Snowflake (SNOW) 29 Nov 232024 Q3 Earnings call transcript

Good afternoon. Thank you for attending today's Q3 FY 2024 Snowflake's Earnings Conference Call. My name is Tina, and I will be your moderator for today's call. [Operator Instructions] I would now like to pass the conference over to our host, Jimmy Sexton, Head of Investor Relations at Snowflake.

You may go ahead.

you Call. us thank Fiscal Snowflake's and Earnings XXXX afternoon, for on joining QX Good in Montana, Chief me of Q&A Executive Christian With session. Bozeman, our Product, Officer; and our Kleinerman, us Mike Chief who Chairman will Scarpelli, for Vice our Senior Financial join and are Officer; President Frank the Slootman,

During third for XXXX XXXX. fiscal our guidance quarter the year we quarter and review full financial our will call, of today's results the and fourth for discuss fiscal

presentation, forward-looking make expected future will release the will stock risks XX, actual from not including and them statements distributed duty our business, the we reliance today financial the to the call, concerning today's and Information products prospects. results. our statements, related to are cause subject to our filed long-term or forward-looking risks performance future including differ undue overall available close During XX-Q we Form purposes as comparisons. use of statements recently evaluate posted a Form quarter financial caution the now obligation you in repurchase call, financial SEC. We today events update statements to is the non-GAAP quarter any those will these with earnings place which of to a We in of filings, on program SEC growth, investor our and as refer To internally future materially distributed point I of and measures expectations. We'd new These to for press release earlier could the ended fiscal website. With or measures, addition earnings and today's result presented with uncertainties, and results, undertake to strategy, XXXX GAAP. features, financial replay press and that call like a our after Non-GAAP forward-looking and the market over that XX-Q in means file investors.snowflake.com. no like operational period-to-period ended at for accordance financial on Frank. most information, to not GAAP and our XXXX, July which measures calculated financial non-GAAP A that, report these and measures out we also in see as are turn today's our decision-making for of for in call changes to be and are statements to please posted substitute non-GAAP reconciliations to also both will on results. would our

the Snowflake of customer of the up added these data a in strategy conversations, preparation flow have reach XX mainstream product million broadly revenue a million. customers at AI sequentially. Generative Thanks, The Jimmy. which our Non-GAAP crowd $XXX Welcome We of strong reflect to We adopters. adjusted grew in during drives revenue X% industries revenue new we're macro of stabilizing top early to year-over-year and Snowflake's strategy. newer strategy customers grew plus as growth. free from data million, good and highly renewed natives, afternoon. of representing Snowflake's in $X intelligence QX of compared XX% many execution ever is now and year-over-year cash use AI that the environment. without Container enterprises as AI. mix by direct many while geographies, diverse our forefront the it the turn share technical access quarter, summer, the an serves in on derives no there's was This, translation, digital Results larger $XXX institutions. engineers results without a terms AI machine company the the said which is XX X times, technologies. Cortex emphasis global second quality in data made We've also Services, and Last let pillar service AI introduced mobilize fractured sentiment of of inferencing the This functionality can critical. specialized Snowflake we that highly operations Snowpark for AI managed all Snowflake. large functions. Business challenges any line. landscape. strategy. of are enablement data, opens can a underpinning Meanwhile, models introduced and using leverage hardware than library in aiming More language of services is many container learning governance plethora Developers showcased more with models. boundary vector on flexible has to and technologies the is XX access waiting analysts announced any language, up Snowflake. customers hence inside and already preview in and AI for Cortex of to and new customers

