Good afternoon. Thank you for joining today's Snowflake Q1 Fiscal Year '24 Earnings Conference Call. My name is Cole, and I will be your moderator for this call. All lines will be muted during the presentation, and there will be a chance for questions and answers at the end. I will now hand it over to our host, Jimmy Sexton. Please go ahead.

Good afternoon, and thank you for joining us on Snowflake's Q1 fiscal 2024 earnings call. With me in Bozeman, Montana are Frank Slootman, our Chairman and Chief Executive Officer; Mike Scarpelli, our Chief Financial Officer; and Christian Kleinerman, our Senior Vice President of Product, who will join us for the Q&A session. During today's call, we will review our financial results for the first quarter fiscal 2024 and discuss our guidance for the second quarter and full year fiscal 2024. During today's call, we will make forward-looking statements, including statements related to the expected performance of our business, future financial results, strategy, products and features, long-term growth, our stock repurchase program and overall future prospects. These statements are subject to risks and uncertainties, which could cause them to differ materially from actual results. Information concerning those risks is available in our earnings press release distributed after market close today and in our SEC filings, including our most recently filed Form 10-K for the fiscal year ended January 31, 2023, and the Form 10-Q for quarter ended April 30, 2023, that we will file with the SEC. We caution you to not place undue reliance on forward-looking statements and undertake no duty or obligation to update any forward-looking statements as a result of new information, future events or changes in our expectations. We'd also like to point out that on today's call, we will report both GAAP and non-GAAP results. We use these non-GAAP financial measures internally for financial and operational decision-making purposes and as a means to evaluate period-to-period comparisons. Non-GAAP financial measures are presented in addition to and not as a substitute for financial measures calculated in accordance with GAAP. To see the reconciliations of these non-GAAP financial measures, please refer to our earnings press release distributed earlier today and our investor presentation, which are posted at investors.snowflake.com. A replay of today's call will also be posted on the website. With that, I would now like to turn the call over to Frank.

