Ben’s Portfolio update end of September 2022

Ben’s Portfolio update end of September 2022

Returns:

Portfolio
2020 63.6% (Since May 12, 2020)
2021 13.1%
2022 YTD Month
Jan -19.9% -19.9%
Feb -18.7% 1.4%
Mar -21.3% -3.2%
Apr -37.8% -21.0%
May -50.5% -20.4%
Jun -53.5% -6.0%
Jul -49.6% 8.3%
August -46.0% 7.3%
September -50.7% -8.7%

These are my current positions:

September August July
Snowflake 17.1% 16.6% 13.4%
Datadog 15.4% 17.6% 20.6%
Crowdstrike 12.3% 12.5% 13.4%
Zscaler 12.0% 10.6% 11.1%
Cloudflare 10.1% 10.4% 6.7%
SentinelOne 9.1% 7.8% 7.7%
Monday 5.2% 4.7% 4.6%
Nvidia 5.0% 5.7% 7.3%
ZoomInfo 4.6% 4.6% 4.1%
Hubspot 3.7% 4.2% 6.1%
Enphase 3.3% 3.1% 3.3%
TradeDesk 2.3% 2.2% 1.7%

Company comments:


SentinelOne:

Ok, this is going to be a little tricky: I had expected $92M organic revenue which equates to 18% QoQ, 100% YoY growth. Unfortunately management decided to not break out revenue from Attivo, which was expected to add $8M. What we know is total revenue came in at $102.5M (124% YoY, 31% QoQ growth). One analyst asked “Did you guys give the Attivo ARR number for F 2Q?” and the answer was “In terms of ARR, I think we’d said it was going to be around – was it 35 for the quarter, in general around that”. So with that little hint we can make some assumptions to estimate Sentinel’s organic revenue growth this Q2: $35M / 4 = $8.75M, which is 9.4% more than (but still in-line with) the “guided” $8M revenue from Attivo in Q2 - good. Now we can subtract $8.75M from their reported revenue and arrive at $93.8M estimated organic revenue; Again roughly in-line with my expectation ($1.8M or 2% more) - also good. This equates to an estimated organic revenue growth of 105% YoY (or higher if last year’s revenue wasn’t all organic - there was the Scalyr acquisition!) and 19.8% organic QoQ, which is similar with Q1 (19.2%) and an acceleration from Q4 (17.2%) - very good! It also shows a nice sequential raw revenue add progression (raw QoQ add in million USD): 3.89 → 5.31 → 7.53 → 8.35 → 10.27 → 9.62 → 12.62 → 15.50 (organic), with just one sequential decline in 4Q22 and a doubling from Q2 last year! I find these (estimated) numbers very good and I am actually surprised that management isn’t more transparent here as there doesn’t seem to be anything to hide - quite the contrary! But maybe I got something wrong?

Another interesting metric was large customer growth, although it isn’t clear to me whether that’s all organic or if Attivo added large customers to the mix (probably Attivo added large customers to the mix …). Here they grew 28% QoQ, the highest sequential growth they had since 4Q20 and a steep acceleration from last quarter’s 14%. Unfortunately I can’t really give them any credit for this since I don’t know if this was organic growth or not. One bright spot to credit though was the close of “an eight-figure multiyear deal, the largest new customer contract in SentinelOne’s history”. The only metric where they gave us organic numbers was the total customers, which grew 10.7% organically, just a notch slower than last Q (11.2%) and down from 14.9% in last year’s Q2.

Another piece of great news is that their “net retention rate reached a new high of 137%”, by far the highest it has been: 114% → 122% → 119% → 122% → 121% → 115% → 117% → 124% → 129% → 130% → 129% → 131% → 137%, “driven by significant subscription expansion, especially from our channel ecosystem and cross-sell of adjacent solutions”.

GAAP gross margin held constant at 65%, but up from 59% last year and more importantly, non-GAAP gross margin increased to a record 72%, up from 68% last Q and 62% last year - again Attivo might have helped here -which is all good.