Snowflake the makes and is the case use common AI easy advanced use possible. case is

of are on governance scope our based and enterprise data are adoption hurdles AI the AI and framework. We well cloud ML. programmability positioned There of and scale challenging for

year first customers ago million-plus $X sharing The data sharing and XX% up our customer its is of quality from Snowflake year this data. a architecture. XX% XX% access ago. are share from XX% all broad addresses through challenge data, up to a data of our

and and and model across on. AI Horizon privileges queries. challenge models now can governance on cases, user are the data as only large be including for Snowflake as ad and strictly consistently enforces offers unified applications, models for a enterprise ML Horizon solution another hoc language Security and present adoption data they governance AI. security use traditional smart trained built of AI

data we data Snowflake universal the that Horizon, accounts exists the search, their in Customers discover As now across cloud. of which enables customers marketplace. and part and search can metadata introduced to

up since was win up process of use year-over-year. XXx data Snowflake of its in October. to customers to Over October year. unstructured data XXX% workloads outside of continues XX% new unstructured consumption quarter-over-quarter. Consumption XX% Snowpark's traditional. last grew was in Snowflake Consumption over

Tables entered preview are and public expectations. customers outpacing Our year. this the capability, is streaming Dynamic X,XXX adoption initial newest Approximately earlier feature using

capabilities number in new have available becoming QX. We a broadly of major

preview. go Services substantial These native public expansion Tables will transaction enter new Our Iceberg unlock Container UniStore products all GA, apps framework workload Apache Snowpark for processing, opportunities. and will

audience. to cities This We XX World from Tour events are campaigning Data year. reached Cloud double to traveled nearly In-person our our fall, XX,XXX last worldwide. these at attendance globally expand

Conference anticipate Mike. Build focused in you turn December, pipelines early on Build XX,XXX I Developer our up Next apps, AI/ML there. With the building see call over data to we to that, where hope and registrations. will is We workflows. is

you, Thank In strong a QX, we saw customers. Frank. consumption of from base broad

Our performance largest expected. and as large growth in between was split evenly stabilized QX. small customer Migrations accounts drove

both workloads customers fastest-growing a legacy X are vendor. Our migrating who from

with in than product and on their of X in One XX-month customers on revenue trailing than $X other their accounts the platform. customers expectations. million quarter. Growth is eighth in September the the exceeded year the platform, X in second We more million added more $XX in with year these

XX% grow the For faster years. QX month than any October. billion. the as X XX we to other strong in net of in in drove new growth be and expect to weeks, a XX% grew in billion represented consumption $X.X quarter period revenue the X next obligations bookings. bookings continued in $X.X grew Of RPO, to year-over-year execution. recognized for performance months. Consumption APJ past SMB Remaining the

progress We are significant in margin expansion. delivering making

from have approximately was contributed margin year-over-year. to expansion. Non-GAAP XXX terms Improved XX% points of cloud providers the basis up service product margin gross

Non-GAAP from credit consumption hiring of benefiting flow operating increasing per from of also adjusted outperformance increased of of scrutiny. free price X% XX%, XX% margin In We revenue margin margin higher-priced ahead and of collections. expectations. was Snowflake. year-over-year. Non-GAAP favorable increased from QX, cash benefited benefit was additions Operating timing

short-term the with ended We investments. and in billion cash cash, quarter $X.X equivalents and long-term

shares. X.X an used Our million million shares of QX, us In $XXX.X. Year-to-date, to to we $XXX million price million at position average we strong cash to $XXX repurchase allows opportunistically repurchase used shares. have repurchase X

Now let's to turn guidance.

we the representing guidance have increasing million, approximately our For between full year year-over-year representing between to growth. XX% trends $X.XX product fourth and Consumption growth improved. XX%. XX% $XXX expect revenue $XXX billion, and quarter, million We're year-over-year

expansion are stability patterns. We in seeing customer

assumes and guidance on is of consumption. based Our continued stability observed patterns

we the of X% and margin average shares expect fourth XXX quarter, diluted weighted outstanding. million For operating

associated For products. now expect of the non-GAAP new product margin full gross We we the expect fourth quarter our gross headwind XX%. with non-GAAP year, margin margin We increasing guidance. gross still in a are product

our We are operating now fiscal X%. non-GAAP non-GAAP We operating guidance. margin of XXXX increasing margin expect

fiscal cash non-GAAP XXXX expect are non-GAAP free adjusted now flow free We our margin XX%. margin. We of adjusted flow increasing cash

year, full XXX shares million. the expect diluted outstanding of we weighted For average