Thanks, Jimmy. Welcome, everybody, listening to today's earnings announcement. Snowflake's product revenue grew 50% in Q1 fiscal year 2024 totaling $590 million. Our net revenue retention rate reached 151%, and remaining performance obligations came in at $3.4 billion, up 31% year-on-year. Non-GAAP adjusted free cash flow was $287 million, up 58% year-over-year. We are, however, operating in an unsettled demand environment and we see this reflected in consumption patterns across the board. While enthusiasm for Snowflake is high, enterprises are preoccupied with costs in response to their own uncertainties. We proactively work with customers to optimize their environments. This may well continue in the near term, but cycles like this eventually run their course. Our conviction in the long-term opportunity remains unchanged. Generative AI with its chat-style of interaction has captured the imagination of society at large. It will bring disruption, productivity, as well as obsolescence to tasks and entire industries alike. Generative AI is powered by data. That's how models are trained and become progressively more interesting and relevant. Models have primarily been trained with Internet and public data, and we believe enterprises will benefit from customizing this technology with their own data. As Snowflake manages a vast and growing universe of public and proprietary data, the data cloud's role in advancing this trend becomes pronounced. AI's focus on large language models and textual data, both structured and unstructured, will lead to rapid proliferation of model types and specializations. Some models will be broadly capable with shallow functions, while others will be deep, specialized, and impactful in their specific realm. For years, we focused on the extensibility of our platform via Snowpark, making Snowflake ideally suited for a rapid adoption of new and interesting language models as they become available. AI is also not limited to textual data; equally far-reaching advancements will be seen with audio, video, and other modalities. The Snowflake mission is to steadily demolish any and all limits to data, users, workloads, applications, and new forms of intelligence. You will, therefore, continue to see us add, evolve, and expand our functions and feature sets. Our goal is for all the world's data to find its way to Snowflake and not encounter any limitations in terms of use and purpose. From our perspective, machine learning, data science, and AI are workloads that we enable with increased capability, continuous performance, and efficiency improvements. Data has a gravitational pull. Given the vast universe of data Snowflake already manages, it's no surprise that interest in these capabilities is escalating while their uses are still evolving. Data science, machine learning, and AI use cases on Snowflake are growing every day. In Q1, more than 1,500 customers leveraged Snowflake for one of these workloads, up 91% year-over-year. A large U.S. financial institution uses Snowflake for model training. Facing memory constraints with their prior solution, they chose to move feature engineering workloads to Snowflake. With Snowflake, they can fully ingest all data, replacing a sampling approach, which left models less predictive and long running. Snowflake enables machine learning for a broad spectrum of user types, not just programmers. For analysts, we have introduced, in preview, ML-powered SQL extensions such as anomaly detection, top insights, and time series forecasting. SQL proficient users can now leverage powerful machine learning extensions without the need to master the underlying data science. For data scientists and engineers, Snowpark is our platform for programmability. New here is a PyTorch data loader and an MLFlow plugin, both in Private Preview. PyTorch is a popular framework for machine learning, and MLFlow helps manage the lifecycle and operations of machine learning. Snowflake had an early start in support of language models through last year's acquisition of Applica, now in Private Preview. Applica's language model solves a real business challenge, understanding unstructured data. Users can turn documents such as invoices or legal contracts into structured properties. These documents are now referenceable for analytics, data science, and AI, something that is quite challenging in today's environment. Streamlit is the framework of choice for data scientists to create applications and experiences for AI and ML. Over 1,500 LLM-powered Streamlit apps have already been built. GPT Lab is one example. GPT Lab offers pre-trained AI assistance that can be shared across users. We announced our intent to acquire Neeva, a next-generation search technology powered by language models. Engaging with data through natural language is becoming popular with advancements in AI. This will enable Snowflake users and application developers to build rich, search-enabled and conversational experiences. We believe Neeva will increase our opportunity to allow non-technical users to extract value from their data. More broadly, Snowflake continues to enable industries and workloads. In Q1, more than 800 customers engaged with Snowpark for the first time. Approximately 30% of all customers are now using Snowpark on at least a weekly basis, up from 20% at the end of last quarter. Snowpark consumption is up nearly 70% quarter-over-quarter. The Snowflake Connector for ServiceNow is in public preview. Customers can access ServiceNow data inside of the data cloud without needing to manually integrate APIs or third-party tools. ServiceNow data is significant because it holds a wealth of IT and security data. The Connector is the first so-called native app built by Snowflake. Native apps, which are on Private Preview, run inside Snowflake's governance perimeter and make use of common services. Today, developers waste time convincing customers to expose their data. With native apps, developers can focus on their core interest: application development. They offload security and deployment concerns to Snowflake. During the quarter, we also launched the Manufacturing Data Cloud, which focuses on supply chain management as a data problem. Supply chain management is one of the few remaining realms in enterprise software that have struggled with the platform itself. Supply chains are all somewhat unique, and the data siloing problem prevents supply chain visibility essential to managing it. With the Manufacturing Cloud, Snowflake continues to evolve from being a data cloud to also being an operational hub for large enterprises and institutions. We also announced that Blue Yonder, one of the largest software companies in supply chain management, will fully re-platform onto Snowflake. Blue Yonder is a key participant in both the manufacturing and retail data clouds. They are the first major supply chain provider to make this commitment to creating the end-to-end supply chain platform on Snowflake. Supply chain management is a highly networked discipline, as the chains typically comprise numerous different entities. Therefore, we expect significant network effects from this strategic alliance with Blue Yonder. Our Summit conference in June will feature more significant product announcements, and we look forward to seeing you there. With that, I'll turn the call over to Mike.

