SaaStr 796: The Secrets Inside Google Cloud's Growth with Sarah Kennedy, Vice President Google Cloud Marketing

0:01Welcome to the official Sastr podcast where you can hear some of the best Sastr speakers. This is where the cloud meets. Up today on the Saster Podcast. Is cost a big issue for AI at a meta level? And what I mean is, a year ago, we're like, oh my God, we're never going to be able to afford anything that's so expensive. But everything, not only is everything getting cheaper, but we're getting better. As developers, as users, we are actually less wasteful of resources. So like the scale-ups I work with, the ones at$100 million or above, honestly, I don't hear a lot of worries about cost anymore.

0:35But what are you hearing in general or at the CIO office around cost versus ROI? Cost is always obviously top of mind for CIOs, but I don't think they're as worried about that as they are falling behind and making sure that they've got the right resources to experiment and to build and learn and test. Because they're relying, like at least a lot of our customers, they're relying on us to figure out how do you optimize the compute, the storage and the processing power and the cost of the TPUs that are powering a lot of these models. We are obsessively worried about that in a good way.

1:33number one agent in G2's winner report. Learn more at inter.com slash Saster. That's inter.com slash Saster.

1:46All right, everybody in Sass, this is it. The biggest, most action-packed Sass and AI event of the year. Saster Annual 2025. It's coming this May. Yes, this May. Three full days, 10 ,000 Sass, AI, and cloud leaders, and more tactical, no-fluff content than you'll find everywhere else. Hundreds of workshops, thousands of brain dates and one-on-ones. If you want to scale faster to 10 million, 50 million, 100 million, 300 million ARN beyond, you need the right playbooks, the right relationships, the right connections and the right people in your corner. And Saster Annual is where it happens. We'll have hundreds of legendary speakers from companies and CEOs from Snowflake, HubSpot, OpenAI, Canva and more.

2:26We'll have more networking than you can handle. You'll meet your next VC, your co-founder, the next biggest deal. I was just talking with the founder that closed a$450 ,000 deal just the month after Saster Annual last year. And we'll have a new AI demo and pitch stage where hundreds of you will be able to do quick pitches of your hottest new AI feature or product. And a chance to win up to$5 million in VC funding from Mayfield. Apply right on SasterAngual.com to pitch your AI startup. So don't wait. Grab your tickets now at SasterAngual.com. And if you want, use my code Jason100. Jason 100 to save$100 before prices go up again.

3:02That's Jason 100 at sasterannual.com. See you in May, May 13th through 15th in the SF Bay.

3:18Hello, everyone, and welcome to another Saster recording of our Saster What's New podcast. We're so excited to be joined by Sarah Kennedy, who's the VP of Google Cloud Marketing. We have been long-term, long-year partners with Google Cloud. Excited to see them back at Sastre Annual again this year. You guys have next coming up. You guys are growing like crazy. There's a lot going on at both Google Cloud and Google. So one of the very first things we want to dive into for today between you and then Jason, who's here, our Sastr CEO and founder as well, is Google Cloud has been on a tear, right? Jason has written it up on Sastr.com.

3:57You guys are growing 30 % just in the last quarter alone. At the time of this recording, might be more by the next time we do a check-in. Can you share with us just for the Sastr community and audience, what's driving a little bit of that growth from your perspective on the inside? What have you seen? Awesome. Thanks, Amelia. Thanks, Jason, for having me. First of all, it's great to be here and to chat with y 'all today. But yeah, gosh, Google Cloud is growing fast. I'm actually, it's my fifth year at Google. And I think when we started, we were at a 12 billion run rate. And now she said it's more than tripled.

4:29And it's an incredible journey right now that we're in the middle of, and especially in just the AI wave that has taken over all of our worlds and lives. But it's the core components, really Google's AI stack that set us apart. And that's everything from the world-class infrastructure that we have at our data centers all over the world. I just got to visit one for the first time a few weeks ago, and it was magical. And just seeing the incredible cooling technology and some of the ways that we drive efficiency, it's just, it's hard to appreciate until you're physically in that venue in that moment.