In my July recap I wrote “SentinelOne needs to show improved leverage/profitability in the next ER if they want to keep me an happy investor.” I was very much relieved to see that they exactly did that: Operating income margin increased to -57%, up from -73% last Q and up from -98% last year. Net income margin increased to -55%, up from -73% last Q and up from -101% last year - what a big step in the right direction. Don’t get me wrong, Sentinel still has a long way towards profitability, but the trend is your friend, and the trend looks really good! Even FCF margin improved to -65%, up from -70% last Q and up from -98% last year. And improving profitability by that much while at the same time continuing to deliver triple digit YoY revenue growth is just amazing. One wrinkle I see here, however, is that they might have shifted expenses to later in the year “I think one of the things you’ll see in our results this quarter is just due to the market conditions, I think we were really looking at spending that wasn’t essential and having some controls over that as we sort of waited to see how the market was going to work its way through Q2.” And “there is some shift between Q2, Q3 expenses or expenses from Q2 into the second half as we felt a bit better about the market conditions.” They guided for Non-GAAP operating margin to -57% in Q3, the same they got in Q2. Since they got what they guided for in the last two quarters, I’ll take this number at face value. And while -57% is an improvement from -69% in Q3 last year, I would like to see continued sequential progress as well. But I think what is happening here is that Q2 leverage improved more because they shifted some expenses to Q3 and this is why Q3 is not expected to improve sequentially. Overall though, FY leverage will likely improve significantly in comparison to last year, where I am expecting an operating margin expansion of about 30%, to below -55%, up from -85% in FY22.

I interpret their revenue guidance as around 17% QoQ growth, but realize that we don’t have a long beat history and in the short history we have, it has been jumping around quite a bit, so wouldn’t be surprised if it will come out 2% higher or lower. Full year revenue guidance was raised by 2.7% to $416M at the midpoint - a healthy raise. Yet, “we want to be prudent about the evolving macro environment” while “cost-consciousness and prudence around IT budgets (…) have resulted in marginally longer sales cycles and more budgetary approvals. (…) The impact has been quite modest so far”. I expect organic net new ARR to plateau next quarter, when it goes from about $65M this Q (without the 35M from Attivo), to a “high $50 million range” and again incorporating “the macro factors I just mentioned, which we believe is prudent in this environment”.

One additional, noteworthy point:

*Singularity Cloud remained our fastest growing solution in Q2


Zscaler:

Before I come to the what I think was the most exciting highlight of this quarter (spoiler alert: Billings), Zscaler earned $318M in revenue (61.4% YoY, 10.9% QoQ). I had expected approximately $322M (63% YoY, 12% QoQ), a notch more, but I think close enough, given what else they delivered this quarter: Billings came in at $520M, that is a 50.6% sequential increase! When I saw this number I did a little jump in my chair, followed by a deep breath of relieve: The previous two quarters we saw soft billings in comparison to the prior year and given that, as well as the current economy, I was prepared to see a significantly lower sequential increase. Indeed, I would have been happy with just 45% sequential growth, applying the same relative softness we saw in Q3 in comparison to the prior year’s quarter. For reference, Q4 sequential billings growth had been in-between 47.6 and 48.6% in the 3 prior years. So my highest hopes were 48% - they beat those solidly by 2.6%. Well done! In-line with the strong billings seasonality we saw the usual sequential acceleration in RPO, which grew 68% YoY to now $2.6B, 49% of which is cRPO (to be realized in the next 12 months) and deferred revenue grew 62% YoY to $1021M, which both looks good to me. Another positive I saw this quarter was the relatively strong large customer growth, which grew by 198 new >$100k ARR customers, 10.5% QoQ, up from 8% last quarter and 39 new >$1M ARR customers, 13.5% QoQ, down from 14.7% QoQ last quarter. Because of the noise in small numbers I am not concerned at all about the slight growth deceleration in those >$1M ARR customers (had they added only three more, QoQ growth would have been flat). In contrast, accelerating large customer growth from 8% to 10.5% in the current economy is a great feat. They also now have over 20 customers exceeding $5 million in ARR.

Zscaler’s net retention rate has again exceeded 125%, now for 7 consecutive quarters as “Happy customers buy more, and our net promoter score continues to exceed 70, which is more than 2 times the average NPS for SaaS companies”.

I was also happy to see that their non-GAAP gross margin improved by nearly a whole percentage point, both in comparison to last quarter and in comparison to last year, to 81.6%.