Thank you, Frank. Q1 product revenues were $590 million, representing 50% year-over-year growth, and remaining performance obligations grew 31% year-over-year, totaling $3.4 billion. Of the $3.4 billion in RPO, we expect approximately 57% to be recognized as revenue in the next 12 months. This represents a 40% increase compared to our estimate as of the same quarter last year. Our net revenue retention rate of 151% includes five new customers with $1 million in trailing 12-month product revenue. Q1 revenue reflects strong performance in a challenging environment. We continue to focus on growth and efficiency. We generated $287 million of non-GAAP adjusted free cash flow, outperforming our Q1 target. In Q1, consumption varied from month to month. We benefited from strong consumption in February and March. Starting in April, consumption slowed after the Easter holidays through today. The strength in the quarter was driven by our healthcare and manufacturing customers. Financial services customers outperformed our expectations. From a geographic standpoint, we saw in-line performance globally with the exception of our SMB and APJ segments. It is challenging to identify a single cause of the consumption slowdown between Easter and today. A few of our largest customers have scrutinized Snowflake costs, as they face headwinds in their own businesses. For example, some organizations have re-evaluated their data retention policies to delete stale and less valuable data. This lowers their storage bill and reduces compute costs. We've worked with a few large customers more recently on these efforts and expect these trends to continue. History has shown that price performance benefits long-term consumption. From a booking standpoint, we saw headwinds globally with the exception of our North American large enterprise segment. This is not due to competitive pressures, but because customers remain hesitant to sign large multi-year deals. Productivity is not where we want it to be, and our updated outlook reflects this. Q1 is always a challenging bookings quarter, and the current macro environment magnifies that. We are still not satisfied with our results. We will only invest in areas that yield returns. For that reason, we will prioritize existing sales resources to drive growth before we onboard new capacity. Q1 represented another quarter of continued progress on profitability. Our non-GAAP product gross margin was 77%. More favorable pricing with our cloud service providers, product improvements, scale in our public cloud data centers, and continued growth in large customer accounts will contribute to year-over-year gross margin improvements. Non-GAAP operating margin was 5%, benefiting from revenue outperformance and savings on sales and marketing spend. Our non-GAAP adjusted free cash flow margin was 46%, positively impacted by strong linearity of collections and some early collections of May receivables. We continue to have a strong cash position with $5 billion in cash, cash equivalents and short-term and long-term investments. We used approximately $192 million of our cash to repurchase approximately 1.4 million shares to date at an average price of $136. We will continue to opportunistically repurchase shares using our free cash flow. As Frank mentioned, we are acquiring Neeva. We are excited to welcome approximately 40 employees from Neeva to Snowflake, and the full impact is reflected in our outlook. Before turning to guidance, I would like to discuss the recent trends we've been observing. As I mentioned, we have seen slower-than-expected revenue growth since Easter. Contrary to last quarter, the majority of this underperformance is driven by older customers. Although we expect this to reverse, we are flowing these patterns through to the full year due to our lack of predictability and visibility of customer behavior. As a result, we're reining in costs until we see a consistent change in consumption. We are still focused on investing in efficient growth with a concentration on continuing to sign new customers, ensuring these customers are migrated quickly and successfully, leveraging our PS team and partner resources, and selling our newer solutions such as Snowpark and Streamlit to win more personas in the enterprise. We are confident that this will ultimately lead to the data cloud network effects we have laid out over the past few years. We still believe we can achieve $10 billion of product revenue in fiscal 2029 with a better margin profile than we laid out last year. Now, let's turn to guidance. For the second quarter, we expect product revenues between $620 million and $625 million, representing year-over-year growth between 33% and 34%. Turning to margins, we expect, on a non-GAAP basis, 2% operating margin. And we expect 361 million diluted weighted average shares outstanding. For the full year fiscal 2024, we expect product revenues of approximately $2.6 billion, representing year-over-year growth of approximately 34%. Turning to profitability, for the full year fiscal 2024, we expect, on a non-GAAP basis, approximately 76% product gross margin, 5% operating margin, and 26% adjusted free cash flow margin. We expect 362 million diluted weighted average shares outstanding. We will continue to prioritize hiring in product and engineering. We have slowed our hiring plan for the year, and we expect to add approximately 1,000 employees in fiscal 2024, inclusive of M&A. Lastly, we will host our Investor Day on June 27 in Las Vegas in conjunction with Snowflake Summit, our annual user conference. If you are interested in attending, please email [email protected]. With that, operator, you can now open up the line for questions.

Thank you.

Thank you very much. Frank, do you sense any connection to the cadence of hyperscaler cost optimization activity? In other words, if the AWS and Azure optimizations begin to normalize within a few quarters, do you think that Snowflake's consumption pattern and sequential growth rates would perk up around the same time, or do you look at this as a more separate kind of phenomenon? Then I have a quick follow-up.

Yes. We think that because Amazon is such a large percentage of our overall deployments that they are a good proxy; we just know from talking to them that what they experienced, we experienced as well. So, there's definitely a ripple effect because we're in the stack. So, the answer, generally speaking, is yes, we will see that. Microsoft is smaller, so they're not as predictive of our experience as AWS would be.

Okay. Then as a quick follow-up, and Mike, I'm sorry to ask you questions. It sounds like you've got a bit of a cold. But is it safe to assume that you're completely through the revenue headwinds from Graviton adoption in the warehouse builder product? I think that's the case. But I'm also curious if there are any other analogous developments on the horizon that we could be thinking about that you might have baked into guidance in the next several quarters?

Yes. We've fully migrated all of our customers in AWS to Graviton 2, and that's the bulk of where our revenue is. And I want to remind you, there's really three types of optimizations. There's the optimizations by the cloud vendors, and that's with better hardware, better performance. Then, there are the optimizations that we do regularly in our software, which improve performance and hence are cheaper for our customers. Generally, those two combined, we forecast, so there's a 5% headwind every year to our revenue associated with those. The third optimization is the one that we really saw in a few of our largest customers, with them just wanting to really change their storage retention policies. Like, one customer went from five to three years, and it's a massive petabyte and petabytes of data. So, we lose that storage revenue. But on top of that, now your queries run quicker because you're querying less amounts of data. And we are seeing more customers wanting to do that. I spoke to some of the hyperscalers; I won't say which one, but they confirm they're seeing retention policies change within their customers, wanting to archive more older data.

