I decided to take a small position in Snowflake in spite of the price.
Any company which is growing revenue at 115%, and growing remaining performance obligations (RPO) at 240%, and has a net retention rate of 162%, and has 65 customers with TTM revenue over $1 million, up over 100% from 31 a year ago, and has rising margins and rising free cash flow, is clearly doing everything right. You shouldn’t let price bias get in the way of paying attention to those numbers. Clearly its customers think it has great products.
Look, an average company that had 31 million-dollar clients last year would be bragging if they had 35 or 38 this year. Snowflake has 65 now! Up over 100% in a year! That’s incredible!
A normal company would be happy to have their RPO rise 40%. Snow’s rose 240%. That’s two hundred and forty! That’s incredible! Do you actually realize how incredible that is? It’s more than tripling all that revenue waiting to be recognized.
The rest of our Category Killer companies brag about having net retention rates in the 120s or 130s. Snow’s was 162! That means that last year’s customers spent 62% more this year than they spent last year, BEFORE Snow had to go out and get any new business. That’s incredible!
Yes FCF is still negative. But in FY 2019 it was a loss of 152% of revenue. In FY 2020 it was a loss of 75% of revenue. In the first three quarters of FY 2021 it was a loss of 22% of revenue.
Product gross margins (their equivalent of subscription gross margins) went 58%, 63%, and 68% in the same three periods.
And it’s all like that.
Best,
Saul
Some people are “waiting until after the lockup is released.” The price has risen $100 in a week or so (with the quarter results announced), going into that release date. It’s been an expensive “waiting.”
I don’t really know about Rockset since it’s still a small private company. Although I know about the CTO pretty well though my close friend circle. He’s a brilliant guy (and speaks my same mother-tongue) and did a great job at Facebook previously. Anyways, I’m sure there are many small companies who are creating solutions for the cloud data world but SNOW has already arrived and is perhaps the biggest fish now.
Really interesting that RPO is at 928M while product revenue is at 148M. I did see that a number of customers had switched from one year contracts to three year. They mentioned on the call the weighted contract life is 2.5 years.
What I am trying to understand is if switching to the 3-year contracts is creating an illusion of more massive growth than exists? Or is that just a good sign that customers want the longer term contracts?
There was some detail that some customers are consuming all their credits before the three years is up, based on usage.
Another thing to keep in mind is their net retention rate is measured differently than other SaaS companies, because as they have said, they are a “consumption company” and not a SaaS offering. I don’t fully understand the implications of this, just throwing it out there.
Maybe someone else can elaborate if Snowflake’s net retention rate is an apples to apples comparison with other SaaS vendors?
If RPO is at 928M and spread it out 3 years. That’s 310m each year, more than double olf current year product revenue. Keep in mind the RPO will continue to grow with the product revenue. This is simple math and the market is not that stupid to be tricked by simple numbers.
Sorry for above, typed too fast, the 148m is quarterly revenue not annual revenue. So annual revenue is 148m x 4 = ~600m. Spreading 928M to 3 years, that gives 310m each year. That’s only 50%. With 100% revenue growth rate, it seems half of the growth is coming from previous RPO and half of itfrom new contracts.
But the RPO is growing at a crazy rate. I think the RPO growth rate is a good indicator of future revenue growth rate. The customers decide how much future credit to purchase. They have a good idea of what they need.
For example, for my daily checking account. I just keep a balance of $200 at all time enough for 1 week. I top it up once a week. I think it’s the same with corporations. They don’t buy credit they are planing to use. They’ll use it up within the time frame.
If I understand correctly CEO and CFO they were pushing for longer term contracts bringing average length to 2.5y. Looks like this 240% in RPO growth is related to lengthening of contracts at least in part. If we want to compare with other our companies we should adjust these RPO numbers to average contract durations.
They also guided to deceleration in revenue growth next quarter. Could be sandbagging, but quarterly growth trend has been decelerating yoy, need to look at sequential qoq growth numbers.
But the RPO is growing at a crazy rate. I think the RPO growth rate is a good indicator of future revenue growth rate. The customers decide how much future credit to purchase. They have a good idea of what they need.
