SNOW December 8 Barclays investor conference

The entire Dec 8 Barclays conference with the SNOW CFO is worth listening. I am practically pasting the whole talk in this post as I find it difficult to exclude much.

I thought it was a goldmine of very granular and some new detail into SNOW’s business, with concrete numbers exposed —unlike the usual fluff and recycled old statements spoken by executives at other company conferences (such as the kinda useless last 3 talks given by DDOG…)

  1. How does management have confidence that their large customers will eventually spend tremendous long term amounts on SNOW?
    They have visibility in the massive spending on GSIs for snowflake migrations. The top 15 alone, from Jan to Oct, sold 1.425 billion in services!

…you have to understand, too, we don’t just find a customer in the quarter and close them. The average customer is 7 months, the average across the board. But large customers are 1 to 2 years to land these customers to just get them to get that first PO to start. And these guys aren’t making decisions for the next 3 months, 6 months. These are 10- to 15-year decisions they’re making in what they’re doing…many of our large accounts, our average Global 2000 lands at $100,000. Yes, we’ve had customers that have landed at $3 million as a CAP 1, we call them, but most land at $100,000. And by the way, when they’re doing that, they also typically sign big service contracts with GSIs to do their migrations for us. And I was actually just looking yesterday, our top 15 GSIs year-to-date through October 31 is the data that I had. They actually sold $1.425 billion in services around Snowflake. And those are all customers that are just starting their journey to Snowflake…People can’t be spending that type of money if they didn’t have in their road map and their plans to move to Snowflake.

…it’s funny, you can see when we’re trying to sell into a Global 2000. And then you can see through their job postings they have people that are looking for Snowflake talent so kind of tells you we’re going to get those deals.

  1. More clarity on what was mentioned in the earnings call, about the three top 10 customers that declined sequentially in spending. Confirmation that at least one? is in crypto.

…we said on the call, and I will give you this disclosure every quarter, 6 of our top 10 customers grew faster quarter-over-quarter than the company’s growth rate. That tells you. And I did have 3 customers who – this is actually the first time that sequentially declined quarter-over-quarter. And those are 3 in the technology crypto space.

Those 3 customers, actually from when we guided at the beginning of the year, they took out $41.8 million of revenue from us for the full year. These were 3 customers who have gone through some challenging times, but also did some heavy optimization. And by the way, all these 3 customers, we’ve been telling them for a while, we can help you save money, but they were growing so fast, they didn’t pay attention and now customers are paying attention. But I would say those were 3 of our worst in terms of how big they were and we knew there were big savings.

  1. Some more information on the competitive landscape for on-prem migrations.
    Google is the toughest competitor.
    Microsoft Synapse doesn’t seem capable of keeping customers!
    As for Amazon, SNOW is now the LARGEST independent software vendor for AWS (!!). And, Amazon Redshift is really not able to win in the large customer deals because they simply can’t operate at a large scale like SNOW.
    It’s also mentioned that SNOW beats out 9 times of 10 on price-performance with POCs

When we are competing for an on-premise migration, always we’re competing with Google, Microsoft. AWS tends to partner with us more out of the gate. Google is definitely the most competitive there. Many of the times, we only win in an Azure account when Synapse fails. And I’m not saying there isn’t any, but I have been asking my salespeople for a successful implementation of Synapse in a large account, and they can’t find one. I have a lot of companies we’ve heard trying, but generally, we get in there once they fail. I’m sure they must have some, but we can’t find any.

We are actually AWS’, I think we are their now biggest ISV. We are becoming very meaningful to AWS. We just had to reinvent and Frank and me, and they’re leaning in heavier on us, and we’re looking at renegotiating our contract with them right now.
And it’s not just about pricing. It’s about more commitments from them to work with us.
So as I said, AWS pretty much partners with us day 1. They don’t try to compete.
Yes, they still compete in the smaller accounts with Redshift, and that’s a pretty good business for them. But within the large accounts to get all the data in one place, Redshift just can’t do that…We can operate at scale [unlike Redshift].