Yes, thank you, Mike. That's very helpful.

Hi. Thank you so much for taking my question. If you could just offer to the degree that you can, what are your customers that are going through consumption optimization telling you with respect to when that's likely to plateau and when they are likely to come back to normal consumption, if you can? Thank you so much.

Yes, Kash. I would say, look, there's - just to put a little more color on, there's optimization, which is just how do we run what we're already running more efficiently and driving a level of savings that way, but there's sort of another layer on top of that. I would call it rationalization. One of the things that we've seen happen over the last couple of quarters is that the CFO is in the business. This is sort of an expression that we use in enterprise, and they're selling is that there is a level of oversight scrutiny that's normally not there. This is not a frequent occurrence. You only see this happening in fairly severe episodes. In the beginning, it's like, 'Hey, we do smaller contracts, short-term contracts,' but then it's like, 'Hey, you're going to live within your means. Here's the amount of money you're going to spend, and you're going to make it work. And you can figure out where you're going to cut to fit into our box.' So, that's really dynamics that we've seen playing out there. Now in terms of your question, when is this all going to be over? These things do run their course because, in the end, we're settling in. I said in my prepared remarks, things are unsettled, but eventually they will settle. We will settle into new patterns, and then we sort of resume from there. But as of right now, I think things are still unsettled, and people are adjusting, and we don't have real strong visibility in terms of 'Okay, when is it all going to be different?'

Thank you so much.

Excellent. Thank you, gentlemen, for taking the question. Mike, this one's for you. And it might be a little unfair, but it's the one that I'm getting most from investors, and it's about kind of guidance methodology and if anything's changed in that. We've seen the forward forecast have to come down a couple of times over the past couple of quarters. And there's a lot of moving pieces in both the macro environment and kind of how your customers are acting. How can we give investors confidence that this is the last cut that we're not going to be running into new types of optimization on a go-forward basis and further taking down our forecast for the fiscal year?

There is no change in our forecast methodology, and we forecast by looking at consumption trends on a daily basis, literally four weeks prior to the earnings up to yesterday. What was unique in the past is that we saw four weeks in April where there was no week-over-week growth, or it was not material. We believe this was largely driven by some customers making significant optimizations in their storage retention policies. In a consumption model, customers have the flexibility to reduce their usage or increase it as their business confidence grows. I can only provide guidance based on the data we have available.

Got it. Thank you.

Thanks for the question. I have one financial question and then a technical one. How do you see the balance between growth headwinds from optimization versus rationalization? In other words, how does the tendency to reduce spending compare to the need to invest more? Has this balance changed over the past six to nine months? From a technical standpoint, what advantages does Neeva offer? Why is it significant, and what benefits does it provide for your customers in relation to generative AI?

In terms of what customers are doing, the number of queries grew 57% year-over-year in the quarter, which is outpacing our revenue. The queries are running more efficiently due to some optimizations. Reducing the amount of storage used for running queries allows them to execute faster. Additionally, we are benefiting from the full implementation of Graviton 2 this year compared to last year. As a result, the growth in the number of jobs is outpacing revenue, and we are becoming significantly more efficient for our customers. Regarding Neeva, we'll go to Christian, who is here.

Yes. So, hi, Christian here. The broad vision that we communicated to all of you over the last several years is that Snowflake is on a mission to extend its capabilities so we can bring computation to happen close to the data. It has evolved us into an application platform. A core use case for applications is not only search and search-enabled experiences, but with the advent of generative AI is the notion of conversational experiences. The folks from Neeva are the ones that are going to power or help us accelerate the efforts around Snowflake as a platform for search and conversational experiences, but most important, within the security perimeter of Snowflake with the customers' data so that they can leverage all this new innovation and technology while ensuring the safety and privacy of the data.

Understood.

Yes. Thank you. Mike, I hope you feel better soon. I have a quick question. Last quarter, we discussed the newer cohort expanding slightly at a slower pace compared to the more established one. Have you seen any change in momentum there? Or are we still observing slower expansion from the newer cohorts and more optimization from the older ones? Are those the two factors at play, or are there other factors influencing this?

No, good question. The newer ones are growing faster. The older, obviously, are the larger dollars. So, when they do optimizations that have a bigger impact. It's interesting to the net revenue retention growth within AWS; those customers are materially above where the overall company is, and that's because we're relatively new to that. The Azure Cloud is really starting to take off for us as well.

