Cloudflare is living on the edge (Q4 report)

Investment Analysis Clubs Saul’s Investing Discussions
1

My favorite rock bands loved to live on the edge - drugs, sex, fast cars and destroyed hotel rooms. Some of them (Guns and Roses) got too close to the sun and melted away. Others (Aerosmith, U2) had staying power to become some of the greatest of all time. Cloudflare reminds me of these musicians that produce awe inspiring work yet could crash and burn if one or two or three things go south.

I am trying to understand how a technology company, on the forefront of the adoption of cloud computing and the growth of internet, can deliver such consistency in revenue, gross margins and market share, while having anemic cash flows.

I realize that they are investing in R&D and sales; they are expanding their product and services suite and grabbing marketshare. However, at some point, you have to stop giving away the shop and make money for the shareholders - who, btw, are your primary stakeholders.

Read the opening comments by CEO, Matthew Prince - “we are not in a rush to be significantly profitable”…“we have done something wrong if we beat significantly on EPS”……smh

Investing in NET takes a bit of faith in the future. Prince promises that they are dialed in and H2 2022 will be better. My patience is running thin.

Here is what stood out to me in NET’s Q4 earnings report: (Link to detailed financials map below)

Revenue growth - Q4 revenue grew 54% YOY and 13% sequentially QOQ. 2021 annual revenue grew 52% YOY. NET guided for $206M in Q1 2022 and $931M in full year 2022. Deferred revenue grew 113% YOY and 27% QOQ to $117M - this metric has accelerated 4 quarters in a row now! NET has beat guidance by 4-6% for several quarters now - which is a sign of a deep revenue pipeline, effective sales force, DBNRR strength and mature fiscal mgmt.

Profitability - Gross margin has been steady in the high 70s for 12 quarters now - another sign of a highly efficient product set and strong pricing power. EBITDA is trending towards positive for 3 consecutive quarters. However net profit / loss remains lumpy and operating margin remains high in the -20s.

Cashflow - Cashflow remains all over the place. Operating cash flow increased to $41M for Q4 and $65M for the year 2021 (482% YOY). Free cash flow also jumped to $9M for Q4 and -$43M for 2021 (53% improvement). The roller coaster ride in free cash flow margin shows where NET needs to improve, especially in the light of the high debt to cash ratio.

Market share - Customers like NET’s products and they keep buying more. DBNRR has improved for 6 straight quarters and sits at 125%. Customers with > $100k contracts have been growing by 70%+ for 4 quarters and by double digits for 7 quarters in a row. Customers with >$500k contracts grew 70% YOY and $1M+ customers grew 75% YOY. And I love the global growth trend.

Debt - remains high at $1,294M versus cash at $1,822M. This is a bit too close for my comfort, especially given the uneven cash flows. They should have no problem raising funds if needed, however this leads to further erosion of shareholder equity and keeps the share price in check.

Product - NET is building a better internet, enabling broader, quicker migration to the cloud and improving the consumer’s experience. 250+ cities, 100+ countries, 10,000 networks, touching 95% of the world’s population - the numbers are easily overwhelming. They keep the product improvements rolling, more frequent than any other company, supported by the mass free testing by the large developer community.

My conviction in NET stays the same with this Q4 2021 earnings report, however like I mentioned above, my patience is running thin. I re-scored the company using my analysis method and their score increased by 1 point aided by the positive trend in EBITDA improvements - a small sign of hope.

Elevator pitch - I own NET (6.5% position) because they are the most innovative cloud and internet company that I am aware of. I am intrigued by their edge computing roadmap, cyberthreat (zero trust) services, IoT and 5G opportunities. At some point, lots of everything else becomes a commodity and the telcos, hyperscalers, AWSs etc. start undercutting them.

I boarded the Cloudflare train in 2020. I am not fully convinced that they can get me to the promised land.

Beachman (Beachman.substack.com)

Detailed financials map https://twitter.com/Iwannabeontheb2/status/14975772637973463…

46 Likes
2

Without going into technical details, see Hhhypergrowth.com or Softwarestackinvestimg.com, I’m unable to provide more than perhaps some perspective? The following perspective may sound outlandish. I think Cloudflare’s third act will be.