Coming to profitability they had great progress this quarter: FCF margin (24%, up from 15% last quarter and 14% a year ago), operating income margin (12%, up from 9% last quarter and 10% a year ago) and net margin (11% up from 9% last quarter and 10% a year ago) jumped to their highest percentages in 5 quarters. So really good progress here as well. And looking towards FY23, they “expect free cash flow margin to be 20% or above”.

Driven by those excellent business results emerges a very strong narrative: Zscaler has a large and increasing TAM in-front with emerging products such as ZDX to manage digital user experience, and Zscaler for Workloads to secure servers and workloads. Those emerging products “contributed 14% of our new business in fiscal 22” and are expected “to contribute high-teens percentage of our total new business in fiscal 2023”. In addition, Zscaler has a >6x upsell opportunity with existing customers on ZIA and ZPA alone. Just as a reminder, Zscaler can service users and workloads. ZPA, 19% of total product revenue, growing 80% year-over-year, stands for Zscaler Private Access: “ZPA is an easier to deploy, more cost-effective, and more secure alternative to VPNs. ZPA allows you to give users policy-based secure access only to the internal apps they need to get their work done. With ZPA, application access does not require network access. ZIA stands for Zscaler Internet Access. ZIA is for connecting users to public applications on the internet: “It monitors the cloud and provides a central location for software and database updates, policy and configuration settings, and threat intelligence”. Both ZPA and ZIA have now FedRAMP High authorization (ZIA just got it), “the only cloud security service to have two products at the highest level of FedRAMP certification. In addition, ZPA is the only Zero Trust solution with DoD IL5 certification”. These certifications help attracting government customers: In Q4 “we added over 25 new Federal customers” and now they “have landed 10 of the 15 Cabinet-level agencies as customers, with plenty of opportunities for upsell at these large organizations”.

Applying the usual beat, I interpret their Q1 guidance as 12% QoQ revenue growth, which, due to tough comps with 1Q22 where they grew 17% QoQ, would come in at 55% YoY growth for Q1. However, because of the tough comps due to this one outlier quarter, we should focus on sequential growth or annualized sequential growth, the latter of which would be at 57% YoY (1.12^4-1). I am not sure though how much conservatism is baked into the Q1 guide because if I update @wsm007’s “4Qt” metric which is the average of last 4 quarters of billings, (which correlates well with the next quarter’s revenue, more on that here: Zscaler’s remarkable predictability and here: Ben’s Portfolio end of July 2022 - Saul’s Investing Discussions - Motley Fool Community), 4Qt points to $370M in revenue, a 16.4% sequential increase. For comparison a 12% sequential increase would result in $356M revenue, 4% less than predicted by the 4Qt metric. Looking back, a 4% over-prediction with the 4Qt metric has happened before, but it would be somewhat on the high side, given the 3.2% standard deviation of differences between 4Qt and next Q’s revenue over the last 20 quarters. It is also noteworthy that this was their first guide for the new FY and so setting the expectations low to allow for all those beats and raises yet to come in FY23 seems like a sensible (and not un-common) strategy. In answering an analyst question about how much conservatism is baked into their full year guide, Remo Canessa (CFO) said “We did see higher deal scrutiny in Q4” and “many customers have not put their budgets in place for calendar ’23, [which] create[s] a level of uncertainty in the second half of our fiscal '23”. This resulted in a “prudent” full year guide of $1.5B which corresponds to a 37.5% YoY growth, similar to the 40% YoY growth they guided to for FY22 in their last quarter of FY21.

So to me, Zscaler, now in its 11th year as leader in Gartner quadrant for security service edge, continues to look like an extremely solid (~60% YoY) revenue grower which has accelerated YoY revenue growth from the low-50s to the low-60s in the last two years and shows great potential to keep growth durability in the high 50s to low 60s throughout FY23. And this while being non-GAAP profitable, generating a large amount of cash and showing continued, despite slowly, improving leverage, as “we remain confident of reaching 20-22% operating margins in the long-term” (which would be roughly a 2x from where they are now).

Some additional points:

*We are entering this fiscal year with a record pipeline.

*In fiscal 23, in our guidance, we intend to deliver operating margin expansion of approximately 150 basis points.

*ARPU has been increasing on a year-over-year basis. Our ARPU has increased 20%.