Okay. And then, one, maybe to help you with your voice, for Frank. But Frank, if you think about the changes in policy in terms of storage retention and stuff like that, there was a reason why people stored their data for a certain number of years, etc. Do you think that what you're seeing now is more of a temporary thing? Or do you think that's kind of the permanent move that's happening here? Thank you.

I don't think it's permanent. Look, like I said, the CFOs in the business are giving very direct guidance in terms of 'Here's where you need to be,' then the operating teams are starting to look at, 'Okay, how do we implement this?' Sometimes the low-hanging fruit is we'll just cut the data back. The processes might actually not be running as well. So, there is actually cost. But you know what, the cost concern is prevailing at the moment because of the general sentiment that we are. In 2020 and 2021, there was growth at all costs, and the mentality was let it rip. Now we're in the complete inverse of that situation. We know where we have strong certainty, predictability on cost, and so on. I don't think that will last. We're just on the other side of the spectrum right now, and we will reconvert to the mean at some point here.

Okay. Makes sense. Thank you.

Great. Thanks so much for the question. Mike, I know in a consumption model, obviously, it's difficult to predict the number of new workloads and transaction volumes; a lot of that we know is tied to macro. I just wanted to come back to the optimization topic. You talked about the three different types of optimization. Is there any way you can compare your total customer portfolio to the most optimized customer that you have just to get a sense of maybe what the downside is if everyone were optimized as your most optimized customer?

That would be so hard to do. I don't have that data. Each customer is different.

Can I ask you a question? Please go ahead. Sorry.

No, I was going to add that in certain instances, some of these optimizations in the third category that Mike described and what Frank was alluding to, it's changing how the business thinks about their needs. So when we made the decision to re-evaluate our storage policy, there's a business impact that only customers can execute; hence it is difficult for us to estimate that type of decision.

Thank you, Christian. It's helpful. Mike, a question I know you do have the answer to. Since you forecast the trends every week, any commentary on how May looks relative to April?

That's reflected in the guidance I provided, the $2.6 billion for the year. There were a few strong periods in May, but overall, it's acceptable. However, it's not at the level we aim for, and that is captured in the guidance now.

Okay. Thanks. Mike, if I could just build on Brad's line of questioning, the spirit of it is what assumptions you're embedding in your second half guidance. Are you essentially reflecting the April, May environment you saw and straight lining it, or are you taking a little bit more of a conservative approach and sort of haircutting that assumes that it or maybe the financial services vertical gets a little bit weaker? That's question number one. Question number two, maybe this is best suited for Frank. Frank, Mike mentioned in his comments that sales productivity is not where Snowflake wanted it to be. Could you elaborate on that? Because that sounds like some of the pressure may not be entirely macro, but might be sales execution. So, I'd love to hear a little bit. If I interpreted that correctly? And the steps you're taking maybe to turn it around? Thank you.

Sorry, Karl. We are expecting that there will be week-over-week growth on average with our customers that will compound, but it's at a much lower pace than it was prior. What's been apparent in the last four weeks is what we're expecting inside there. I'm not expecting a straight line from where we are today at the end of the year.

Yes. On the sales productivity side, I do think that's very much a macro thing. There comes a point where you can't push any harder. We have applied the resources, but we're not converting on the resources in a way that we think is optimal. So is there an execution aspect? There always is; that's just day-to-day sales management. But in all the years of doing this kind of work, I've felt like I've always sort of under applied the resource. In hindsight, I thought that I could have done more. This is definitely a situation where I feel like we have applied tremendous amounts of resource and we've been very, very successful at it. Still, there comes a point where we need to become more selective and more prioritized on driving performance. I definitely think it's a macro thing. The sentiment out there is of a sort that you just can't push us any harder than up to a certain point.

Okay. Thank you both.

I thought that I could have done more. This is definitely a situation where I feel like we have applied tremendous amounts of resource and we've been very, very successful at it. Still, there comes a point where we need to become more selective and prioritized on driving performance. I definitely think it's a macro thing. The sentiment out there is such that you just can't push us any harder than up to a certain point. Thank you both.

We can't hear you, Pat.

If I remember right, Blue Yonder is JDA, and as i2 and Manugistics. So, anything about why that's so interesting would be great.