We are year, X,XXX more for this the up open inclusive now track can you than employees to line on operator, of add M&A. With questions. that,

Instructions] from Our question the with [Operator is first line Mark JPMorgan. of Murphy

XX-minute the love earnings result. I and call, on a fantastic congrats

stopping some for spill and data find might they're Snowflake you the hearing of generative So and perhaps are customer like estates on they their good edge products or tangible conversations aren't because starting that shape. and Are Frank, in at other AI we concept that more projects be around that Snowpark of drive pipeline there that AI over uplift into sensing data, are broadly your more and sharing? Snowpipes enough conversations with seeing front data

they're showcasing now speaking, all of run Generally things productively can things is technologies, because have them the well-governed used advantage now that to turn the newer estates getting can also and with to with data yes, frameworks get where delivering preoccupied that which from are they their they just have it are the into X take we they customers shape them and interesting where of to Snowflake. on

platforms. And on of they interest in in of access able also are to people optimization data the curation and degree not their knock-on looking true the data but what it's be data of the So organization have a also that at frenzy definitely quality the the to the strategy, and high effect data has need to. of of interest kind just data AI

a as So taking terms of well people what what's data it of should it. view in are be in and in estates much their broader

Stanley. from with Keith question Weiss next Our Morgan is

in come off sort for with on optimization, how you on products? yet maybe level, color, you into question migrations versus more we've looking of as account coming from your in of us any kind peers call, to leaning of mentioned slow sort And us sort of [indiscernible] times more stabilization can like what idea are obviously hyperscaler the new the just much heard who that Maybe you -- way what year? and Keith. are a at that into existing from sort leaning could of give parts your When the you're is new on next already several seeing playing of an growth. that there more start an you cost the versus well. give alleviating out are quick stabilization customers the which with customers scrutiny Is

the typically the of unbridled much institutions things, evolved who larger to and What mentioned recent years in X that, and year is reset and resets enterprises have and prone Keith, then Because it has are to these is that base quarter. in to saw I rationalize. expansion and drastic later incremental that our over customer they each of earlier Frank. not less optimization to with getting the is less include we consumption exposure

themselves rationalized have place through forward. good and now people are driven a themselves and in Secondly, processes really to these move

You are to things get on point, us interesting, can only much. it's preparing ML At rationalize some and optimize new enabling and tired AI diminishing of for technologies. returns. moving People they're namely so that now and and

of why add too, customers that see stabilization as out other quarterly shift, XX our top grew to a are all Frank that, that seeing make mentioned. is we And the point I'll we I'll is sequentially. X

enterprises scrutinized always have we're enterprises always mature and seeing are customers biggest will. mature now They cost. Our

the nothing And business. that's so continue. will just run there, there's new and a way that anyone should And

you sales X of some seems maybe flat sort and are efficiencies X it's only then regions healthy. light And terms question like leaning quarter. again, investing quick of you helpful. more Because your But marketing or it's areas headcount. maybe what into in your in then this That's noticed basically number. growth getting still? Obviously, any versus that more obviously, of you're on pretty sort shed Can it. quarter-over-quarter we Got

So just would you're be where on helpful. investing, color some

Yes.

there. beginning even are is So say can those in particular, heads, We in so in very so would first our general, ago of marketing with involved the are so, sales added people of marketing function, heavily QX at to a of in all, added at most kickoff. invest our sales more or end we and And about or QX, get X right I continuing we leader sales what of and in new months talked Europe. typically the

people have invest some We we continuing to people investing, go underperformers the forward. we're and another to, that changing right to more is APJ continue and prune will X and and in. globally out invest we been we in as continue And

Our is from next Lenschow line Raimo question with the of Barclays.

