company like you guys, is that it’s about connecting people. And so I think if you’re pointed in the right direction and I hope that we can start to be that way as a society, if we’re trying to connect and serve people, then all of our technology and all of our experiences that we’re designing will fall in line with what our goal is. We’re going to get what we focus on and what we chase.

Kyle Hollaway:

Love that, man.

Dustin Raney:

Yeah. Great answer. Thanks. Well, that’s it for this episode, we appreciate everyone joining and certainly, Rob, we appreciate you being on and sharing your insights and information about Snowflake. Hopefully it was informational to our listeners and we really look forward to future podcasts. And Rob, I’m sure you’ll be back on the list to come join us as we see some of this transition happening within the broader ecosystem. So, thank you very much.

Rob Saenz:

Glad to. I’d be honored, thanks.

Kyle Hollaway:

All right, well thank you everyone. And look forward to joining you all again on the next podcast.