Look, I have a long-term fascination with supply chain management because supply chain management has never really been platformed in terms of software; it's an email spreadsheet operation. It's incredibly inefficient, and it's an incredibly high volume opportunity. The reason that it couldn't be platformed is, first of all, each supply chain is different. So, it's tough to have a standard solution for something that is so variable. Secondly, there is the data problem. If you can't establish visibility across all the entities that make up the supply chain, you stand no chance of solving that problem. The reason that I find it so interesting for Snowflake is that all the entities in the supply chain will become Snowflake accounts because that's how everybody will have visibility to everybody else. We have a real fighting chance of solving it. Secondly, the processes that run in supply chain management are extremely computationally intensive and they run in very high volume. Of course, Snowflake is ideally suited for taking on those kinds of workloads. I really think supply chain management will be the most networked segment of all industries that we're operating in. Today, the most networked segment that we're running in is financial services by far, but I think it will be overtaken by manufacturing and retail in the fullness of time. There's absolutely no penetration right there; these are unsolved problems very much in almost the history of computing. That's how serious that is. So, it's a fantastic historical opportunity for the technology to address.

Okay. Thank you.

Yes, thanks very much. Frank, with sort of the explosion in questions around AI over the last six months, do you think that buyers or executives are tying the opportunities with AI to the data yet? Meaning, I know conceptually, they might get that, but are any of your conversations with customers sort of starting to percolate because of AI and the need to get your data sorted out to take advantage of that? Or are the most people still sort of in the discovery phase on that front? Then, Mike, can you just talk if Neeva impacts the op margin guidance for the full year at all? I was just kind of curious; you've mentioned savings, but margins are sort of flattish year-over-year. I was just kind of curious if that had any impact. Thanks, guys.

It's Frank. Obviously, customers make the connection between data and the ability to take advantage of the large language models and the natural language interface and all that kind of stuff, and that's already happening. The services that are today available on Snowflake are also available in the AI space; you can already rig things together and make some interesting progress. But the thing is you need to have highly-curated, highly-optimized data. That is what we do at Snowflake to really power these models. You cannot just indiscriminately let these things loose on data that people don't understand in terms of its quality, its definition, its lineage, and all these kinds of things. I think we are in a really great place. I said in the prepared remarks, data has a gravitational pull. We will attract tremendous demand for these types of workloads, and our strategy is to enable that to the maximum extent possible.

Regarding Neeva, Kirk, that is completely included in the guidance. They have a team of highly experienced engineers, all located in the U.S., and they come at a high cost.

Thanks, Frank. The idea of making Snow available to everyone with a simple chat-like GPT interface in front of the Snowflake data is key to reaching the broader market. How long do you believe it will take before this is widely adopted beyond just power users within our organization? When do you think we can start seeing this on everyone’s desktops?

I believe that the more straightforward approach, which is perhaps not the exact term, involves utilizing Salesforce data within Snowflake, something we are currently doing internally, and will be widely available in the second half. People will appreciate this, as will I, and I prefer it significantly over using traditional dashboards because it allows me to ask questions. These are relatively simple inquiries. The complexity arises when you begin to pose much more challenging questions; that's when the limitations of these technologies become apparent. We are still in the early, exploratory stages of developing this technology. The content generation aspect is intriguing and engaging for users. While tough analytical questions might traditionally take weeks or months to solve, this software can provide answers in seconds, enhancing productivity. We are currently at the peak of the hype cycle; the true work is just beginning.

In the sales organization, we're only doing backfills right now, and we will look at performance management and upgrading people. We could reallocate heads from one region that's underperforming to another region, but no net new hires. Sorry.

Great. I hope you feel better.

Yes. Sorry.

All right. Thank you for taking the questions. You mentioned the change in data retention as a more prevalent form of optimization recently. What about the refresh rate? Are you seeing customers pull back on the frequency with which the data are updating?

No, Christian here. We have not seen changes there. If anything, because of our cost model, the economics are fairly similar if people are updating more versus less frequently or reasonably similar, and we don't see changes in the patterns.

All right. That's helpful. Thanks, Christian. And then just a follow-up on Neeva, I guess, either for you or for Frank. So, I think of the technology as fairly horizontal in terms of potential appeal. I'm just wondering if you think this can be an avenue to help land new enterprise customers going forward? And then, secondly, how much of a value-add do you think that this can truly provide to the installed base? Thanks.

This is Frank. We view search and chat as a complete evolution under the influence of AI of our relationship with data and how we interact with it. I think most of us remember when search first became available how that dramatically changed our relationship with data. I'm personally a search junkie; I can't leave it alone. I find it incredibly empowering. The problem with search has been that it matches its own strength; it has zero context. It's not stateful. Now we have the technology to make search incredibly powerful; to the point that when it can't find it, it can actually generate the code to answer the questions that are posted in search. This is incredibly important to what we said from the beginning: Snowflake is about mobilizing the world's data, and this is how we're going to do it. Search and chats are sort of morphing into a single natural language interface. But the other thing I would caution you is that this is not at all about what the natural language interfaces are. A lot of the intelligence we're talking about is going to be manifested through the interfaces, not just through natural language.