well. as me from Congrats

areas forward? the about data products going different to lots exciting approach your going sales Does market? all it forward more address products you You kind sales to new going expansion how do of like that And finding crazy of forward? out. sound change, think classic you have like How new want to a you're be do are setup here here mean going of are have doesn't amount here. actually and needs these Because AI. Frank, sales a going coming do from like crazy to that able to now that ensure force Like the to you

there's -- excellent actually an Yes, actually observation. that's

understood. and very We dominant XXXX, us changed. as had the It market lanes has really have X has selling and we deployed very you defined historically sort served well were Snowflake observed, really anywhere. well. go sort in swam that everywhere narrowly of to that back you swim motion When But of correctly

AM of our broad-based all the that market, coming recent in directions. framework, in AI been onslaught delivering in whole absolute as these enable you've that I programmability we've working hard, we're in fruition. applications mega very and capabilities, of and the capabilities the very, These incredibly are towards in ton seen are of these And all just capable Now our to very platforms mean, to and right? many an drive is platforms directions. different years,

ton Now all irons and they're the getting a hot. now of have we fire, in

person we So and these because that be to standpoint, a from going world in these much is all a the specialization be all sales is to over happening all impossible disciplines. expert and for technologies literally single it going have more to

of wherever we So that groups and sort to skills, then have if core of we're you they're teams have people purpose capabilities, specialists that general have will, augment going will needed. these will

and motions selling rapidly year. basic they evolve are our coming supported in So how will the

question the Evercore Kirk ISI. line Materne is Our from of with next

all, it, some these quarter. very still I very, of the unstructured just overall of could newer part small we they to about workloads I Can on year. look you dimensionalize that think sort stabilization? at we the the turn the meaning Congrats talk that that get are is acceleration Anything assume having 'XX? has but it's into with on next data Mike, for out you you catalyst sort as guess, impact impact into at as consumption when of consumption a helping of to

what doing, to not now. that's unstructured helping It's but Snowpark I can't with us well also stabilization. more that. definitely quantify Well, they're exactly doing effects very of starting the it's for is see just we're right

new I is Some uptick really, of streaming the things early We year. we're in tables. forecast and say, Streamlit built our already seeing next would very into dynamic for

The we going that really build catalyst can do these that we for GA a think in be we just history don't because new are in us next and other products new, much preview for they're clearly, because lot a And before we need XXXX. into now will growth for public year products of that.

working year while, we and there really for it been technology out we've over Document preview. is document. acquisition ago. into to that quite returns on going Kirk. of now semi-structure on did the it basis And was a what document incredibly unstructured does a into and an is popular Document an a is follow-on, AI One That AI a

in in So it full a ton participant analytical can become interest is that. processing. of a There

of XX% into the entire analytical so which estate the sphere. data, world's And unstructured the is really brings of that data

become will driver are will we data, something data, very least focused expect in today, and at do important, be AI unstructured and our to because be So heavily on so a that LLMs that business. textual especially this

waiting opportunity on to federal that? FedRAMP. on you of know sort update can I back guys us hear Any an give on you the And stands? update where Mike, were just

because huge but literally I would upside piece, Federal it authorization have is have it's we a might a I our should for and say opportunity, us such any high thought I today. actually it's small had day. FedRAMP

tuned. stay So

You'll see soon. an very on announcement that

with line the question our Stifel. Brad Reback of is from next

Just X there? announced a think should we any of Amazon, with potential now quick Graviton question being headwinds by technical

improvements. we was and headwind our X announced improvements really benefit told a software tested X% hardware we have Graviton not hardware recently. not associated as just count before, software. with all year, Well, you that every on We

some, know. when We stay we we'll don't but and expect have It's update you we there will more information. too just tuned be early,