The other companies:

You can find my earnings recaps of Monday, Datadog, Snowflake, Cloudflare, Hubspot, Zoominfo and Crowdstrike here: Ben’s Portfolio end of August 2022

And the earnings recap of Enphase Energy as well as some pre-earnings thoughts on the others here: Ben’s Portfolio end of July 2022 - Saul’s Investing Discussions - Motley Fool Community.

In addition, here are some scatter-thoughts I had this month for some of our companies:


Monday:

TLDR: Question: Is Monday’s NRRs for customers contributing >$50k ARR pretty much meaningless? Answer: probably yes(?!)

One question I had was about how Monday calculates its DBNRR (Dollar Based Net Retention Rate) for the different customer cohorts and how this will change the interpretation of it. To provide some context from their last report: Net retention rates held steady in all categories with >125% for all customers, >135% for customers with 10+ users and >150% for customers contributing >$50k ARR and from the call: “we believe we are now at the stage where it’s stabilized”. Now the question is whether there are customers that are counted towards the >$50k ARR cohort for the purpose of figuring the NRR that actually contributed less than $50k ARR at the beginning of the evaluation period.

Why does this matter, you might ask? Let me draw a few scenarios which are not very realistic, but efficient to make the case:

Scenario 1: At the beginning of the evaluation period Monday has 5 customers contributing $51k ARR each. At the end of the evaluation period (12 months later) those same customers are contributing $60k ARR each. That means, at the end of the evaluation period Monday has a NRR of 117.6%, or (1/5)*(60/51 + 60/51 + 60/51 + 60/51 + 60/51) = 1.176 = 117.6%.

Please let me know if I got even this simple example wrong, but this is at least how I understand how it works.

Now what happens if we have

Scenario 2: At the beginning of the evaluation period Monday has 4 customers contributing $51k ARR each and one customer contributing $20k. At the end of the evaluation period (12 months later) those same 5 customers are contributing $60k ARR each. How wound Monday calculate its NRR now? I see two possible ways:

a) (1/4)*(60/51 + 60/51 + 60/51 + 60/51) = 1.176 = 117.6%.

b) (1/5)*(60/20 + 60/51 + 60/51 + 60/51 + 60/51) = 1.54 = 154%.

To me, a) would make the most sense since the one customer wasn’t contributing more than $50k ARR at the beginning of the period. In this example you can see the issue with b), where one customer that ramped up from a small contribution to above $50k ARR can introduce a huge bias in the NRR: You can come up with

Scenario 3: where Monday has 399 customers contributing $51k ARR each and one customer contributing $1k ARR at the beginning of the evaluation period and those same 400 customers are contributing $60k ARR each at the end of the evaluation period. Here you can calculate Monday’s NRR as follows:

a) (1/399)*(399 * 60/51) = 1.176 = 117.6%.

b) (1/400)*(60/1 + 399 * 60/51) = 1.32 = 132%. Quite a different picture just because of one customer ramping spend from essentially nothing to >$50k ARR! (You can replace $1k with $100 and see what happens then!)

In any case, since I wasn’t sure which way Monday really does it, I asked Investor Relations and this is what they replied (it needed a follow up email):

Q: Does the 150% NRR you quoted in your last earnings include any customers that were contributing below $50k ARR at the beginning of the evaluation period?

A: Thank you for your interest in monday .com. For our net dollar retention calculation, it is a four quarter average based on users at the end of the evaluation period.

Follow-up Q: So, does that mean to ‘qualify’ for the “$50k ARR cohort NRR calculation”, customers do just have to be in the $50k ARR cohort at the end of the evaluation period? Or in other words, are there customers in the “$50k ARR cohort NRR calculation” that were contributing less than $50k ARR at the beginning of the evaluation period?

A: To be part of the $50k+ NRR calculation, the customer will only need to be over $50k ARR at the end of the evaluation period, not necessarily at the beginning.

To me this was quite surprising and if I understand this correctly their NRR is subject to bias from few customers that ramped up above $50k ARR from a much lower amount. This IMHO makes this NRR cohort completely meaningless, but I would like to hear if there are differing thoughts about this out there. In any case, even if I remove Monday’s >$50k ARR customer NRR from the picture, Monday is still an incredibly strong company with strong revenue growth paired with high gross margins and a clear trend of improved leverage - in a space where especially the latter seems to be extremely difficult for their competitors.