5:00But it's also then the powerful platform that we provide hundreds of different AI models through from world-leading models. And it's not just Gemini models, it's Anthropic, it's Llama, it's so many of the other models out there to give people choice and access to everything through a single platform that's Vertex AI for us. And then also our Gemini models, obviously we want to have the world's very best and most powerful models. And we're doing that, but we're also doing that all with the privacy and security protections that are true to who Google is built in very much from the start. So I think when we combine really that level of just innovation and the choice and flexibility we're giving our customers, it's a recipe for growth.

5:38And that's why we've seen that massive tear over the last, really in the last few years. Sarah, can I push on that a little more? Just because you are thinking about it from a marketing perspective. I would say the vast majority of startups I work with are running part of their infrastructure on Google Cloud, right? For sure. But sell a little harder, but more specifically. What I mean is you're gaining market share. And yes, there's AI. And I want to spend the rest of our time talking about AI because honestly, it's the only thing anybody wants to talk about. But be both objective but a proponent.

6:09Why are you gaining share? But you don't gain share at this scale, right? Unless there's two or three core reasons folks come in and they just pick, at least for more of their budget, if not 100, because we are in a heterogeneous world. But why are you gaining share? A lot of times, Jason, we have customers who will come in and move one workload over. And because, again, we started out with our core tenets of who Google Cloud is, we wanted to provide choice. And we didn't want it to be this all or nothing. You have to move everything to Google on day one or not. And so a lot of what happened over the years is that, especially in the last three to four years, we want a lot of business just with single, mostly like modern app workloads.

6:50For example, people would come over and build a net new app on Google Cloud, start to work with us and see how powerful actually that relationship was. And what you're now seeing in the last one to two years are all of those who three to four years ago took a bet on us and then have started to migrate more and more workloads over the like more recent 18 to 24 months. and with AI, they're looking at it. We were the third place player for a long time in the cloud space, but we really are the first choice when it comes to AI. So now when all of those CIOs and CTOs had started to consider, now what do I do?

7:22They're getting pressure from their boards. Who is the player that actually can keep me the farthest ahead of the curve? Almost all of them came to us, whether they were already working with us for their cloud infrastructure, compute storage, et cetera. They came to us to actually ask the question first, what should I be doing with AI? And I would say, if you think about the power of what's happening across all of Google, we run some of the most sophisticated compute infrastructure in the world, answering billions of search queries, streaming millions of hours of YouTube videos. We have seven products that reach more than 2 billion users.

7:53That's like crazy. And the answer to our customers is very simple when they look at who they're going to partner with. And as we're infusing AI into YouTube, into search, they want to come to use all the same technology and infrastructure that's been available to Google for the years. And so that's where I think we're winning business. We're building credibility. But we're, most importantly, keeping them ahead of the curve. And that's how we're winning and growing right now. Can I ask you one nerdy, tactical GTM question? And I know that your team is big. But when I look at the public folks that talk about migrating workflows over, we often see it's like a seven to ten year journey.

8:27Yeah. Because you'll migrate one workflow and then two, and then you'll often see some of the pure play infrastructure players. It'll be like eight years until they maximize out an account. So question for you, having seen all of it, with the sales team and the GTM team, like how do you get it just right? So you remain super customer centric? Because if you're in sales, I might love to, I might push a little too hard. Bring everything up for tomorrow, Sarah. How do you help the team manage that? Because it could be the better half a decade, right? Yeah. Yeah, I honestly think we're playing the long game.

8:57And I think the sales team fully appreciates that. And we think about our investment and time spent with customers as the payout doesn't happen immediately, nor should it, because they have to build trust. They also have to learn. And we have to figure out what it is that we have to help support them with to learn, to then migrate. And so there's a ton of work we do right now that's around just training and how do we think about investing in their developer community within those companies? How are we educating them? How are we bringing them along on the journey? Because it's one thing to sell something to a CIO or a CTO, but it's a darn big another to go deep into the org and figure out how you actually get a lot of those developers.

