My thoughts on all this!

My thoughts on all this!

After reflection, I find that I agree completely with this question of Monkey’s: Why basically does DDOG seem to be the only “reasonable” stock to hold anymore because everything else seems to have one “problem” or another.

I have noticed also that that has been the way it tends to look because everyone seems focussed on analyzing every tiny detail of each company’s report, searching for tiny faults, and then saying “Disastrous report! I’m selling out!”

Please excuse any little mistakes in the detail of numbers that I make below, but I’m going from memory here.

For example Zscaler’s billings “only” growing at 50% or 60% when they had been growing faster, to me says that I might adjust my position size slightly, not that I’d sell out.

For another example, the assertion that we should exit our fastest growing company (Sentinel) because new customers dropped from 800 last year to 750 this year. (Oops! Sorry! Last year that 800 included all the customers from an acquisition! The 750 this quarter is an all time record!)

No, Billings and New Customers for a Quarter are not the same essential reasons for which we exited Fastly, Zoom, Docu, etc

We exited Fastly because, with all their “great tech” they only added 9 new paying customers in a quarter while their competitor, Cloudflare, added 80 times as many! No, that’s not a misprint, it’s not 80% more, not even 80 customers more, but 80 TIMES as many. It was 9 compared to 720 or so! There were half a dozen other major reasons, but that one sticks in my memory.

We exited Zoom because they did it all, and signed up all the world, in a couple of Covid quarters in 2020 and there was nowhere else to go. For the next few quarters they had all those customers and all that revenue that they hadn’t had before, but it wasn’t growing hardly at all. They had already conquered the world in their niche! I kept warning that sequentially they they were hardly growing at all. Revenue growth went sequentially from something like 120% QoQ to 17% QoQ, to 8% QoQ to 5% QoQ, but they were still showing 350% growth year over year each quarter because their revenue was so much above pre-covid. And then, after four quarters, year over year collapsed as well, as they were now comparing with the big post-Covid-growth quarters.

We exited Docusign for the same reason we exited Zoom. When you have already signed up 50% or 60% of your TAM during Covid, even if your product is great (it must have been or you wouldn’t have signed them up like that), you can’t keep growing at 50% or 60% per year. Period.

For me I’m investing in hyper-successful companies and I want to keep holding them as long as the story and numbers are intact, and I’m NOT going to obsess, and go crazy trying to analyze every tiny little metric! But that’s just me, and if you want to go ahead with that, go ahead.




Hi Saul,

I believe this forum thrives when there are honest and open debates, so I’m going to engage hoping to understand your position better.

Even though I take your point that obsessing over minutiae is not productive, I believe there may be solid reasons for exiting both SentinelOne and Zscaler and I would like to try to articulate those and hear counter-arguments.


The customer growth story in terms of the number of customers added remains intact as you rightly point out above. However contrary to the story with Snowflake - which is that while the growth in the number of customers added may be declining, the quality of those customers are improving - I believe the opposite may be happening for SentinelOne. This could be indicated by the growth of the # of customers above $100k:

**>$100k	Q1	Q2	Q3	Q4**
2021	122	145	173	219
2022	277	345	416	520
2023	591			

2021		19%	19%	27%
2022	26%	25%	21%	25%
2023	**14%**

So the number of large customers added, while still a lot, has decelerated quite a bit this quarter. And this while it is their strategy to target larger customers. Perhaps they are not as successful as they would like to be and Crowdstrike and PaloAlto are grabbing the big ones?

The second point relates to operating leverage, which I would argue also went the wrong way:

**Op %	Q1	Q2	Q3	Q4**
2021	-126%	-101%	-102%	-103%
2022	-127%	-98%	-69%	-66%
2023	**-73%**

**FCF %	Q1	Q2	Q3	Q4**
2021		-70.0%	-80.5%	-85.6%
2022	-87.4%	-97.6%	-37.0%	-10.7%
2023	**-69.9%**

Whereas the decline in operating margin could be put down to seasonality, the FCF break is a lot bigger. I therefore question whether the increasing operating leverage trend is still in tact. And this is probably my main concern and the key thing which I think has changed.

A narrative around these numbers could be that they are having to spend more and more, to attract relatively less valuable customers than the competition.

In addition, as with the comparison of Cloudflare and Fastly, a comparison with Crowdstrike is appropriate. I have previously thought (and written) that the likely trajectory of S is relatively easy to anticipate as they track CRWD’s earlier performance pretty well. Lately that has also started to deteriorate. Over the last 2 quarters SentinelOne has become much less efficient in how they generate roughly the same revenue that CRWD did about 3 years earlier.

Comparing Crowdstrike’s Q219, Q319 and Q419 revenues, operating and FCF margins to SentinelOne’s for the last 3 quarters yield the following:

**CQ219	CQ319	CQ419**
**SQ322	SQ422	SQ123**
CRWD	55.7	66.4	80.5
S	56.0	65.6	78.3
CRWD	-50%	-43%	-35%
S	-69%	-66%	-73%
CRWD	-64%	-20%	-4%
S	-37%	-11%	-70%

The above spend trajectory seems to invalidate my previous hypothesis that SentinelOne will track Crowdstrike’s trajectory. So again, this to me would indicate that the numbers have changed, thereby calling into question my original thesis.

SentinalOne now seems like they are having to spend like hell just to try to keep up with Crowdstrike’s earlier trajectory, not to mention Crowdstrike’s current trajectory.