Wolfe next Our Alex from with is of question Zukin Research. the line

directionally, Bruck guiding, what we just guess, would is think stabilize next I expect? This nice and biggest same think if Congrats how puts we that's Ethan would for Alex when you're on and level about Zukin. in the the what be you're should through the I the consumption trends, NRR quarter. trough given we year? question assuming and And when growth is, would the My persist takes about to consumption like XQ, stabilizing to around accelerate guess, NRR, vein, expect you when trends

year the what a right next lot going -- with right guide to do So it time, biggest see is and puts with to growth. be next going of clearly, to to It size impact to dip year, stabilizing. takes to a our for we've revenue expect see now converge we'll we're and to tell we to over to seeing I'll stabilization as our and NRR, be more. said, I have too And start could guide of I'm the are now it's at do bit and not initiatives the new just that. at. the the little -- going continue NRR. we're I early

next call when stay tuned we for our So for February guidance give year.

from Jefferies. of Thill with Brent line is question the next Our

what you're Mike, you is higher the seeing you offering. higher on there? Can I extrapolate a that tiers mentioned utilization there just think you're

and see which we becoming bigger more what and from companies bigger large SKU, there. very ones being revenue using critical who using and that's enterprises business derived higher Well, for as customers. of a in we're typically, those has seeing are higher margin do are are our And result, BPS, our guys large these us, our the

Okay. just 'XX? feel think And the adoption this impact you? in 'XX? in early bigger thought hit How quickly you'll in helping back are for coming Frank, it terms When from about the think process half this tailwind of is monetization a way, do that you Do you that overall AI when does impact, your you a in see perspective? of

and not AI unfold a the is nuances business. to be all consumption you how on into like turning just translate a It's switch to going and There's our to effects the lot itself incremental going into see sudden, of of happening.

we As already data seeing the remarks, in of the of that has quality the interest consumption. bringing we're are data a said the also the the during universe in prepared then is strategy, that also which on driving data knock-on effect in AI initiatives. people expansion It's interest

necessarily be All ton incremental of even of of AI, is proverbial the happening workload though the water these that above And state are the characterize that AI, AI to prepared the is everything customers data and below actually going legacy going huge might be decades a to have advanced in you environment, that. We well where that's AI, to utilization their that it would cases, in data the workload to very from literally ML be the tackling ticking because very, to beneficiary years. but years, service optimizing that as think is a years you to AI not for are going iceberg really to Snowflake, it huge trusted the because a and our only and support estates is value prior some happen is models. but we're normally Snowflake are and to our and in of spent data super is that tip of enable then that a well and things that us bring has and these estates and platforms building sanctioned start curing when completely organizing

of Colville line is with Scotiabank. question the next Patrick Our from

far October. talked consumption you about strong month? this Can you November. very trends of remarks, through prepared talk grow We're your continued in Consumption thus XX in In September. days just now to

guidance. have What in big a consumption U.S. holiday to remember, typically with and say it said, of I happy slow the I'm has that's trends but is that With is are we're reflected Thanksgiving also week. seeing, would you in and our the week good, the a

Our the is next question Radke with of Tyler from line Citi.

to about wanted I ask you container services.

some data? that are characterize those good that curious seeing momentum I'm you're XX large of are for that, consumption kind like Snowflake models early And that? when customers private in or type do preview. what how you those customers how more uplift sounds it of on there, directly than So doing they language actually of many deploying the

for good Yes. seen of GPU silver have This is cases, prior use instances preview those Christian. and are customers of large use of Yes, a number but payer variety we a leveraging language in models. services

it's sure, the early of encouragement preview. the for we've and and a lots cases the of got rollout of continue meaningful use So excitement adoption part to

Great. And just a question on go-to-market.