9:36Many of them are certainly certified in using other cloud platforms if they're not already using Google. So it's a long game and it's actually something we talk about this a lot. You don't see this immediate return necessarily from investing in students, but it's the right thing for the future of where cloud and AI are going. You talked about the CIO and together with Google Cloud training the developers on the team, which actually haven't lived too much, versus the classic developer bringing it in and saying to the boss and the CIO, I've made this decision, please approve it. How has that changed the buying pattern there?

10:08We go back to the old Twilio billboards pre-everything on the 101, ask your developer. It was a powerful message. And it's really interesting to me how the CIO empowers the developers and educates them versus the other side. So it's especially apparent in the last couple of years that we've seen most CIOs, just like most CMOs, are going and looking to, it's whether it's your developer, like your most passionate developer who actually has been working in AI and ML for a decade. Now, all of a sudden, when your board's calling you, you're like, they're going straight. And it's a bottoms up conversation where they're the ones bringing in which models are we going to test?

10:47How are we going to experiment with which AI platforms? It's almost primarily a developer led motion in pretty much every environment. CIOs have to be accountable and responsible, but they're learning. They're being taught by their developers the things that they've been distant from for quite a while. And so that's a very, for us, marketing to developers has always been important. It's never been more important though right now in this moment because that is a truly bottoms up moment. And so that is, it's different from like the traditional historical approach of compute and storage when you're dealing with eight digit, nine digit deals in some cases of a massive, it's like procurement, the CFO, the CIO, they're the ones out there making some of those choices about their long-term cloud strategy and who's going to provide security, stability.

11:30And that runs so many different facets of their organizations. And that is not just a developer-led decision or motion, but AI, frankly, is. And that's an exciting, actually, time to be, I think, both a developer, but also the CIOs who are learning from them. And they're probably also seeing a different path for succession planning based on just the passion of people leaning in. I mean, so we're seeing that shift. And I think that's true in almost every function. So it's interesting, just from a marketing perspective, you're seeing a resurgence in the power of the developer making the decision in AI, right?

12:03Absolutely. And maybe if we're just talking about cloud in general, outside of AI, maybe that's migrated to the CIO. These are big decisions now, right? And the CIO, if I'm going to put more money into Google Cloud, just generically, that's a CIO level decision. But if all the budget's going to AI, it's an interesting, it's not a tension, but it's a huge dynamic because so much budget's going to AI, right? So I guess it's empowering the developers more in the org. It is. And I think they were empowered already with the cloud decision. I do think they were weighing in a lot. And I think they were heavily weighing that.

12:32But when it's a 10-year, 5-10-year contracting commitment, it just gets bigger. And the influencers and the decision makers that are involved in that buying decision just balloons. Versus AI, it's all experimental. And then, yes, there's going to be the larger, bigger enterprise deals over time. We've got many of those, too. But the developer usage is what's teaching us what the future of this platform and our product set and our models and our infrastructure really needs to look like. Yeah, I can't. I'm sure there are many that are able, but I can't imagine most CIOs can keep up with the pace of change of models.

13:04It's fast, right? It's hard to Google and it's hard. In Google, it's hard. Even in Google, it's got to be hard, right? It is. What came out yesterday? Literally. And we have, fortunately, we use AI to summarize AI news and announcements. It's just such a meta environment in that respect, but it's the pace of changes. It's cool, though. What a time to be working in tech and AI. And what an exciting time for everybody to figure out how can you leverage this to your benefit in your job, in your personal life. It's just the possibilities are endless. Yeah. Can we tie it for a minute? Let's tie it to B2B and SaaS and maybe the big deal with Salesforce.