Crowdstrike:

In my August recap I wrote:

Another positive in my mind was that subscription customer growth stopped decelerating just shy of 10% QoQ growth. This number was going down from a high of 17.6% now for 5 consecutive quarters. I was wondering if Crowdstrike was more focussing on larger customers and that this was the reason for this slow down, yet in answering an analyst question along those lines, “we found that, hey, we got new logos coming in from large companies. We have new logos coming in from smaller companies. And as typical, the velocity is driven more by mid-market and SMB.” While this isn’t the clearest answer you could imagine, I interpret it as being neither a focus on SMB nor on large customers.

Watching this video that @rdutt shared with us: https://www.cnbc.com/video/2022/09/26/crowdstrikes-modular-format-appeals-to-customers-facing-economic-downturn.html

we got some slightly new color here (in addition to some other good points):

Kurtz: “The deals [in the last Quarter] tended to be a bit bigger. (…)” Now he didn’t specify whether those are deals with existing customers or new customers, but I thought that was a noteworthy piece of information.


Wrapping it up

Now that earnings season is over and we have all the facts, I am very pleased with my current top-6 positions (SNOW, DDOG, CRWD, ZS, NET and S). Overall, my confidence in those six companies which make up 75% of my portfolio has never been higher. I still have some open questions about my bottom-6 positions, but I am in no rush making any significant changes here and for now, I will just wait and see what the future brings.

Thanks for reading and I wish everyone a great October!

106 Likes

Hi Ben – I just wanted to say I think your analysis of the $50k+ customer NRR is spot on. Kudos for running that down with IR!

Also, as Justin (thesea) noted here, Monday op loss guide - #7 by thesea, the $50k+ customers only make up about 22% of Monday’s revenue. (Guess that makes sense…it’s less than 1200 customers out of 150,000+ total customers MNDY has. Their average customer spends just around $3,000/year with Monday.)

200 customers crossed the $50k mark this quarter, 167 last quarter, and 183 the quarter before that…and it was 143 and 135 the previous two quarters. A year before that they were adding just a few dozen a quarter, so the hypergrowth appears to be leveling off a bit. Who knows if it stays around 200, or even goes back down? To try to form an educated guess I think we need to ask, what’s driving the big spenders?

Even if we accept that the TAM broadly defined really is $56b or whatever, how much of that is already taken? Monday says the deals they win are mostly greenfield – they’re not cutting into the billions of the TAM already being spent with Atlassian, Salesforce, Smartsheet, Asana, and dozens of private companies – so maybe the (green)field just isn’t that big.

How else do you explain the incredible slow down?
Q3 2021: 95% YoY revenue growth
Q4 2021: 91%
Q1 2022: 84%
Q2 2022: 75%
Q3 2022: 66% (includes a 5% beat)
Q4 2022: 60% (this would put them at 70% for the year – they’re guiding to 63% now)

Would be interested to see some thoughts on this from Monday bulls.

Bear

27 Likes

I don’t think your NRR calculation is correct. Let’s take your scenario 3 for example and the way I understood how it should be calculated according to monday’s definitions in the press release:

Last year:
399 cust @ 51K ARR = 20.349K ARR
1 cust @ 1K ARR = 1K ARR

Total of 20.350K

Current year:
399 cust @ 60k ARR = 23.940K ARR
1 cust @ 60K ARR = 60K ARR

Total of 24.000K ARR

NRR = 117.9%

12 Likes

I think you are correct, @Rubenslash!

This is from their F-1 filing with the SEC:
We calculate Net Dollar Retention Rate as of a period end by starting with the ARR from customers as of the 12 months prior to such period end (“Prior Period ARR”). We then calculate the ARR from these customers as of the current period end (“Current Period ARR”). The calculation of Current Period ARR includes any upsells, contraction and attrition. We then divide the total Current Period ARR by the total Prior Period ARR [just like @Rubenslash did] to arrive at the net dollar expansion rate.“, page 8, https://d18rn0p25nwr6d.cloudfront.net/CIK-0001845338/c17b48e6-deb4-4726-bfc3-b293e5a42ce4.pdf

And this way of calculating it makes much more sense. phew, I kept telling myself that the way I calculated the NRR can’t be right unless you want to make it look bigger than it should be. So I am glad we worked this out. That’s why I love this board!