…When we actually sit down with customers and look at real data not benchmark data, put actual workloads and do POCs, 9 times out of 10, we are winning in price performance and by a considerable margin.

  1. More detail about the seasonality as SNOW gets bigger and that we are now opened up from COVID. Reiterates, as said since day 1, that 70% workloads are scheduled vs 30% human interaction.

We constantly fine-tune our model based upon the history. And we have seen, as we’ve become more of a global company and especially what we have seen, for instance, we sell to a lot of U.S. companies, but they have employees around the world. They have a lot of users in India. And you actually see during the Indian holidays, and there’s a lot of holidays in India, I’ve really noticed this year, that the downtick daily consumption tied into those holidays. And so our models are all, the more history we have, the better we are at forecasting that stuff.
…And unfortunately, during COVID, it was very different when people were working remotely. People aren’t taking holidays when people are sitting at home, they were continuing to work.
…I do see seasonality in consumption. Clearly, this quarter has a lot of holidays, and around July 4th and then the summertime holidays, we do see that because remember, about – roughly about 70% of our consumption is done through scheduled jobs. No human interaction, about 30% is human interaction with the system. And you can see that on the holidays. I see it on Saturdays and Sundays how it drops off, but it’s saves that, that base roughly 70%.

  1. Just as @laneylawyer quoted in the other thread: SNOW helps companies make more money, which helps SNOW stay resilient in growth in an economic slowdown. It’s not simply about cost cutting like when selling observability products in DDOG or cybersecurity in CRWD.

…when you have challenging economic times you…be able to sell on value. And it’s not just a cost replacement. It’s what’s the value you’re driving. And data is becoming one of the core assets of most companies today, whether you’re a media advertising company or a financial institution, getting real-time data and insights into your business is a pretty important piece to companies. And we’re getting data. There’s one large oil and gas company. I know we sold them $3 million worth of Snowflake. They have consumed, and they themselves have said they see $30 million in value. My comment was we sold that too cheap.

  1. How does SNOW give guidance? As mentioned above, they gain confidence in their forecast with looking at the spending on global system integrators for Snowflake.

I’ve been saying since day 1, we try to guide the business to try to have a 3% to 5% beat, I think, it’s a good beat. And given the challenges and what’s happening in the economy right now, I’m pretty pleased with how the company has performed.

My FP&A team, I have a team of data, small team of data scientists that build the model we reforecast by customer. We kind of go down to the top 200 – I forget if 280 customers. We actually go and talk to the reps: ‘Does this make sense? What’s going on?’ I sit through a lot of the big accounts. I sit through EPCs, I read, I get – there’s all kinds of deal information I read and what customers are doing. That’s what gives me the confidence.

But I get a lot of confidence when I started looking a lot more into what people are spending with GSIs. People don’t spend that type of money unless they plan on doing something on Snowflake. And I’m digging into who are those big customers, where they are on their journey.

  1. Some more detail about gross margin expansion and growth in FCF generation.
    It’s all about getting the largest customers to spend huge amounts over time.
    That makes the S/M spend efficient, it allows usage of paid for cloud reserve capacity at scale, and permits SNOW to continually renegotiate with the hyperscalers.

I’ve said this since day 1… [gross margins] may get to 80%, but it’s not going to go above 80%. I think we can get into the high 70s. Free cash flow will continue to expand.
There’s still more room in gross margin – in the gross margins. We still – we have, I think, 32, 33 deployments around the world. We have a lot of them that are at scale. When you set up a deployment say in AWS, Singapore, we’re paying for a certain amount of free pool, we call it. That’s the reserve capacity. We’re paying day 1 before we have a customer. We have to get to a certain level before that starts to become profitable for us. We have a lot of deployments that are not at scale that will help drive margin up.
And as the bigger customers become a bigger percentage of our overall business, that are on business critical, those guys, we have better margin profile on those customers. So that will continue to drive gross margin along with renegotiating with the cloud vendors over time…, as we get into these bigger customer relationships, you’re going to get a lot of leverage out of sales and marketing. If you can have $50 million-plus customer relationships, those are very, very efficient from a sales and marketing standpoint.