So leaders key as some and year, there execution like do you've do this it this I at for? about point? the still you're the are sales in I execution guess, looking well? good roles been have those APAC. get turned results feel partially corner Or issues you guess, on And you right ultimately, place in I that's all you like you some seems issues like resolved think earlier feel to talked

and We're It's we're and That's much in the on-demand large from we're motion for more and selling and always improvement. feel the ability a here. by running running every true stronger, I said, becoming then incrementally that's terms better, process. if I've always Look, in That an deeper SMB, as to There's day. extremely been we're consistent Yes. ISB as organization of sell. to in productive work more Frank. you a all long getting customers. more been mid-range way our -- opportunity for that unattended, to it's The large salespeople sales will, to an global

of never stability very But over. things. how It's something of and a Anybody up dynamic So run company, a have to a or this deliver mix that this and business billion lot magnitude a organization quarter. lean of knows hard we on is whole, who's order $X of on show in lot every we results to, of something that's works. our rate, this as a progress. been But the those

becoming our of terms a it's go-to-market formidable So enterprise capability. in

is of Brent question next Piper Our line Bracelin Sandler. from with the

we'll Frank, here. start with you

quarter, twice. was mentioned data unstructured I last think

it times. You've than out more a dozen here called

You It talked increase October focus certainly hasn't this, trends here. out a been think, How going an about area Document I think AI you movement Clearly, the game coming supported Snowpark in change being year? up, a the data sea year-over-year. relative is as past, and particularly changer is a changer game but XXx Containers opportunity it next growth here. unstructured feels like about we of support there's a of much to forward? much How

things well bringing before. a full as really really all that and and data, is enriched the slide that been is with large Look, earlier, recently, unstructured you it's just haven't both It's can't purposes, in language market borderline It's this is to data. models Document exciting. these the And of that of become that unstructured to data that relatively unlock I dealing AI world's going of analytical textual analytical unusable reference participant going had extraordinarily types and is also data to as the on it's super it insights as through and the workloads. workloads. and Snowflake outcomes majority of until it ironic because It's of for we developed, said in

to of terms this in as computing. it driver a our type we of know this going of is world So be

time those How I slightly now to platform? Mike, And growth factors just driving at consumption there? on per if new the net the of year. over a helpful. What this first versus perspective, improved look on headwinds the average the are coming is ease in optimization metrics customer, much starting year-over-year for workloads revenue That's a

big that X, Well, numerous of of to they happen. day what the been part happen not said repeat history but is to any that's was investors, now, is at and growth are at won't about biggest has life. the I our optimizations happening Snowflake. our happen will don't of see because customers legacy time. they we're times happening optimization. within Nothing X would to I'll Snowflake I shown with They've continue new say, in say on-prem I've all ones future, talked what customers seeing growth Most we the they migrations into our from

there's seeing within migrations, expansion as we're customers too. existing but initial So well, also workload new

thing X there, seeing. there's is it. starting meaningful consumption So in that's revenue. long it's XX% driving to for it's for -- just I to It's our us. us. And That's say we're get will general a still not no but Snowpark kick of Still ways

Derrick question is next from Wood line of with the Cowen. Our TD

wanted off in I re-platform for the And guess Frank, we when did I a dovetail I cause I to on-premise it saw slowdown to that of legacy migrations. customers I looking hit, mean, guess macro on-prem think to from cloud.

But XXXX macro the accounts? are wondering new now a the Just curious, kind consumption in seeing up start pipeline what Global the if environment you're gotten back seeing improving here. legacy you stable? that And more you're I highlighting kick trends initiatives to guys I know mean, bit has little of replatforming for

of hardcore surprisingly the get on-premise that of is migrated there almost no mean, enormous sort still is legacy cloud. an I just there's our And business. still amount the to Well, workloads replatforming is that sitting waiting doubt the to of

a So had cloud, important. enabling those haven't get But of very architectures period data opportunities into I states are that Once technologies our expect said for people you what this Mike's time. to new very before. continue our now and long new is as the

workload expansion. that drives so And

constraints the possible. people AI is in programmability, is from all business. day the run in have and we're the what unlimited problem really terms X your just customers, I telling definitely we because of limit. of that. the used I that no to new going of with grows We don't like, cloud us doing workloads. technologies And foundation the was is longer from sky not of the story is is foundational be It be Snowflake to now but doing where it's So the okay, the that's and can the now what's on-premise, with technology. capacity in numbers conversation That's mean, our opportunity been and

problem budget, back your And but holding not you and things, that it those to Your Not pay are us have course, the ability for your imagination well. to It's is then, technology. our right? harness anymore. as easy is technology of

