Retrospective Lessons from Crowdstrike's 202

Warning: For a briefer version of this post, please save your time and just read Saul’s post #82229 [1].

The Purpose of This Post
As a less-experienced investor, the purpose of this post is to analyze how Crowdstrike, my highest-conviction position entering 2021, returned -3% for the year. While a lot changes in a year, I found this exercise intriguing as many of the board leaders (who I consider “smart money”) held at least a medium-sized position in Crowdstrike for most of the year. So, this is my attempt at “learning out loud” to identify what some of this “smart money” might have missed.

A Disclaimer
This is not to say Crowdstrike is a bad investment, a bad company, or anything similar for that matter. This is also not meant to oversimplify the past – of course it’s going to be easier to analyze the past! There are members of this board with all sorts investing styles – some more technical, others more financial – some with a tighter leash, others willing to wait for stories to unfold. Crowdstrike may very well fit in most of these portfolios! The reason why I’m doing this is that while in the short-term a share price is unimportant, a year gives us 4 quarterly reports and 250+ trading days, which I deem enough to use a share price as a signal for retrospective analysis.

The Summary
In short, I believe that: (1) we received a couple of amber flags in Q1’FY22, and (2) several red flags in Q1’FY22 for Crowdstrike’s high standards. While some of us noticed some of these flags, the majority of us failed to attribute enough importance to them, probably as a result of confirmation bias from our existing conviction in Crowdstrike. Sure, hindsight is always 20/20; but the point of this analysis is to underline these flags so that this can become part of our arsenal for evaluating future reports.

Pt. 1: Q1’FY22 – Initial Signs of a Slowdown

In June 3rd, Crowdstrike reported its Q1’FY22 results. After reading every post on this board after their earnings, the consensus seemed to be a “good but not great” report. Despite that, Crowdstrike remained the #1 most held position on the board by the end of June [2].

The most evident flag was Crowdstrike’s smallest beat over guidance, at 3.7%. Before the quarter, Crowdstrike had an average beat of 6.2%, underlining the magnitude of the “miss.” Their guidance wasn’t very optimistic either, suggesting 7.1% QoQ at the top-end – down from the three previous quarters, and the lowest since their “peak-Covid” quarter.

The second flag was that for the first time ever, Crowdstrike’s Operating Costs as % of Revenue increased. They showed a beautiful decreasing trend until the quarter kicked in. Similarly, Crowdstrike’s EPS failed to increase for the first time ever. Credit to XavierH for spotting these two issues [3]. Note that even subtracting the additional costs of Humio’s acquisition would also yield an increase in Operating Cost as % of Revenue.

	Q1	Q2	Q3	Q4
2018	125%	117%	110%	102%
2019	93%	92%	85%	78%
2020	75%	71%	68%	64%
2021	67%			

Further, we received the first sign of a slowdown with their sales machine. Previously, their sales efficiency [4] kept improving every quarter. For every $1 spent in Sales & Marketing, Crowdstrike added the following in net new ARR (from Q1’FY21): $1.13 → $1.19 → $1.23 → $1.35 → $1.28 (Q1’FY 22). For the first time, they reported a decrease here too.

Again, retrospect is always 20/20. There were plenty of upsides from the call. Notably, it was the first time that ARR did not decline sequentially from Q4 to Q1. Their cash flows, profitability, and customer adds (even removing Humio’s) were impressive too. A lot of these factors could have been attributed to their integration of the Humio acquisition. However, looking back, there were definitely some warning signs (for their high standards) that I overlooked.

The market reaction from their results seemed to follow the “good but not great” sentiment. Between their respective Q1s and Q2s, the average share price increase from high-growth SaaS increased 51%, while Crowdstrike’s share price rose 30% [5].

Part 2: Q2’FY22 – Underperformance Becomes Evident

While a few of the Q1 mishaps were could have been categorized as minor flaws, Q2 confirmed that the flags were very much there – and Crowdstrike’s best days were definitely behind.

First, the slowdown in net new ARR was drastic. In Q2’FY21, net new ARR increased 77% from pervious quarter’s 65%. Meanwhile, in Q2’FY22, net new ARR decreased from 68% to 44%(!!). Management had warned us that ARR seasonality would be less pronounced, and that they were moving downmarket to SMBs – but this seemed like a “change in story” moment.

Further, operating expenses increased by the highest percentage ever YoY – at 58%.

Next, while RPO remained strong, for the first time – current RPO as a percentage of total RPO dipped from Q1 to Q2. The previous two years, the proportion of current RPO increased from 69% to 72% and from 70% to 72%, while in FY’22 it decreased from 72% to 71%. Seems minor – but it was just another piece of evidence that Kurtz’s Ferrari might not be as unbeatable as it seemed. This was raised in an excellent post by wsm007 [6], where he even highlighted the CFO’s complete deflection to an analyst question about RPO.

Despite the evidence that they were moving downmarket (potentially reaching more, smaller clients), Crowdstrike also reported a slowdown in customer adds from Q1 to Q2 for the first time. And this is including the inorganic customer adds from the Humio acquisition!

Once again, we got more excuses from management about seasonality – something that was previously very rare to hear from Kurtz & team. “While we do not specifically guide to ending or net new ARR, we expect seasonality in net new ARR to be less pronounced relative to prior years as we moved from Q2 into Q3 given our steady climb at a much higher scale in recent quarters. Additionally, please recall that our net new ARR in Q3 of last year included approximately $6.8 million in acquired new ARR.”