We are very much focused on growth, but it’s not growth at all costs. It has to be efficient growth. I actually just had a 4-hour meeting yesterday with our – all of our executives who are all hounding that they need more money, I need more money, and it’s not unless you can justify how it’s going to drive revenue and actually see the payback you’re not getting it.

  1. The unending emphasis on the large, large, large customers. As said repeatedly, this is what separates SNOW from other high growth SaaS companies. It’s why I think SNOW deserves a higher relative premium valuation to other SaaS (even though technically, SNOW begins each quarter with zero dollars in revenue!)
    The ability to deliver such great value to customers that many of them can spend $10 million, 50 million, 100 million in contracts means you’ve got something very special in the leadership, product, and mission criticality.
    It’s how you get 160-170% in NRR, and it’s how SNOW has shown BOTH the highest YoY and highest QoQ growth during an economic slowdown for the past 2 quarters, while generating good FCF:
    SNOW 18% QoQ → 12% QoQ at >2 billion run rate (positive FCF)
    CRWD 10% QoQ → 8% QoQ at >2 billion run rate (positive FCF)
    DDOG 12% QoQ → 7.5% QoQ at 1.75 billion run rate (positive FCF)
    ZS 11% QoQ → 12% QoQ at 1.4 billion run rate (positive FCF)
    NET 10.5% QoQ → 8% QoQ at 1 billion run rate (negative FCF)

…we are very much focused on large customers. I will say, when we joined – when Frank came on to the business, I’m not going to say they [SNOW] didn’t have the large customers, they did, but they had a disproportionate amount of, what I’ll say, commercial customers. We are very much focused on the largest companies in the world. And that’s why I always tell people, I don’t even look at how many customers we have. I look at how many Global 2000, how many Fortune 500, who are the marquee accounts in certain regions, countries, verticals. That’s what I’m more concerned about the quality customers, the ones that I know can be $10 million-plus a year in revenue.

And so we’re very much focused on those large customers. I’m not saying we’re ignoring the low end of the market, when I say low end, the SMB. We have a group that focuses on those 500 and less employees in the company. I called out that we saw some weakness in those. And what I want to stress – those guys are still growing and consuming. They just are not growing at the rate we are forecasting the growing consumption.

  1. CFO reiterates 2% per year dilution guide for the long term

We’ve been running at pretty low dilution given our growth. We’ve been running around 2%. I think this year, it’s going to probably be closer to 3%. But longer term, my goal is to keep it at 2%. And clearly, you have a trade-off is do you pay people more cash and less equity or vice versa. And I’m spending a lot of time looking at all of that. But clearly, our cash balances we have – we had $4.9 billion in the bank at the end of last quarter, and that will continue to grow. While we’ve been doing M&A, we have done a quasi-buyback in terms of we now do net settlement on RSUs with people.

  1. When asked about what they plan to do with cash (besides M&A to acquire talent/technology):

I’m looking at a number of things right now, it’s earning a pretty good interest. But I’m not naive. I’ve been looking at different things with share buybacks and other things like that, but haven’t formalized anything.


Thanks JonWayne. That was an incredibly useful and informative post. I’ve copied a lot of it into my own notes on Snow. I really appreciate your contributions to the board.


I have to second Saul’s praise. That was beyond excellent, Jon.

How are you thinking about Q4? Even a 5% beat would only be 8.5% growth sequentially…I guess that’s priced in by now but it still looks a bit shocking, but maybe it’s a one-time effect of cost improvements they’ve made that save their customers money? (I love that they do this and still grow so mightily!)