In Cloudflare’s recent CC, CEO/Founder Mathew Prince broke out some milestones for Cloudflare. As I’ve recently stated Cloudflare and Zscaler go head to head regarding Zero Trust. In the Call, Prince stated that Cloudflare is currently positioned well and is obtaining the lions share of its revenue from their zero trust solution. But, it’s Cloudflare’s third act wolf capturing the astronomical TAM of the programmable edge-Cloud with their Workers product that will, IMO, catapult Cloudflare into the next level of Hypergrowth. Their movement toward this third Hockystick, when Cloudflare’s Workers via their fourth Cloud is why I’ve overweighted Cloudflare in my portfolio compared to Zscaler. How soon? Well in the CC, Prince said they’ll be adding their 1 Millionth Developer this year. What’s the tipping point to when they’ve sufficiently built out ecosystems and produce an avalanche of increasing revenue growth? Dunno. I’m not waiting for past numbers to be presented when the market is typically six month ahead. I’ve recently increased my allocation to 17% making Cloudflare my highest allocation.

I believe Worker will be every bit as large a revenue growth driver for Cloudflare as Data Sharing will be for Snowflake. The strong Network Effect Moat will likely be much smaller for Cloudflare than what Snowflake has created, just due to the differences in company culture, IMO. Cloudflare has done everything right to prepare and now entire companies are built on their Workers platform (this is sticky). Yes, Cloudflare’s third flywheel of Hypergrowth was a massive undertaking and has taken a long time to get going.

I hope that this didn’t come across like a pump piece.

Best,

Jason

35 Likes
3

Great thread.

BeachMan2115 writes:
I realize that they are investing in R&D and sales; they are expanding their product and services suite and grabbing marketshare. However, at some point, you have to stop giving away the shop and make money for the shareholders - who, btw, are your primary stakeholders.

CEO/Founder Matthew Prince has larger aspirations. The mission statement for Cloudflare isn’t “make money for the shareholders,” it is “to help build a better internet.”

My patience is running thin.

Then maybe NET isn’t right for your portfolio. That’s fair. There are enough other high growth stocks in which to invest. Saul himself passes on some that are discussed frequently here. Cloudflare’s mission is going to take time and require large investments. My view is that Cloudflare has a sustainable model for growth, so I’m sticking with it.

WillO2028 writes:
I believe Worker will be every bit as large a revenue growth driver for Cloudflare as Data Sharing will be for Snowflake.

I disagree. More in a bit…

The strong Network Effect Moat will likely be much smaller for Cloudflare than what Snowflake has created, just due to the differences in company culture, IMO. Cloudflare has done everything right to prepare and now entire companies are built on their Workers platform (this is sticky).

I agree with most of this, except the degree of influence of company culture on Network Effect or Moat.

The Network Effect for Data Sharing should be obvious, as it’s a network sharing feature. The more companies that have data to share put that data on Snowflake’s platform, the more companies that want data will go to the platform to find data they want/need. The more companies look to Snowflake to find data, the more companies that have data will put their data on Snowflake to be found. It’s kind of like EBay in the early days, and once EBay established itself as THE marketplace it snowballed. Pun intended.

This has very little to do with company culture. Workers doesn’t get the same kind of supply/demand Network Effect, except that as more companies use Workers and are happy, the more Workers gets discussed and advocated among the programmers. That’s not nothing, but it’s not the same kind of company use directly driving other companies to use Network Effect.

And in just general terms, I think the market for Data Sharing is much larger than the market for Edge Computing. Now, maybe that’ll change in the future, but in my view every company wants relevant data to analyze to improve their business. That data can range from customer behavior to market dynamics to all sorts of things. But, Edge Computing is somewhat esoteric in that the only reason to compute there instead of centrally is to save latency, which is the delay in sending signals to a further central location than a closer edge location. But, there are and will continue to be many many applications with a few milliseconds of latency means nothing compared to the burden of programming for the edge.

Just think about a site login. Sure, it would be great to log people into your system faster, but if you try to do this at the edge then where do you keep the repository of accounts, usernames and passwords? If you keep it at your central location, then the edge computer needs to talk to the central computers and you’ve lost your latency advantage. If you cache the information, then you need to do so at each edge location, which means more redundant data spread out among many edges that needs to be kept up to date. It’s harder to do and so the question becomes just what applications are worth the benefit of shorter latencies? And then if you’re supporting mobile applications, what happens as your user moves from edge to edge during a session (like he’s in a car or on a train)? Now, don’t get me wrong, there are applications for which processing can and should be done more locally, but even then results are often sent back to a central location for later analysis.