Okay. Thank you.

Great. Thanks very much. Mike, I hate to pose this to you, but you're probably the best to answer it. Beyond the week-to-week usage patterns in the installed base, are there any other operational data metrics that you're looking at to give you confidence on when NRR will bottom?

I consider various factors such as pipeline generation and weighted pipeline, typically looking out three to four quarters. I participate in sales calls every Monday and spend a lot of time with representatives discussing their accounts. The most crucial aspect is that current consumption patterns are the best indicator of future performance. We are also monitoring new products that may be released, which, while difficult to predict, provide a degree of confidence. We have significant announcements coming soon, with Streamlit being one of them. We previously mentioned Applica in Private Preview, but we believe Streamlit will have a meaningful impact. Currently, we are pleased with the daily credit consumption we are observing with Snowpark.

That's great. Thanks very much.

Yes, thanks for taking the question. I'll pose this to Frank to give Mike a break there. But just on Microsoft, so obviously, they're hosting their Build conference this week and a ton of new product announcements, including in data and analytics. But I wanted to ask you more about the partnership front. I think you commented on just seeing some better traction there. I think they've evolved their partner program, including adding you as a Tier 1 partner. Could you just talk about kind of the status of that relationship? How you're fitting in, given some of these announcements, like Fabric, which kind of unifies Microsoft's own products, but just the status quo on that relationship and the opportunity with this new partnership?

Yes. Our relationship with Microsoft has been growing at a faster pace compared to the other two cloud platforms we work with. It is clear that Microsoft views Azure as a platform rather than as a single integrated proprietary stack. They consistently emphasize their commitment to choice and innovation. While we will compete with Microsoft from the very beginning, we've experienced significant success for a variety of reasons. People continue to choose us, and we anticipate this trend will continue, which is beneficial for everyone as better products lead to more choices. The positive aspect is that our relationship has reached a level of maturity, so when issues arise or when people don’t adhere to the guidelines, we have well-established processes for addressing and resolving those matters. This is crucial as we move past earlier, more chaotic stages of operations. I have confidence that this trajectory will persist, and Azure will keep growing and outpacing the other platforms.

Great. And on Snowpark, it sounded like you're pleased with the consumption this quarter. Could you just give us a sense of expectations on the revenue ramp there? What are the big use cases you're seeing today? Is it Hadoop migrations, or data engineering? Just give us a sense of how you're expecting that ramp-up and what are the main use cases driving that?

Yes. Here's the important thing to understand about Snowpark. Snowpark is the programmability platform for Snowflake. Originally, I know Snowflake was conceived with SQL interfaces, and that was the mode through which you would address the platform. This has really opened up a whole host of modalities, if you will, onto the platform. Our posture is, look, if it reads or writes to Snowflake, we want to own these processes. Snowpark is the platform to achieve that. Now, the supply chain, if you will, how the data comes into Snowflake is through data engineering processes. Often, these are Spark workloads and processes. We think they ought to run on Snowpark. They're going to be cheaper, they're going to be faster, they're going to be operationally simpler, and they're going to be fully governed. We think if you are a Snowflake customer and you're not running these processes on Snowpark, you're just missing out on all four dimensions that I just listed. On the consumption end, it's the same thing. If you're doing analytics, data science, machine learning, AI, if it reads from and writes back to Snowflake, we think that's Snowpark. We have taken a very emphatic posture to this. We're campaigning Snowpark very, very hard around the world. The interest is tremendously high. As I said in the prepared remarks, we went from 20% in one quarter to 30% of our customers using it on at least a weekly basis. We think that's going to go to 100%. I think Snowpark will become extremely prevalent around the use of Snowflake. Beyond that, there's a whole wide world that we're obviously also very interested in, and we're going to start at home and own everything we can over there.

Thank you.

Good afternoon. Frank, maybe for you. I totally get the current cost concerns and optimization efforts underway. I'd be more curious to hear what you think could get us out of the current slowdown. Are there products or workloads that you would flag as the key ones to watch that drives the re-acceleration of the business? Just thinking through what's in your control? Or do you think we have to wait for further macro to improve? Thanks.

I guess the number one issue is sentiment out there, just the lack of visibility, the anxiety. Watching news all day doesn't give you any hope. That's absolutely number one. What we're seeing is that when we’re dealing with CTOs and Chief Data Officers, these people are chomping at the bit, but they are getting stomped. As I said earlier, by the CFO being in the business and saying, well, I guess that's all good and well. But here's how much you're going to spend. You know you're going to get a new contract. You're going to live within the confines of the contract that you have. This is really artificially constraining the demand because of the general anxiety that exists in the economy. These things run their course. We've been through these episodes before. I think that's really the requirement. There's plenty of demand out there, absolutely. AI right now is going to drive a whole other vector in terms of workload development. It's going to be hard to stop, CFOs or no CFO.