My approach to SNOW isn’t so much QoQ based investing, unlike how I view all other companies. I believe Slootman said it best (quote from 2 quarter ago’s investor presentations, where I wrote a separate post SNOW: it's not for QoQ investing):

CEO: We give 10-year guidance. Why? Because we don’t want people to ‘pick the fly s*** out of the pepper’, so to speak, and you have my views of the quarter.
But say, look, is your thesis intact over the long period of time. And those are always the questions that we try and answer.
I cannot run a company on a quarterly basis. There are quarterly aspects to it. Fundamentally, almost everything that we do has a much longer time horizon. When you look at P&Ls, most of the money we spend is not related to the current period, right? And sometimes it’s hard to convey that.
I don’t really feel that encumbered by public market…if you don’t like it, you can go buy IBM. We’ll tell you what we’re doing. We’ll give you a long-term view. We hope you want to sign up for the journey over a long period of time. Those are the kinds of investors we want. But if you want to get out in 90 days and make a bunch of money, I have nothing to tell you.

To be more specific, I’m not going to be too worried about how SNOW performs next Q, when one quarter ahead is really driven by seasonality, short term macro impacts, and how SNOW acquired their customers from 1-2 years ago…a fast ramp on newer customers is not going to come from the record global 2000 lands from this recent Q3, right? I’m more interested in a truly forward outlook on how SNOW is going to do 4 more quarters from now - where they are going to see the supposed fruits of today’s labors of this year’s Q1-Q3’s large customer acquisitions - and management has already given a sneak peak via their confidence in >47% revenue growth guidance for next fiscal year.
They have long term visibility into the customers they just freshly or recently landed.
Unless a thesis breaking event happens in the next Q report, I see no reason to reduce my largest portfolio position out of just one quarter’s fear.

To expand further, I see little choice outside of SNOW in today’s environment.
I think growth durability and FCF growth, and secondarily, limited dilution, is paramount in today’s market, and probably for the longer term moving forward.

Even if interest rates drop down over the next several quarters, it’s hard to imagine investors on wall street are going to suddenly make the same mistake by piling all-in to ‘growth at all cost’ stocks, like they did in the last couple years’ bubble mania.
I’ve learned my lesson and I’m sure the hedge funds of today did too (Tiger Global is down over -52% YTD, for example!)

You’d probably have to wait until this existing generation of investors die off so that the fresh recent lessons are forgotten, in order to get another mania like from a year/two years ago. (The 2000 dot com bubble didn’t really ‘repeat’ itself until 20 years later in 2020-2021, right?
Similarly, housing was in a slump for many years in the aftermath of 2008 GFC — the banks learned their mistakes and tightened lending standards— even though a recession was long passed).

Back to company specific points - I’m not going to bury my head in the sand and buy up companies, for example, like SentinelOne, just because it’s the fastest grower (for now, and at a small scale) and cross my fingers that we get zero interest rates again and that everyone forgets about the 2022 bubble pop. SentinelOne is deeply FCF negative and has just shown massive deceleration in growth.

Instead, I’m going to greatly prize companies with durability of growth and +FCF. SNOW meets this benchmark with flying colors, relative to other SaaS today.

SNOW might double their FCF next year, and even hopefully show re-acceleration in revenue growth in the latter half of the year depending on the beats (and like everyone else, depending on the overall economy).

I would like to know - who else are we expecting to do this next year, and at such a massive >2 billion revenue scale?
Who are we expecting to remain in HYPER-growth several years from now?

Who else has a (in my opinion, believably conservative) $10 billion revenue target by FY29 while still growing 30% that year with 25% FCF, and not only that, but also a plan for <2% annual dilution?

Here’s an interesting chart from Guggenheim last month:

Note in the fine print, they list the following as their examples of current “hyper growth >40%” stocks:

Now let’s weed this list out. Let’s remove those who are not yet hypergrowth (>40%) at scale (>1 billion run rate), and remove those without proven FCF positivity.

We now have CRWD, DDOG, SNOW, ZI, ZS

Of these five, I’m already skeptical that CRWD and DDOG can maintain >40% growth next year, or the year after that. Remember, CRWD is guiding for 30% next year. And ZoomInfo already guiding for what, 17% growth next year or something (yikes - what a drastic fall off the cliff from 45% last Q)?
And I’d be concerned DDOG is in the same boat.

Of these, I can really only see SNOW and possibly ZS doing >40% in two more years.