The basic building block for Edge Computing is typically serverless computing, which is a misnomer because there is a server involved, it’s just that the programmer doesn’t need to worry about it. That’s great in many ways, but it also imposes limitations on what can be done. For instance, no session information is retained since each API call might go to a different physical server that doesn’t have the history data from the last API call. And if you try to store history, then you’re going to a central repository for that store, which loses the edge latency advantage.

And then there’s variations in what “the edge” actually is. In many cases, the end device itself becomes the edge. Think about Apple’s face recognition phone unlock. The facial data is kept on the edge device (the phone) and the processing to determine a match is also on the device. This maintains privacy, and since Apple’s phones have pretty great computing power, increases performance. But, it’s not a use case for Workers since it doesn’t use POPs (Points of Presence), it’s all on the end device. So, not even all Edge Computing use cases can be handled by Workers.

When talking about moats and stickiness for Cloudflare’s Workers, I see that as both a positive and a negative. Sure, once on Workers, as long as Cloudflare keeps the performance up, there will be little incentive to move off of Workers, but that same re-programming effort is also a disincentive to move onto Workers in the first place. There are more than a few options for computing at the edge today (from Amazon to Fastly and others), and all sorts of variations in how much is done where. Amazon, for instance, offers a variety of Edge Computing services, from Outposts and Local Zones to Wavelength and Snowcone (doesn’t compete with Snowflake). Amazon even offers something called IoT Greengrass, which puts some Amazon capabilities on your IoT devices to enable the kind of end device processing that needs no data connection, as well as APIs for communication to either Amazon POPs or central computing resources.

If you’re not confused about what Edge Computing really means, then either you’re an expert in the field or you don’t really understand all the possibilities. Essentially, programmers are constantly looking for ways to optimize workflows and there’s now a whole gamut of not just possible places on which to perform compute, but way to communicate data between those possible places. Workers is great for POP applications, but Edge Computing is far more than utilizing POPs, as POPs don’t make sense for every workflow - or I dare say, most workflows.

OTOH, Data Sharing between organizations not only has a snowballing network effect, but also doesn’t appear to have any kind of usefulness or applicability limit. All kinds of organizations will want data to analyze and most are willing to pay for that data, and companies that have data are willing to share that data for a profit when it’s super easy to do so, and they are able to control how much of that data is shared, and to whom - all which Snowflake provides.

So, Workers is fine and good, but in my view the size of the revenue opportunity there for Cloudflare is nowhere near the size of the revenue opportunity of Data Sharing for Snowflake - under just about any measure you can think of.

46 Likes
4

Smorg,
Any thoughts on how to think about data sharing via Snowflake vs 1st party data anaysis, which i think is what AMPL was focused on?

I admit to never digging into SNOW too much just due to valuation. So i was curious if they partnered or worked on DOCN, but it looks like they are currently big 3 clouds only:
https://docs.snowflake.com/en/user-guide/intro-cloud-platfor…

DOCN has a new managed service for MDB, so i was curious if expecting SNOW to bother with a smaller smb-focused cloud like DOCN would even make sense:
https://docs.digitalocean.com/products/databases/mongodb/

You have a link to a laymans view of what use cases look like for snowflake? Seems like data warehouses have been around for a while, so i am not sure what snowflake is doing that is so unique.

Looking at a random article from 2015, it feels like it could be written today. https://www.ironsidegroup.com/2015/02/02/top-5-trends-in-clo…

Dreamer

3 Likes
5

Thanks for sharing your thoughts and nice rundown of areas of interest!

I’d like to address the points about earnings importance…at least the way I think through it:

I realize that they are investing in R&D and sales; they are expanding their product and services suite and grabbing marketshare. However, at some point, you have to stop giving away the shop and make money for the shareholders…

We are not getting dividends so the only way we make money here is when companies increase in value so that the price goes up for the day we eventually sell. For this reason, if they fail to reinvest cash to fuel growth, we actually make less money, not more.