Thank you!
Ben

13 Likes

Am I the only one who lost a bit of sleep over this answer from Monday’s IR? I agree with @PaulWBryant and @SlowAndFast that their >$50k NRR seems problematic. And I also agree with @Rubenslash’s calc above.

However, the reason this is so problematic, is because it is either incorrect, or they are overstating their NRR.

Here are the details. Assume the following:

Yr 1: 10 customers with on ave $60k ARR each = $600k ARR - cohort a
Yr 2: Of those same cohort a customers, only 8 are left; they lost 2 customers. And of those 8, 2 now have ARR of $40k and 6 have ARR of $100k = $680k ARR for the 8 that are left of cohort a.

→ NRR is 680/600 = 113%.

This is correct: you take a cohort and see what happens with that cohort over the course of a year. You include upsells (the 6 customers that went from $60k to $100k), contractions (the two customers that went from $60k to $40k) and attrition (the two customers they lost). This is what their definition in their 20-F says and is the correct way and the way our other companies also does the calc.

Now let’s do what IR says they do. Take exactly the same data and start by selecting only those customers that have ARR>$50k at the end of the period.

There are only 6 customers at the end of Y2 with ARR>$50k from cohort a. Yr2: BUT we also now have 2 customers with $60k ARR at the end of Y2 which were not part of cohort a as they had ARR of $10k each in Yr 1. Now we select all of the customers at the end of Yr2 with ARR>$50k. The 6 from cohort a and the 2 new ones and look at their ARR in yr 1 and compare that with yr2.

Yr1: 6 cohort a customers: $60k ARR x 6 = $360k
Yr1: 2 customers, not part of cohort a: $10k ARR x 2 = $20k
Yr2: 6 cohort a: ARR $100k x 6 = $600k ARR
Yr2: 2 new customers with $60k ARR each = $120k ARR

Yr 1 total ARR for customers who had >$50k at the end of yr 2: $380
Yr2 total ARR for customers who had >$50k at the end of yr 2: $720

→ NRR 720/380 = 189%

If they select only those customers with ARR>$50k at the end of the period by definition they are not including any attrition/churn and also not including any customers who contracted to below the threshold!!

So either their IR gave an incorrect answer, or their NRR for >$50k ARR customers is overstated.

-WSM

12 Likes

Thank you for helping with the math @wsm007! And I agree, IR’s answer was problematic and unclear (maybe if a customer drops below 50k or even churns they still count that). Sadly, I don’t expect more from IR. But call me a cynic.

But whether that’s true or not, you could still have 200 customers going on average from $10,000 to $75,000 or so this quarter (maybe some went from $3,000 to $200,00 or something and some just went from $30,000 to $55,000 or whatever). But the point is, 200 measly customers move the needle pretty hard when the whole $50k cohort is 1,160 customers.

But I’m worried Saul will call these trees. Let me be clear about the forest: it doesn’t matter what the NRR is on 1,160 customers. It doesn’t matter what the NRR is on 22% of revenue! Monday’s management talks about this $50k+ group because the numbers sound good. It’s window dressing.

I’m seeing revenue slow too dang quickly and I just don’t see how Monday will have a different fate than Smartsheet or Asana.

Bear

15 Likes

I would assume that their CFO is the one who calculates this statistic, not the public relations guy who works in Investor Relations, and thus I assume that they do it correctly. I mean, REALLY guys…

As far as the NRR as a whole, for the whole company it is over 125%. For customers with more than 10 employees it’s over 135%. (That means excluding the tiny 3 or 4 employee companies). What the heck are you worrying about?

On the other hand, Bear’s concern about the business slowing too rapidly is a valid concern. If that continues it would be a serious worry. They are still at 75% growth this last quarter, and while they are guiding low, they raised their annual guidance by 9 points in one quarter last quarter, and there are other indications that they are growing like mad, like tripling their US headquarters in NYC and their European headquarters in London.

Finally, as far as their $50k customers, they do have two and a half as many as a year ago any way you look at it, and if they are only 22 % of revenue it’s because their revenue from smaller customers must be growing rapidly too.