Yes, that, Yes. varies of I quarter. Global X a a campaign, and see years, at really times be sales by about XXXX too, Global we But X, on-prem first-gen I'll of as And for on to added QX. number Yes, migration have that coming them customer. many, to lot an an but a estate, happen just to Global said, still almost that first. all with It to migrate on-prem we they it's Snowflake. always them migration. over the doesn't had to of to years. not on-prem with Global X-year sometimes starts many start many last that be solution with done purchased Many of cloud said, we for going as I've also it's XXXX what With is a your want selling XXXX. Global And too, question a a XXXX to add see, into before, said X have XXXX cycle. our pipeline every in Derrick,

looking And more companies approach. a Got you, it. RPO. And pay-as-you-go heard we've a to really take for Mike, follow-on around of

shifting speak and to commits forward? could your play customers RPO growth annual multiyear trends that looking versus are what into So you approach extent to that may or how

pricing Actually, who see signed have had pretty are run get to see with time quarter if continue top of to or continue like customers a our same April of to a is to And actually to they the capacity. as is a are X-year before just going our with then X don't they they out up we they they have. expect do that QX then contract X-year going only significant of They and ones The contract, go. that. they pay that And be pay-as-you-go XX that, to I customers. renewals bookings customers for want have that I number contracts much. No. another until push something. you are do

payment credit. to per term, discount payment really, anticipate is and expect though do I want all customers that's are X said longer price lower I'd up to price And what terms, terms, per annual it It on pay want more most credit available price the pay to willing rather monthly want customers. is credit. Payment do that things advance. will to I along, you per in to people down in And that give all terms to and I've arrears and of to the comes date,

is question Joel next Truist. of with from Our the Fishbein line

announced is joint with you. I you product partnership going, an guess, And this one you I'm Frank, also Summit, that integration. curious announced just how Snowflake Microsoft. and partnership some is for expanded

about So I'd how forward? going that an to you get that's how going love and feel update

numbers measurably exact standpoint know we better This don't is super quite is The Microsoft much compensation really ticked on the that quarter. it the this Frank. Yes, super, seeing is from with of upticked field the in the platform at we're things quarter. but alignment a the Azure the platform I we that effects actually in worked And that. in And bit energy just during important renewed hand, coming a Microsoft really up. saw of the from world. our relationship

we're So sentiment in in the behavior the field, field. seeing developing we're and overall encouraged by the the encouraged in we're that's

been with than Microsoft has So healthier it's the relationship definitely historically.

pleased we're optimistic it. So about and

we milestones from partnership? should the for there be Any looking

Not really.

Our the Mike next question Needham. of from with is Cikos line

comments momentum some our of with of And around good heard consistent we've growing to which the earlier Snowpark, the checks. is of some hear in some

what or probably sizable data a Snowpark is to more Christian. start I does customer customers? time larger as Snowpark guess, these how those part for think becomes of I adoption for over about versus Snowpark think about maybe how non-Snowpark base maybe talking there the usage Mike customers' we consumption yet workflows? But empirical of a enough question Or should or is building enough customers, for Snowpark

them think roughly we a have X I done haven't I too the that consumption to which on a X will production a identified has Snowflake million their to now, still increase that -- but migrations. into are tell. those year. of which, it's see legacy I vendor. customer doing there number POCs, significantly by customer their significant we way, the we've will migration, in early say $X.X on saved right going And

to the trend quite the we consumption customers, that -- starting year. into see, expect we're some meaningful, next with continue and So will it's clearly,

with Patrick from question JMP. next Walravens is Our

Owen Pat. quarter. is for on Congrats This Hobbs on the strong

the is how between hyperscalers? have evenly there the different much X So there? kind comes share the majority Is kind from does consumption I of that it compute starting off, of or guess, split X of