Investing in cash-generating companies for dividends is great if the investment strategy is to earn passive income while, hopefully, achieving equity growth that beats inflation; the ever-depreciating value of the currency or ever-increasing cost-of-living. There is a problem here though. It implies that company is either so big it can no longer grow at a significant rate regardless of re-investment, or worse, simply don’t have any better ideas for using the cash for growth (growing teams/infrastructure, innovation, marketing, share buy-backs, acquisitions, leverage, etc). This is why we never see companies in the prime of their growth cycle introducing dividend programs. There is plenty to spend money on to keep growing. With great management, the spending has significant impact. With bad management it has less effect on growth, or they go the way of diworsification (misguided spend). This is why I feel one measure of good management is their ability to generate revenue growth: it means a lot of cash comes in and ends up spent wisely.

– A side note: This does NOT necessarily mean the growth percentage has to go up (acceleration). Even staying the same is a huge achievement, as each quarter the company is larger and it is harder to maintain a growth rate. In fact, I absolutely guarantee 100% of our companies’ growth will decelerate over time. This is the (investing) Law of Large Numbers. –

My point is just that I would love to see every one of our companies earn very little (while maintaining a “healthy” (sustainable) balance sheet of course). I want them to spend every top-line dollar on growing the business. This is, I believe, what Matthew Prince meant in that quote:

“we are not in a rush to be significantly profitable”…“we have done something wrong if we beat significantly on EPS”

…if they beat significantly on EPS it is likely a surprise that they made way more than they had budgeted to spend. No doubt a very nice problem to have, but still unplanned-for.

In fact, I do this personally too. I re-invest all of my earnings in growth (finance, personal or family). If I have enough safety-net, what does stashing one more dollar under the mattress do for me? It is the same for our companies. I also do not invest in dividends or fixed-income holdings for the same I don’t want a company to pay out dividends. I am after compounding growth.

(I realize this is in danger of going off-topic in to portfolio or financial management. Please do not reply to these aspects of this post. I was only using the concepts as color to explain why I agree with the comments by the CEO. …and I felt this thread, while containing some great info, felt like it started off on the wrong foot, IMO of course)

25 Likes
6

Thanks all for the interesting dialogue on this topic

WillO2028 wrote: Prince stated that Cloudflare is currently positioned well and is obtaining the lions share of its revenue from their zero trust solution.

Yes - this is of high interest to me as I try to forecast their revenue pipeline. Would be nice to see them breakdown their major revenue streams so that we have better visibility into how each one is accelerating or not.

Smorgasbord1 wrote: CEO/Founder Matthew Prince has larger aspirations. The mission statement for Cloudflare isn’t “make money for the shareholders,” it is “to help build a better internet.”

Herein lies one of my concerns. Matthew Prince runs a publicly held company that has shareholders. He is not running a charitable non-profit organization.

Smorgasbord1 wrote: Then maybe NET isn’t right for your portfolio. That’s fair. There are enough other high growth stocks in which to invest.

I might agree with you on this. At the end of earnings season, I will likely make a decision.

RafesUserName wrote: We are not getting dividends so the only way we make money here is when companies increase in value so that the price goes up for the day we eventually sell. For this reason, if they fail to reinvest cash to fuel growth, we actually make less money, not more.

I am not looking for a dividend for shareholders. What I would like to see is NET increasing its enterprise value by improving cash flows and reducing debt, all while continuing to invest in R&D (product) and growth (sales and marketing). I suggest that NET’s share price will stay relatively muted if they fail to do so, especially in the 2022-2023 environment.

Other cloud SaaS companies in similar or adjacent markets are finding the right balance.

See the comparison between NET and DDOG below.


NET  (expenses as % of revenues)   Q420   Q121   Q221   Q321   Q421
Selling General & Admin Expenses  68.97% 70.77% 68.83% 66.46% 67.16%
R&D Expenses                      27.60% 28.63% 27.13% 27.14% 31.90%

DDOG (expenses as % of revenues)   Q420    Q121   Q221   Q321   Q421
Selling General & Admin Expenses  43.31%  42.38% 37.93% 36.08% 35.57%
R&D Expenses.                     37.59%  39.03% 39.68% 41.09% 39.95%

DDOG is not “giving away the shop” which leads to lower SGA expenses…AND…it allows them to invest more in R&D (as a % of revenue).

These two companies were founded around the same time (DDOG 2010 and NET 2009) and are vying for adjacent cloud $ budgets of their enterprise clients.

They both have very similar gross margins. Its what they do with those margin $s that makes all the difference.

21 Likes