 Hi, I'm Kyle Rourke, VP of Platform Strategy here at Snowflake and I could not be more excited to be here today with Margaret Sherman, who is the head of data strategy for Sonos. Now all throughout the day today, we've been hearing from lots of customers from all over the world and hearing about their journeys and how they've transformed their business by embracing the data cloud. And of course, this next story is one I am personally very excited about because I'm a huge Sonos customer and I'm sure many of you are as well. Thanks, Kyle. As you mentioned, I'm the head of data strategy at Sonos and what that means is that I set the data priorities for the company as well as guide the company on where to invest now and in the future to get the most out of our data resources. I've spent about four years at Sonos. Previously I was the head of product data and worked a lot with setting up Snowflake and the IoT data and I've been in technology for 20 years. And the last 10 years I've spent in data analytics and machine learning. That's very cool. Now, how does your company view data in general? How does your team fit in the overall strategy of how Sonos leverages data? Yeah, so Sonos is a sound experience company and we really pioneered multi-room wireless audio and we made that experience amazing and truly changed how people listen. And our mission is to inspire the world to listen better. And everything that we do is in service to that and data is a part of that. So we really believe that data is fundamental to helping us achieve our mission. Data helps us build a better business, build better products and ultimately we think it helps us make our customers happier. Now before Snowflake, maybe let's go back in time, you've been at Sonos for the last four years. Maybe go back in time a little bit for the time pre-Snowflake and what were some of your challenges that you guys faced before you started with us? Yeah, it was really challenging. So I was actually leading the product data team at the time and we used Snowflake primarily for our IoT data. And so we're collecting tons of data, but we're really struggling to leverage it. Essentially what would happen is if we wanted to answer a single question about what was happening with the customer experience, we would have to have data engineers go and write some code, spin up clusters. This could take weeks just to extract the data and get it into a shape where analysts and scientists could go and work with it. And so we really went from being in a place where it was taking weeks just to answer a single question to now we can do things in hours. So it really changed things for us. Wow. And so the benefits of, I mean, obviously being able to go act on information very, very quickly. What are some of the other benefits that that's driven for you guys? I mean, our data people love it obviously because if you think about the process of data science, it's very iterative. So you're going to ask a question, you're going to go and investigate your data, you're going to do some data processing, you're going to do some visualizations, and then you're going to come up with more questions. You're going to want to dig in, you're going to want to pull more data, and you're going to want to join it together, do different cuts and pivots. And before that was just off the table because as you can imagine, if every time you have to go weeks to go do that, it's just impossible. Whereas now, our data scientists can churn through these problems very quickly. And the data engineers love it because they're not sitting around waiting for jobs to finish for forever. They are able to get through their code faster. And as one of my engineers likes to say, she told me, she said, Snowflake, she's a beast crunching through the data. You mentioned IoT. I think that's obviously a very challenging space for a lot of customers. And there's a lot of interest in it. Maybe give me some more thoughts on how much data are you guys bringing in? Is it small, large, has the volume been something that you could handle with Snowflake? Yeah, I mean, that was why we chose Snowflake. So prior to having Snowflake, we were really struggling with a lot. I mean, we have very large volumes of data, as you can imagine, from an IoT device because we're collecting from over 10 million homes across the world. So it's quite a bit of data. And all of that doesn't fit in a traditional sort of data warehouse. We were trying to push some of it into SQL. And we were essentially taking just a handful of our telemetry events. And we were boiling them down to daily and weekly aggregates. And even trying to push that into SQL was just too much for it to handle. And with Snowflake, we had processing jobs that were timing out in SQL server after running for hours and hours. And then Snowflake could just crunch through it in a few minutes. So it was, yeah. I mean, I literally almost fell off my chair. They showed me the comparison numbers. Well, so now, well, so it's been a good experience for you. But let's talk about your customers. And I think I can say, as a customer myself, I've always had a great experience with Sonos. That's probably why I keep buying more and more and more of them. But talk to me about how has data really, really helped you guys drive that customer experience in some very tangible ways? As a Sonos customer, we really hope that you enjoy the great sound and freedom of choice and ease of use that the product brings. But obviously behind all of that, behind that really easy experience, it's very complex. You've got a lot going on when you're interacting with your Sonos system. So it's not just a single piece of hardware. You've got you've got a mobile device potentially that you're using to control it. You have third party voice services. You have music service providers, Wi-Fi, network traffic. All of these things are going on. So for us to really make sure that