For example, for my daily checking account. I just keep a balance of $200 at all time enough for 1 week. I top it up once a week. I think it’s the same with corporations. They don’t buy credit they are planing to use. They’ll use it up within the time frame.
I am still failing to see how it’s an apples to apples comparison on previous RPO when the contract length may have been extended by 2-3x.
Let’s take an example of a company which is spending 1M a year on Snowflake. They currently have a one year contract and now they decide to switch to a three year contract for 3M. Suddenly RPO has increased from 1M to 3M, but nothing has really changed, since this company would be purchasing a million per year anyways.
To keep things even on the RPO comparison the contract length would need to be fixed. For those numbers listed above, we’d also need to know average contract length at the end of each quarter, otherwise the comparisons are misleading.
On the call they said the contract extensions are primarily not to go through procurement each year, since they know they need the service, they might as well go for a longer contract. This probably gets them a slightly better rate anyways, so it might be going from 1M to 2.8M in spending for example with that example above.
RPO can be inflating as they see longer term contracts…
RPO growth rate is the only figure that is staying high in 220% to 240% for SNOW.
Their Y/Y revenue growth was 154% for Oct 19 quarter ($73M/$28.7M)… has now decelerated to 119% for Oct 20 ($159/6M / $73M)
Similarly, here are growth rates comparing y/y for Oct quarter
Total customer growth rate decelerated from 175% to 84%
Customers with $1M revenue growth rate decelerated from 121% to 110%
DBNER decelerated from 189% to 162%
Deferred revenue (Current = what can be recognized over next 12 months) grew only 34% on y/y basis in this latest quarter
Agree, all of these are showing law of large number deceleration… nothing really worrisome… SNOW business looks really well set for continued growth for many many years…
Question really is not even current valuation… it is really what is the upside from here… if you look forward one year, and expect growth rate to decelerate to just 100%, with a TTM revenue of $1B, you are paying $113B in market cap today… is it justified? may be… but can it double from here to $200B+ market cap? That is hard to expect.
fast forward two years, assume 80% growth rate and $2B TTM revenue, you are paying 55x of that in price today… that is slightly lower than CRWD trading currently at 65x of revenue at ~80% growth rate… so again, that may be justified at that time, but what is upside?
I understand Saul’s philosophy of ignoring valuation when you see incredible business… but I contend SNOW is showing sign of same laws of large numbers catching up in not too distant future and at current prices, upside is not so high.
I have tiny position in SNOW knowing well I can and am wrong many times, however, hoping that with lockup expiry, we will see better entry point… if it never happens, its ok, there are plenty of high quality growth companies to choose from where I can be more confident of higher upside.
For the contract duration change, I don’t think it’s a sudden process. It’s not like everybody suddenly decide to extend to the maximum contract length possible. It’ll be a overall gradual change for the whole customer base.
Some people may worry about the SNOW consumption model. SNOW may remind them of FSLY. However, SNOW is very different than FSLY. FSLY 's content consumption is retail consumer based. We know retail consumer behavior is cyclical, unpredictable. Companies can also switch out of FSLY to other vendors very quickly. Tiktok is a good example.
It’s not the same with SNOW. SNOW’s customers are data driven not consumer driven. So the data demand is more preditable. Companies do data analsys for all kinds of things. And data will only get bigger and bigger. Personally, I use GoogleDrive to manage my investment. I have more files and more files over time… SNOW’s platform is also sticky. Let’s say companies purchased 3 years of credit. They’ll stay with SNOW for at least 3 years. It’s like a phone contract. So I am not too worried about SNOW’s consumption model.
For the high valuation:
Different companies are valued by the market differently. A shoe company growing sales at 100% per year is not going to get the same valuation as SNOW. Even companies in the same industry are not valued the same because of favoritism. For this reason, I don’t try to use formula to do exact calculation to arrive at a fair valuation. SNOW gets current valuation because the market supports it. Going forward, what moves the stock price will be the revenue growth. So we have to keep an eye on its growth. If there’s a QoQ slow down, is it temporarily or is it permanently trend and continue to slow down in the future. It’s a tough question to answer. I try to make obvious choice: what is happening now?