A lot has been written about Zscaler, so I would just like to add one thing: this is not a one quarter billings blip. This is the second quarter of decelerating billings growth. Last quarter this was the main orange flag I saw in an otherwise very solid set of numbers - and I did not sell, but this quarter the trend got worse - and I did:

**Bill	Q1	Q2	Q3	Q4**
2019	65	115	85	126
2020	88	135	131	195
2021	145	232	225	332
2022	248	368	346

**YoY	Q1	Q2	Q3	Q4**
2019	35%	17%	54%	55%
2020	64%	72%	72%	70%
2021	71%	**58%	54%**

So while the last 3 quarters of revenue growth has tracked above 60%, billings growth has now been in the 50-s for two consecutive quarters.

And my expectation, based on what they’ve done in the last 3 years (a 48% qoq billings increase) is that billings growth will again hit mid/low fifties yoy next quarter.

Because billings drive revenue, it is inevitable that revenue growth will drop below 60% in the next couple of quarters, imho.

That, for me, is again enough to come to the conclusion that something has changed. They are simply not as successful at billing for new work as they were a couple of quarters ago. Their >70% billings growth 4 quarters in a row, which has fuelled their increasing revenue growth rate over the last several quarters, has now gone in reverse and will inexorably drag down their revenue growth in quarters to come.

I thought that the reasons for both companies were enough to sell, but perhaps I was too trigger-happy and am missing something.

Appreciate your thoughts, as always.



1) SentinelOne

  1. Customer growth > 100K ARR: Yes, there was a slowdown here. However, it was still growing 113% YoY (vs. 127% in Q1 2021). Over the last 8 quarters, YoY growth was between 111% and 140%. I don’t see how this changes the thesis, they are still growing >100K ARR faster than overall customer growth and faster than revenue growth. They added 71 large customers sequentially, the second strongest ever! Why do you consider this bad?

  2. ARR per customer: SentinelOne increased ARR / customer from 27.8K in Q1 2020 to 45.5K in Q1 2022 and it has been increasing every quarter. Crowdstrike’s ARR / customer decreased from 110K in Q1 2020 to 107K in Q1 2022. So yes, the trend that CRWD is moving downmarket and SentinelOne upmarket is still existent.

  3. Operating leverage: Why do you compare with previous quarter when there is so much seasonality involved? If you compare YoY, they improved operating margin by 54 percentage points and FCF margin by 17 percentage points. If you want to compare QoQ, at least use TTM to adjust for seasonality, it looks like this on a TTM basis:
    Operating margin: -140% => -128% => -116% => -107% => -111% => -108% => -96% => -85% => -75%
    FCF margin: -88% => -77% => -79% => -78% => -82% => -89% => -73% => -51% => -52%

Overall, the trend of improving profitability is intact. ?If you read the earnings call transcript, management hinted that they expect to improve profitability by 30 percentage points every year. They hinted that by applying that 30 pp improvement, they should be profitable in CY24, and even sooner on a FCF basis. SentinelOne is not discounting either - I recently read some customer reviews and they were all very positive about SentinelOne, and the most common complaint was that SentinelOne is more expensive than the competition, but that they still prefer to pay up because the solution is so much better. This indicates pricing power to me - I don’t see any reason why SentinelOne won’t become profitable. In their long-term targets, management guided towards >20% operating margin. (refer to their first shareholder letter)

Recently, IDC released their market share report in the endpoint protection space, 2020 & 2021:

Crowdstrike: 9.7% => 12.6%
MSFT: 8% => 11.2%
Then a bunch of more legacy tools like Trend Micro, Sophos, Symantec, Kaspersky, Carbon Black etc - all losing market share.
SentinelOne: 1.1% => 1.8%

The only ones not losing market share are: CRWD, S, Cybereason, PANW & Lacework. The legacy tools still have 50% or more market share, that is 50% market share that SentinelOne can still capture with their superior solution.

The link:

My conclusion: I was hoping for better margins, but nothing has changed my thesis here. Is this Crowdstrike? No, CRWD will continue to dominate the endpoint space and capture most of the market, that’s why I am also long CRWD. They have approx. 13% market share and might eventually get to 25%. My thesis for SentinelOne was never that they will replace Crowdstrike, they both have >97% gross retention rates. However, my thesis is that SentinelOne will get significantly more than the current 2% market share, and that SentinelOne will be a significantly larger company than the 6.5B $ market cap today. Not to mention that they are also moving towards cloud workload protection, identity protection & much more. The ARR in cloud workload protection grew from approx. 1M $ last year to 30M $ today, that is 30x in one year!

2) Zscaler

There were great discussions about Zscaler here recently so I don’t have much to add. I think that Zscaler will be a significantly larger company in the future than today, so I don’t see a reason to sell out. No thesis change happened for me.I am watching Cloudflare’s entry in the Zero Trust space, but it doesn’t seem to be a winner takes it all market.


General thought: I am not obsessing over QoQ metrics and try to look where the company could be in a few years. I have an investment thesis for every company and as long as the thesis is still there - no reason to sell out unless a significantly better opportunity arises. QoQ obsession over metrics has mostly resulted in poor investments, like Amplitude that suddenly shows accelerating QoQ growth for 2 quarters before hitting a wall or Lightspeed that accelerated coming out of the pandemic but has a poor, low margin business model (Some might add Upstart here, I was lucky enough to sell out on time). In general, the competitive environment & competitive position of a business doesn’t change (dramatically) in just one quarter.