Anyway, these are my two cents.
Of course, there is tremendous opportunity out there for those willing to risk it - people might do much better by allocating more to SentinelOne or BILL or DDOG or CRWD etc, than towards the ‘always expensive’ SNOW, but I am personally not willing to take the same ‘risks’ from what I’ve learned this year - I had drastically cut my allocation to everything couple months ago except to SNOW



I agree with a lot of your thoughts, including skepticism on a lot of “our” companies. I’m down to less than 2% total in SentinelOne and Crowdstrike (just hanging on to some taxable shares for now to avoid wash sales). Cash is again my top position, and it’s been my best performing in 2022, so I’m ok with that for now. No need to push it with companies we don’t feel qualify as the best.

My question would be, why not have more skepticism for ZS and SNOW? I was just thinking back, about the ones we saw as the favorite at different points in time. At one point our #1 was Alteryx. Before that, Twilio and Shopify. Later Zoom…then Upstart. I think we’ve been great at changing our minds when the story changes. But I don’t think we’ve expected our minds or the story to change as often as they have.

That said, if SNOW and ZS are the only companies that you feel qualify as the best, well ok. I would argue for Datadog and Cloudflare and BILL, despite their flaws. Others would argue for SentinelOne and others. Maybe some would put a slower grower like Crowdstrike in the mix, or Veeva or Axon.

My guess is that until most of these companies do well, none will. So my main point here is simply that we should always expect things to change. New companies will join the list of “potentially the best,” and we’ll have make decisions and adjust. I just personally have had much better luck picking a group of “the best ones” than picking “the #1.”



I just personally have had much better luck picking a group of “the best ones” than picking “the #1.”

Hi Jon,
I have to agree with Bear on this. Think back. When, in the past, you have put almost all your funds in one stock it has turned out disastrously (that was Upstart, and you were just as convinced about it). In fact, when you wrote that, it made me actually reconsider whether I should keep Snow as my largest position (I did). Jon, it may work out fine for you, but it’s dangerous! For an unhappy comparison of my own, whenever I’ve been so convinced that I let a position get over 30%, it always has done worse than the rest of my portfolio from that point on. (I can remember Crowdstrike and Upstart as examples).


This is probably the most valuable lesson that I have learned from this meltdown. I would say from investing in hyper growth stocks in general. As far as I’m concerned, the lesson splits into two sub-lessons:

  1. Don’t let any single position – no matter how much you like the company and are convinced about it – grow too high in your portfolio (even in a very concentrated portfolio of 6-8 stocks). What’s too high? I don’t know. What I do know is that UPST at some point got to 26% of my portfolio (10 stocks), and that, in hindsight, was too high.

  2. I sometimes forget about a very simple dynamic: (a) stocks go up because people (or machines, nowadays) buy them (not because they’re undervalued – this is subtle); (b) people buy stocks because they are enthusiastic about them; (c) to be enthusiastic about them, people look at a few key, simple metrics that are accessible and understandable – the investment thesis must be simple (sometimes, splitting hairs in utterly complicated analyses may only be a honorable intellectual exercise if the market is unable/unwilling for any reason to see the same things); (d) these key simple metrics must generate positive surprises (this is critical – nothing is good per se, in absolute terms; it must be good relative to expectations); (e) when they’re no longer enthusiastic about them, people sell and stocks go down. Rinse, repeat.


I don’t disagree with anything you said in your entire post reply. My belief on SNOW is exactly as you describe. I do think it is the best company at this moment in time and holding it with the intention of keeping it long term, that’s no different than how I have always approached my high conviction companies. Unexpected change is the only constant in the real world. (Less than a year ago I scoffed at the idea rates could hit 5% so quickly and yet here we are! Fed guiding for 5.1% and we are already at 4.5%, how wildly wrong I was!)

I am fully prepared to accept if the reality demonstrates SNOW is no longer “the best” company to hold. I don’t get married to a stock.
Just like for example, UPST, right? Last year I held it as the largest position for a couple quarters while the numbers didn’t show anything thesis breaking yet…then Nov 9 2021 results shattered that and I dumped it immediately, within seconds in the afterhours on seeing the revenue print miss.
To others, might be a “disaster”, but to me, that was still a staggering 200%+ profit on most of my shares.