13:41Yeah. But how big was it on two and a half billion or something like that? Big Bs, yes. Before we get there, I just want to, when I was watching the earnings call with Mark yesterday, or maybe it was a CNBC, just to your point, he was so excited. Yes. Oh, my God. I haven't seen him like this in a decade. And that's saying a lot because Mark gets excited about a lot of things. But it was authentic. But the thing, going to your point, like people at Google are excited. Yeah, he can get up there and sell. Even if they shrink this quarter, he's going to get up there and tell you it's the best thing that he's ever seen.

14:12But you can smell the authenticity in someone that's been doing it for two decades. It was rejuvenating for him, right? Rejuvenating. It's an exciting deal. He's placing a massive bet, and we are too. I love these kind of partnerships because there's so much good that can be brought to the world, frankly, whenever two companies of this size and scale come together with that authenticity. And that's what it's all about. Listen, it's been a journey with Salesforce to get to this point. But the deal that we announced, it spans everything from cloud infrastructure to agentic AI, customer service, collaboration, everything.

14:47And now all of CRM as well as agent force inside of Salesforce are going to now be available on Google Cloud. But this is actually a lot of this came from it's like a customer led thing for both of us. A lot of the enterprises who were already working with Salesforce and already working with Google demanded and said, this is where that demand came from. We want to run Salesforce on Google's infrastructure because of the security features that we offer, because of the compute and storage and the efficiency and all the other things that are the goodness of it. And when we come together, we bring them more value in doing that.

15:17But also them adding Gemini models to Agent Force is going to be a huge unlock. I think everybody's really excited with the potential of that. And I have been from day one when I saw Mark announce Agent Force. Okay, what are we going to do together? Because this is really the potential there is endless, but having access to Google AI to power their agents for all the companies that work with Salesforce already, that's going to be a big unlock. But also, this is all about letting customers work where they want to work. We're talking about this with Workspace. For years, large enterprises have had a lot of different opportunity and choice when it comes to collaboration and productivity platforms, but typically only two of them people really want to use.

15:56and that's Google Workspace and Slack. And so to create a better experience for our joint customers there, we're going to be integrating those platforms. I personally, I've been waiting for this for a long time. Accessing Google Drive files inside of Slack, accessing Google Calendar within Slack, bringing Slack and Google Chat together. I think this is like a marriage in heaven, those two products, because most of the users who are passionate about user experience, those are the two products that they already use and want to use more of. And so bringing that together for the first time is going to be really cool.

16:25I hadn't realized things like application. Integrating chat in Slack is pretty interesting, right? Yeah. It is funny as the world has changed over the last few years, I live more and more in Google. I always did, but more. I don't want to go to, honestly, I don't want to go to Slack. It's a tax on my time. It used to be wonderful and there's no knock on Slack. It's a great business, but it's become another browser or another app I have to do. And I just, I want to simplify my life. So running that, even that integration would actually be a game changer for us. Yeah. Okay. Let me ask you one other nerdy question on this, but just my, I'm interested, but ignorant.

17:01The first point, the Google Salesforce application integration, we could spend a whole bunch of time on. But at the first one, when you talked about running Salesforce on Google Cloud, if I've got that right, and that, what does that mean? And then I'll ask you a follow-up question, but I get, you get, I'm running on Google Cloud. It's highly secure. I've got my data, but Salesforce is pretty secure. There's some pretty good enterprises running in Salesforce. Like I haven't heard a lot of complaints, right? So I'm just very interested. I've purchased both, right? I have a massive commitment to both.

17:34I see why I might want to pull my data, but what's the real driver? Like, why am I waking up and this is like a P1 priority to not have isolated clouds? I do think security is at the heart of it. I'll be honest. Is it? Yeah. Yeah, and listen, it may be different for very large enterprises versus small businesses that are thinking about different problems that they're solving when they're actually asking for this. But we've seen this in a lot of places. It's, number one, where your data lives and how security is going to be even more important in the next year. And that'll be true. I'll say that every single year until I retire.