Very helpful there. And then, Christian, I wanted to follow up on Neeva. I completely understand the acquisition of Streamlit. However, I find it a bit more challenging to fully grasp the details of Neeva. As you assess Neeva and its technology, what aspects stood out the most to you? Was it the team? Is there a unique search engine technology involved? Is it related to their expertise in large language models? What motivated the acquisition of Neeva?

Yes, it's a great question. It's the combination of traditional search technology with LLM technology. I think most of us have seen numerous demos of people that take an LLM in a couple of days or hours, produce something that looks good, but then there are problems on how precise that search is and how reliable those results are. The Neeva team did extremely well at combining LLM and generative AI-type technology with traditional technology to provide attributional results. It's very interesting in an enterprise setting where you want more precise answers. That combination was very appealing. And then, of course, it is a world-class team. The combination of those two aspects was appealing to us.

Helpful. Thank you.

Great. Thanks. I wanted to ask about the competitive and the pricing environment out there. I guess on the competitive side, have you guys seen any change in win rates or workload shifts to different platforms? And when it comes to pricing, you talked about customers focusing a lot on cost savings. How is this translating into your ability to hold kind of unit pricing, especially on renewals?

Frankly, I'll let Mike weigh in once he stops coughing. The thing about pricing is, look, physics are physics. A read is a read; a write is a write. Economics cost a certain amount of money, right? There's just not that much room other than playing games or temporarily sponsoring different parts of the business to really get a sustained pricing edge on one player or another. We're all converging to very similar economics. Where you see huge differences is in the total cost of ownership, and that is not the cost of computing and storage. That's the cost of running that technology. This is where Snowflake has a huge advantage, and our customers know that. It's just reduced skill sets, far fewer people, not having to touch the complexity of the underlying platforms. We're more descendants of Apple and Tesla than being descendants of Hadoop, like some others in the marketplace. We've really abstracted the complexity, and that's what generates this TCO advantage. But the raw cost of computing and storage, there's not that much opportunity to be understated.

I want to add something to highlight what Frank mentioned in his Snowpark answer, which is that we're seeing relative to competitive platforms, Spark and by Spark, we're seeing Snowpark providing only better performance but also better price performance. So, interestingly enough, we see customers giving us technical wins and wanting to migrate because of the better economics of the competitive dynamics.

Great. And if I could squeeze one more in. Just in terms of LLMs, you guys are obviously sitting on a lot of data to be mined and training models. Do you envision kind of building up GPU clusters and offering training and inference on your platform? Or do you think that's really the place for hyperscalers to be doing that?

We're doing all of it. We alluded in the prepared remarks to Applica, which is a multi-model collection of models being built at Snowflake that requires GPUs. We're doing our part, and we're also working, and we'll show more at our conference on how we surface GPU resources into the platform. So all of the above is an important component of this generative AI wave of innovation.

Okay. Thanks. Mike, feel better.

Thank you. Our next question is from an analyst with Scotiabank.

Hi, thanks.

We're having a problem here.

Yes, sorry about that, Mike. It's hard to hear the operator.

Yes.

Just wondering, you've called out financial services as your largest vertical. Wondering how much of an impact that vertical had on the consumption patterns that you pointed out post the Easter holiday?

I guess the financial services vertical is doing fine. It was very strong for us. It's still 23% of our revenue and growing quite fast. It was in some of the other areas with some of our bigger customers outside of financial services.

No, the operator is fading. I would agree. I appreciate you sneaking me in. Just going back to the revised guidance. It suggests growth falls below 30%, but you did mention confidence still in the longer-term $10 billion target. Could you just spend some time on what you're hearing from customers that drives confidence around what you're seeing is temporary, which suggests growth bounces back? What about the bookings commentary? It sounded like North America large enterprise is the area that's standing out favorably. I just want to make sure we have the right context there and if there's anything else you can add on what's driving that. It's appreciated. Thank you.

What I would say is we have a lot of customers. We have only moved a fraction of their data that we know they have multi-year plans to go on Snowflake, and that's what gives us the confidence as well as the pipeline of deals. I'm not just talking about the pipe now; there's deals for next year that I know their long sales cycles, these big customers. That's what gives us the pipeline on top of a lot of the new products we have coming out over the next couple of years.

That will be the last question. Thank you for your time and your participation. That concludes the conference call. You may now disconnect your line.