I will emphasize again that there were still several positive takeaways from the call. Revenue growth remained at 70% YoY, RPO remained strong, and operating leverage was kicking in better than ever. But things were very clearly different than a year ago. After Crowdstrike’s quarter, it remained the #2 most owned position (after Upstart) right after their earnings, and tied for 2nd after September, too [2]. At this point, the market stopped agreeing with the level of conviction seen on this board. Between Q2 and Q3, Crowdstrike’s share price decreased -29%, while the high-growth SaaS basked increased +23% (that’s a difference of 52%!).

Part 3: The Lessons

Now, this post becomes totally fruitless if we don’t take any learnings from it – as this should hopefully make us all better investors.

Lesson #1: High valuation = High scrutiny
Last month, I wrote a post about how my perspective on valuation changed when I started thinking about it investors placing a bet that a company’s performance would surprise the market [7]. Well, after growing at 89% → 85% → 84% → 86% in the quarters reported in 2020, Crowdstrike was practically tied as the 5th highest valued SaaS stock in the market [8]. In other words, it was one of investor’s highest bet to continue its steady flawless performance throughout the year. This becomes a Catch-22, because the more investors bet for the outperformance, the more the company is expected to outperform. So as soon as the company stops surprising investors as much, the market quickly responds. By the end of the year, Crowdstrike had dropped to the 18th most valued SaaS stock [9].

Lesson #2: Acquisitions are really hard (for a concentrated portfolio)
I have no doubt that Humio was a fantastic acquisition for Crowdstrike. I trust its management, and the techies who have written fantastic overviews about what it represents. However, this is the nth example of acquisitions showing adverse effects on the numbers of hypergrowers. It is not difficult to see why – any minor disturbance to the numbers or the narrative of a company are easy to spot. Crowdstrike had to onboard approximately 100 employees, and seamlessly integrate them into their sales and innovation machine – it is not easy! There are comingled systems and dispersed employees and upselling strategies and R&D coordination that needs to take place. In the long-run, this is most certainly good for shareholders, those that hold a significant portion of their net worth on such a company are going to suffer the short/medium term hiccups that are intolerable in a concentrated portfolio. We’ve seen this story too many times.

Lesson #3: Amber flags become red flags overtime
As this example shows, some of Q1’s issues could have easily been justified. Especially when some of the key figures for Crowdstrike painted a positive picture. But it is essential to take note of some of the issues and watch for their appearance on the subsequent quarterly earnings. We often think about a story changing when it is evident, ala Fastly or Lightspeed more recently – but a story can also be changing right under our noses although it may be harder to decipher.

Lesson #4: Management always gives clues
As I skimmed Crowdstike’s earnings reports, I definitely noticed management’s change of tone in Q1, and even more in Q2. Managements don’t like surprising analysts, as they get sharply punished by the market – so these changes are subtle, but they cannot be understated. It is totally fair for companies to warn us of laws of large numbers, and deals that cause outliers in a quarter. But when questions are deflected, or excuses are pronounced in the prepared remarks – management often communicates something without expressing it in words.

Lesson #5: This crowdsourcing community is full of phenomenal analysts
As I spent time looking into this analysis, I came across many posts (some of which I referenced here) unearthing some of these findings at the get-go. I am sure I read them when they were posted, but I clearly didn’t give them the attention they deserved. Sometimes, there is so much quality posted here that one must fight the urge to “keep up” by skimming, rather than truly reading, digesting, and understanding.

In Conclusion
Crowdstrike remains a fantastic company. In the last quarter, they added a tad bit less revenue than Monday, Amplitude, SentinelOne, and Cloudflare COMBINED. And they’re still growing at 64% YoY. But as Crowdstrike continues to get larger, it naturally becomes more difficult to convince investors to “bet” that it will continue causing massive surprises quarter after quarter.

Will they become the fourth “category-defining cloud platform” that they continue to tout on their investor presentations [10]? Maybe…they very well could. Could they be a Shopify-esque example of continued share price outperformance at a latter part of the S-curve? Could be. But what’s clear to me after undergoing this analysis is that the degree of confidence that Crowdstrike is one of the best ideas to place our capital today is in doubt. In fact, it was very much in doubt many months ago. Which leads to Saul’s question – “If I wasn’t in Crowdstrike and someone came along and proposed this company to me out of the blue, would I start a position in it?" I warned you that you could have saved some precious time by simply reading his post!

Visit… to maximize your learning of the board before posting

[4] Using the SaaS Magic Number to compute this
[5] Basket includes: Datadog, Docusign, Cloudflare, ZoomInfo, Zscaler, Snowflake


Good analysis but I wonder if it is too late to sell CRWD now. Let me explain. CRWD should do 1.45B this year. Next year if we project revenue growth of say 53% we end with an ending P/S of about 20. This is the same as Service Now which is growing at 25-30% growing at half the rate.

Let us compare CRWD with ZS and NET. Both ZS and NET have slower current revenue growth rates than CRWD but their valuation as measured by P/S is much higher (CRWD, ZS, NET PS =30, 45, 55 by end of this Q). Do we think ZS and NET are going to show much more revenue outperformance in FY 22 and beyond than CRWD to justify their premium? If not the market will rerate these 2 companies quickly.

Following the numbers is one thing which we are doing here. The more interesting thing is the technology, competition, and moat of these 3 companies. If we believe CRWD is the weakest of the 3 in this aspect then it is quite likely that its revenue should decay the fastest. It seems to me ZS has the best moat and least competition in enterprise companies. Can ZS maintain its revenue growth despite the friction? CRWD may have the greatest competition among the 3 from S which is a company 1/7th size.

Long CRWD (8%)and S (6%)


CRWD has decelerating revenue. ZS is re-accelerating. You can’t analyze them with numbers that are static, as they are changing in opposite directions. NETs valuation counts on their growth maintaining for years, not quarters. That’s the low end thesis, although I think most expect their revenues to accelerate handily.