18:09Yeah. But it's never been more important than right now, number one. Number two, everything you do with AI is built around where your data estate lives. And so that's the heart and soul of what companies are going to be able to do to unlock even more value. And when you think about the kind of data that lives inside of Salesforce and then powering that with the superpowers of Google's infrastructure, even being able to train custom models and using BigQuery for that, and that be your storage layer of your cloud ecosystem, it's just much more seamless. And again, the security layer around what you do with AI will become, those two things cannot be separated anymore.

18:44And so that's, I think, a big part of the choice and the decisions a lot of these CIOs are making is, who do I really trust from here forward to secure my data and to be the foundation of everything and every application I'm going to leverage, whether it's through agent force or whether it's building their own custom agents on top of Google Cloud? That's never been more critical. Yeah, it is. It is. First of all, security is more important in the age of AI, right? At the end of the day, I think half the CIOs' decisions are about security one way or the other. They are. Half, right? If CIOs are looking and saying, listen, AI is changing the world.

19:16We've standardized a lot of our work on Google Cloud. And yeah, we're investing mass amounts in Salesforce, but we want that data to reside on Google Cloud. Yeah, it's the big driver here, right? It's heightening the concerns here, right, of where my data lives. Let me ask one follow-up on that. When you're talking with CIOs and senior folks in the office, are they asking about agents? Are they asking about work? What's their language for AI at the application level? Are they all talking about agents the same way we are on LinkedIn and X or are they describing this in other ways? It widely varies.

19:51Some of them are still talking about it the way that they have been because they've been deeply investing in this for years, like automation, because a lot of it is just standard on a bit. But the intelligence layer on top of it and the autonomy of agents and how they think about, some of them are more advanced than others and their competitors are driving them to be. And that's actually a healthy thing that they're talking about it in the language of agents. Giving them that language as a CIO to partner with their peers is a really important part of the value we bring. And we do that behind the scenes with them and educate them on.

20:20And it's not just about what language they want to use. It's about the language they need to use to be able to communicate with less technical peers who are all the business case, use case drivers of a lot of the investment they're making in AI right now. Got it. And maybe one last question on the CIO, and then maybe we can just have a little fun and talk about Google Max and anything else you want to hit. Is in the real world, when you're talking with CIOs and stuff, I have my own opinions here, but I'm not talking to CIOs. Yeah. Is cost a big issue for AI at a meta level? And what I mean is a year ago, we're like, oh, my God, we're never going to be able to afford anything that's so expensive.

20:54But everything, not only is everything getting cheaper, but we're getting better as developers, as users. We are actually less wasteful of resources. So like the scale ups I work with, the ones that 100 million or above, honestly, I don't hear a lot of worries about cost anymore. They've gotten good at it. But what are you hearing in general or at the CIO office around cost versus ROI? Cost is always obviously top of mind for CIOs, but I don't think they're as worried about that as they are falling behind and making sure that they've got the right resources to experiment and to build and learn and test.

21:27Because they're relying, like at least a lot of our customers, they're relying on us to figure out how do you optimize the compute, the storage and the processing power and the cost of the TPUs that are powering a lot of these models. We are obsessively worried about that in a good way. That's why when I visited the data center that I mentioned, I learned so much about the way that we have the most efficient data centers in the world and specifically the way that we're training models and using TPUs. And the cooling systems provide less power surge and power pull for us to our data centers that makes it just cheaper.

21:59And it allows us to actually do more. And that's something that we don't talk about it that much, but it's so central to the future of cost optimization for AI. When you get AI that's that accessible, I feel like we're going to enter an era maybe next year where there's no limits. At the application level. I'm not saying that ML engineers can't consume every GPU and GPU. But what I mean is at the application level, I think an era is going to come where it's going to be so cost effective that we can do almost anything we want at the application level. As long as we're smart about how we use it.