we're creating an amazing experience because we're super customer obsessed. So Margaret, you know, one of the things that I've always experienced as a customer of Sonos is that, you know, frankly, it's just for me, it just always works. And that's one of the best parts of the customer experience, whether I'm pulling up Spotify on my phone to go listen to it or whether I'm plugging in a sound bar into the TV, everything seems to just work. So maybe just walk me through why it's been such a good experience maybe for me. Yeah. And how you use it with it. So some of the, you know, kind of at a high level, an example of what we do is we use Snowflake and because of the power of Snowflake, we're able to bring together different kinds of telemetry about what's happening in our products and our services. And then we can try to pinpoint reliability issues and determine what's happening, like what's causing them. That was kind of the first thing that we attacked with Snowflake, was really to go in and dig in and join some different events together and start slicing and dicing and looking for, you know, what are the main problems that we're seeing and what are the fixes to them. And the product team was able to find some reliability issues that they were able to fix and they shipped the fixes and we were able to significantly reduce the rate at which some of these errors were occurring. So just by having all the data in one place and being able to go actually act on it quickly and of course in a cost-effective manner, it really did let you guys really pinpoint any issue when it did occur and go support a fix quickly. Yeah, I mean, we were able to find things that we didn't even know were happening because we could really drill into the data. So it wasn't just having one place, it was also just being able to go dig into a different level of fidelity that you didn't have before. I mean, you should see some of the Tableau dashboards that we've put on top of Snowflake, so it's pretty impressive. That's awesome, that's awesome. So what really set it apart? I mean, you've been in the business, you're an expert in this business and there's a lot of options out there. You know, what really set apart Snowflake from everything else at the time and even now in your opinion? Just like you said about Sonos, it just works. We were able to stand it up really easily, we were able to load data into it really easily. You know, it's pretty flexible in terms of what you can do. So we, for example, we use the variant column quite a bit and that allows us to take kind of this semi-structured data and throw it in there and have it indexed and then we can work with it as we want to and we don't have to have like a real complex data pipeline upstream before we throw things into Snowflake if we don't want to or we can. We really like, obviously, I mentioned the speed, I keep mentioning that because it's really powerful and it holds a ton of data, but even better is the cost, right? So we're pushing tons of data into Snowflake and we're not having to pay that much for that cost. We're basically paying S3 costs and then, but then you pay, you just pay for what you use, so you're just paying for your processing costs and even that is pretty easy to optimize and you guys provide tools for that and support. I mean, you helped me save thousands of dollars for some months this year, so. That's great. It's really great. What's the plan for the future? Where do you see Sonos and Snowflake and the data cloud? Where do you see all that intersecting in the future? Yeah, I mean, we really want Snowflake to be kind of the center of our data platform and bring all of our data together. We want to live the data dream, right? We're going to get our data together and do all sorts of cool analysis. So we have a bunch of different projects that we want to be able to do. For example, one thing that we're looking at doing is bringing together our product telemetry data and our customer support data so that we can try to find patterns in terms of the types of errors or sequence of events that happen before somebody calls us so that we can potentially intervene and fix the problem before somebody even has to reach out to our care team. And then another place that we're looking at using Snowflake is connecting it to Salesforce. So for people who are interested in hearing from us, we could do things like when you set up a new product, we can send you information about how to use that product or if there's new features available for the product that you have, we could send you information about that. And with Snowflake as our back end, it really helps us be able to tailor the customer experience. So Margaret, you've talked a lot about Snowflake and using Snowflake and it quote, just works. I mean, you guys are running a very large, a very, very large implementation of Snowflake using IoT data coming from, as you mentioned, millions and millions of devices. What's been the overall lift to the organization just to go manage Snowflake and keep it going and keep it up and running? It almost runs itself. Like I'm almost surprised because, you know, we obviously use a lot of different third party tools and most of them require quite a bit of intervention on a regular basis. I would say Snowflake, it stays up and running and you have great tools for managing the whole system. And it's easy to see what's happening. It's easy to see, you know, when different clusters are spinning up and spinning down. We have Tableau dashboards that we use to monitor all of our usage. And so, and you guys provide all the data for that to make that really easy. So, that's really great. Can't say it enough, we love Snowflake. So Margaret, thank you so much. It was great hearing your story about Sonos and how you leverage Data Cloud. So again, thank you for being a customer and thank you again for being here today. Thanks for having me, it was a pleasure.