XX% the No. of up AWS one, right by then GCP. far now. our Microsoft business still being fastest-growing GCP Microsoft is with XX%. to is followed our Azure X% and AWS but biggest, Azure is is by

X%. to result, as will as X is customers not said, you, more reasons And in I for as is and I expensive why do As salespeople GCP. AWS big much of And inclined to our in in GCP are our GCP really so not than Azure. much is a the it just GCP operate tell

picture, And move then well. scientist of question maybe going going as this that now the era is see the of through the of as the data sort Christian guess And bigger learning machine kind AI? into in you of I for is data, a we role you future? How where do big and changed

and of obviously We interest on both announcements for hosting see expanded language Container cases, capability Cortex. interest models we via Snowpark but on Snowflake a also use obviously generative do the new of both but actual via number ML Services, where inferencing we traditional see AI seamless science the have those to we've more And large platform. lots such Snowflake data had Yes. at and

types use our from we cases. demand see for of So broad base customer strong both

question is from next line Our of the with Moskowitz Gregg Mizuho.

incremental to be in next large as have budget of you execs, since will IT primarily company IT? speak dollars budgets for 'XX spend AI entirely on even from you a whether areas any or curious other year if hear I'm with thoughts the approved of come there AI as Frank, calendar well maybe will lot or

understanding how architectures to is that right fully words, them holding do where I and is it. comfortable enable back It really budget takes time is in same the are if they're anything In words to tech sentence. to the confident and available it. But to to don't mature the deployment hear ways in it. going AI do and involved everybody make resources other these evolve even

experimentation, iterations So these to a of you kinds and see that. and going go testing all they're compare of through things. benchmark many contrast of And many, lot

that I the but very field will proliferated also deep, foundational become and purpose. will that large specialized narrow And very, are We models as models very, well. subsector very with think in very, are many, many, many very

C-suite enterprises, field. really going to large for centers, are a enormous an that to supply example, people it So in economics, I promise, opportunities. you of looking These when pricing you big excited sort is impact and attention. will a marginal tell optimization, not management, of very talk has that we're very, for Data it's the contact steroids the that I influence always of the like, become reset in all under now chain has the are really incremental but mean, had newer about. on technologies

did X Labor strong Very helpful. though, Or really did years. time Over which that period period? consumption? Mike, encouraging, than grew faster past hear Day? for a different X X to other And that you that any weeks post weeks, the then span of consumption it course, particularly for Was see over

Post Labor Day.

of line Baird. is question last from Our Will Power with the

nicely that? X we're you use or kind behind customers, threads other Mike, there like that guess or were of the cases wonder, color of strength XX common any common Any quarter-over-quarter. top your I there? it sounds products I referenced driving

very others migration, workloads continue X all estates large to -- big Snowflake. just customers but just are the move data was continuing these to are doing to that with No, massive a

industry-specific very as well. That's

QX, couple level Okay. the of I guess, color your given And headwinds kind faster gross expect there but product margins. seems a of sounds question maybe kind forward? target, going may upside And Last might I just will, general think. at maybe you're couple on It kind of gotten quarters, like there a be to already you've surprise, you quick on any longer-term of have of the you than like there is if this it expected? in

public Yes. I upside are this going quarter our of lot is with of into don't a to preview. we're reasons. guidance for in really The our that a these see number seeing headwind big and all products, that new long-term margin

to the take amortize -- contribution we We of to out required have associated software start to new times get to have actually months introduce usually to software actually top until capitalize a the X costs amortized particular, new fine-tune to negative to products, it on now us something lot costs under the are we to that. operating Unistar, many start margins times, year of when like of to up a that takes going In half we're features. to those the costs that be with GAAP development

have that to we are that so products many quarter coming so have out just the a this And is headwind going new on margin.

well as today's for call. you session Thank question-and-answer the concludes That your as participation.

your lines. You may disconnect now