I might be wrong about all of it, but that’s how I see it.

Saul

28 Likes

Fair enough Saul - enough nit-picking on the NRR, but to be fair IR answers generally don’t get answered by the PR guy; I emailed Monday IR before and got an answer from the head of investor relations. Perhaps he didn’t answer Ben but if he did, he should really know that stuff.

But let me try to answer your questions about customers. Are small customers growing rapidly? Are their >50k customers very big? I tried to figure some of it out based on what they gave. Hopefully the following helps.

They don’t give us total ARR but we can estimate it from revenue. And they did tell us that their $50k customers were 22% of ARR in Q1. But unfortunately they don’t tell us that % going back, nor did they give it in Q2.

They did tell us a couple of other tidbits, which I have used to try to figure out their customer dynamics in more detail, to answer the question: has the average size of their $50k customers grown by a lot? Did smaller customers grow by a lot?

Let’s start with the smaller customers.

We can estimate total ARR by multiplying the last Q’s revenue by 4. And they gave us the % of ARR from customers with 10+ seats (which include the $50k+ customers). They contributed 53%, 63% and 72% of ARR in 2019, 2020 and 2021. So we can calculate the ARR growth of <10 seat customers and >10 seat customers as follows:

2019 2020 2021
Est total ARR (last Q revs x4) 106 200 382
% of total ARR from 10+ 53% 63% 72%
10+ seats ARR 56 126 275
<10 seats ARR 50 74 107
10+ seats ARR growth 124% 118%
<10 seats ARR growth 48% 44%

→ So small customers are not growing very fast, in fact quite the opposite.

They also tell us that:

  • $50k+ customers were 22% of ARR in Q1 2022.
  • $50k+ customers were 76 end 2019, 264 end 2020, 793 end 2021, 960 Q1 2022 and 1160 Q2 2022.
  • Therefore $50k+ customers grew by 247% in 2020 and 200% in 2021
  • ARR from $50k+ customers grew by 297% in FY2020 and 239% in FY2021 (from the F-1)

→ This means that most of the growth in ARR contribution from these $50k+ customers came from the growth of the cohort, and not from the growth of the ARR of customers in the cohort (247% customer growth led to 297% ARR growth and 200% customer growth to 239% ARR growth).

We can estimate total company ARR as $434m in Q1 2022 and from the 22% can estimate that an average $50k customer has $99.5k in ARR ($434m*22%/960).

Working back from this, and assuming a continuation of the ARR growth trend which they gave us into Q1 and Q2 of 2022, we can estimate the ARR contribution from $50k+ customers for Q2 2022 as well as for prior years, and to estimate the average ARR/$50k customer.

With a lot of assumptions, I get to this:

Screenshot 2022-10-05 at 15.56.03

→ My conclusion from this is that the average ARR/50k customer grew and continues to grow at around 13% pa. So they are adding a lot of customers to the cohort, but they don’t seem to be generating huge customer ARR growth beyond that - they probably don’t have many/any $1m customers.

So what?

Tbh I struggle to pin all of this down, as it seems like there are a number of gaps (how many >10 seat customers do they have and why don’t they tell us? What is the ARR contribution of $50k customers over time, and why don’t they backfill the info and give an update in Q2? Are they having success with upselling/what % of customers are on pro ($16/seat), what are the average number of seats per cohort, etc.)

Still, it looks to me like the following dynamic is at play: they have a huge base of extremely small (sub 10 seat) customers, and these customers aren’t growing their revenue a lot (±40% pa and declining). From this base they upsell to >10 seat customers, which leads to a nice uptick in ARR of this cohort (>100% but also declining), and then they again upsell those to $50k+ customers - adding seats each time, leading to the biggest growth in ARR (±200%). But the $50k customers never get really large and there is limited expand motion beyond adding seats. They don’t add additional modules, charge for consumption and their upsell to higher tiers may not be as successful as they would want. So the only real upsell motion they have is adding seats. And because their big base of very small customers are not growing exceptionally fast (±30-ish %), and the biggest customers they have, have a ceiling of ARR contribution per customer, their growth is starting to drop quickly as the ARR growth in all cohorts seem to be declining.

Happy to have other views, but for now I’m steering clear.

35 Likes