Anyway- as I’ve already said: for SNOW, I think it’s a little different as I believe QoQ type investing is not as applicable as for other companies. Its business fundamentals are driven by long term management decisions and performance in one quarter is a reflection of what they did a year or two previously…it’s not by any means a “Google back in its early days”, but I do think the closest to a publicly investable “pure” hyperscaler today.

And when I am mentioning Google…look at their 2004 Q3 earnings report. Just absolutely astounding.

Google reported 2004 third-quarter net income of $52 million, or 19 cents a share, compared with $20.4 million, or 8 cents a share, in the same quarter in 2003. Revenue increased 105 percent, to $805.9 million, from $393.9 million.

Read that again…105% YoY growth at >3B run rate, and profitable at the same time. Now that represents they have truly special products…and what’s crazier is eight years later in 2012, they were still growing 37% YoY at $53 BILLION run rate.
That’s the kind of company we should be looking for to buy and hold in any kind of economic environment. For now SNOW, is the closest we have (please tell us if there is something else out there!)

This is why I found the recent earnings report miss and growth deceleration of many of ‘our’ companies, especially a deeply FCF negative and small run rate company like Sentinelone, to be, honestly, quite pathetic in comparison to what the “greatest” companies are supposed to be…

I want to hold the best we can identify TODAY with the hope they can be held long term like a Google, but ready to eject it from the portfolio without hesitation if thesis breaks.

(I hope Sentinelone still works out for holders out there, but I hope it’s clear why I have not been willing to “risk” buying it simply because it’s “cheap”. The harsh lessons are abound everywhere this past year. There were quite a few posts on this board in September, only 3 months ago, that pointed out how inexpensive S was at $28 or whatever and they wrote they were buying more on the dip… and look today: down -50% from then; “cheap” can still be very expensive…)


First time poster here, but thought I’d chime in with my thoughts since I currently work as a data scientist for a F500 company. These thoughts are obviously anecdotal, but I have to imagine my experience isn’t unique based on conversations I’ve had with friends at other companies. Note that we use Databricks and not Snowflake (I wasn’t involved in the original vendor selection–I’m guessing it was based on advanced data science needs), but I’m invested in Snowflake because I believe both will succeed given the trajectory of this space and the value creation in my experience translates regardless of the platform. (I also plan to invest in Databricks when they eventually go public).

This type of value generation does not surprise me at all based on what I see from my work and I think it’s a big factor for why Snowflake / Databricks will continue to succeed–they unlock huge gains in productivity and business value. Some of the models I build have generated new revenue to the business that is probably 2-3x our company’s entire annual Databricks spend and I am one of many, many data scientists using the platform. Not to mention, my team has removed probably 100s-1000s of hours annually of manual work through large-scale reporting and automation made possible by the cloud computing–another source of business value.

This value creation is particularly true in a lot of mature businesses undergoing digital transformation–they sit on a wealth of data, but have always operated (often times in-efficiently) without tapping into the power of it. There’s TONS of low-hanging fruit to improve operations/strategy with even basic analytics & models that when combined with the scale of a large business, can lead to significant profit that dwarfs these costs.

In terms of being “mission critical” and cutting back, our company has only expanded our use of data spend over the past year. We’ve made some efforts to cut obvious waste such as setting our compute clusters to turn off after 30 minutes idle time rather than 2 hours, but otherwise the workflows and models that I build (and required compute) are only continuing to grow. When this data work is returning many multiples on your investment, it’d be silly to not continue expanding its use.

My company started migrating to Databricks >5 years ago and even today we’re still moving significant workloads from legacy Oracle databases. I completely believe them when they say it’s a multi-year journey just to get data moved (that’s probably an understatement) and once that process is finally complete, it would make no sense to turn around and migrate to a new platform. The land and expand model is incredibly important given the scale and trajectory of these relationships over many years, which is why Snowflakes NRR remains best in class.