Cloudflare conference call notes

Investment Analysis Clubs Saul’s Investing Discussions
1

This was an a great quarter for Cloudflare, and they gave a ton of detail in this call. To give a comparison, there was 25 minutes of prepared remarks compared with 10 minutes of remarks from Fastly. Also they mention Fastly a few times in the call as an egregious pricer who gave unexpected bills to many customers, and that they have gotten a lot of feedback and loyalty about the importance of a steady bill.

Matthew Prince CEO

  • revenue 99.7M 48% YoY
  • strength in all customer sizes
  • 100k ARR customers saw 65% YoY growth
  • strong growth across existing customer base
  • covid concessions lower than expected
  • raising Q3 and 2020 guidance
  • predictable and consistent business model
  • 95%+ revenue is billed up front on subscription basis
  • usage basis is more volatile and other vendors are coming to Cloudflare for consistent pricing, this is building loyalty from customers and making their revenue model more predictable
  • covid forced vendors into 2 categories: 1) nice to have 2) must have
  • Cloudflare is a must have service
  • customers which were previously slow to adopt cloud are joining rapidly now
  • Q2 strength in Europe, industrials, and small businesses (unexpected strength)
  • nimbleness of go to market team and shortness of sales cycle
  • total increase in global internet traffic, although saw plateau in Q2
  • cyber attacks are up 37% versus Q1 and Q2
  • new customer growth in mitigated atttacks was 63%
  • May was the busiest month ever for DDOS attacks
  • June saw one attack last 4 days, peak at 750M packers per second, and the customer was not even aware until Cloudflare alerted them
  • hired 257 team members in Q2
  • 47,000 applicants for jobs, up 757% YoY, 96% acceptance for offers, and 99% within sales department
  • progress towards profitability, 2,000 basis points to operating leverage YoY
  • gross margins > 76%, within long term target model
  • European financial services company, 3 year deal at 450k per year with Cloudflare Teams (replacement for VPN/Firewall), chosen for future proof roadmap, can further expand this customer as browser isolation technology is rolled out in 2nd half of year
  • US based industrial, 350k per year, WAF solution replacing legacy on-prem
  • Leading ID & access management company (Okta maybe?), 2 year deal at 500k per year, moved workloads away from AWS due to better speed and flexability of Cloudflare platform
  • Born in the cloud gaming platform signed additional contract from 1.8M annually, using Cloudflare Workers
  • Cloudflare workers launched three years ago, true Turing complete serverless platform that allowed devs to build sophisticated apps and run them edge to edge through entire network
  • 10% of Cloudflare traffic is through Workers and 20% of new customer deals include Workers
  • 22,000 devs wrote their first app using Workers up 44% YoY
  • not satisified being a niche edge computing platform, want to be the go to serverless platform for devs/applications
  • upgrading workers: CPU, pricing 75% less than AWS Lambda, supports many languages: JavaScript, C, C++, Python, Rust, Scala, Cobal
  • customers picking Workers because fastest, consistent, secure, cost effective

Thomas Seifert CFO

  • less headwinds than expected
  • US revenue 49% of total, up 46% YoY
  • International revenue 51%, 50% YoY, driven by EAMEA with 62% YoY growth
  • 3 million total customers, 40% YoY
  • shifted to revenue based KPI from billings
  • total paying customers 96,000
  • added more than 80 large customers (100k+ ARR) this quarter, 637 total with 100k+ ARR, up 65% YoY
  • half of the 637 large customers are on the platform less than 1 year
  • DBNRR 115%, 2% decline, churn in non-strategic accounts in asia pacific (I believe this unusually low number is from the bottom tier customers who switch on/off the solution)
  • Q2 gross margin 76.8% down 150 basis points, long term target is 75-77%
  • able to absord 30-40% increase in traffic levels with little impact to gross margins
  • operating expenses decreasing
  • increasing headcount, 1535 employees
  • Sales & Marketing 48% of revenue, R&D 15%, G&A 17%
  • seeing operating leverage in business
  • net loss per share of 3 cents
  • 1.1B in cash
  • covid impacted segment which is about 10% of customers, grew sequentially but decreased as overall percentage of revenue
  • uptick in sales productivity
  • RPO 274M, 18% growth QoQ, 58% YoY
  • 3Q guidance: 102.5-103.5M, 39-40% YoY growth
  • 2020: 404-408M, 41-42% YoY, projecting -.18 to -.17 earnings for year

Questions section

  • growth in large customers
  • 15% of Fortune 1000 using
  • replacing existing on prem solutions
  • Cloudflare Unbound, edge computing for 3 years now, in this time learned devs don’t want to prematurely optimize, fast is the last thing you need (compares to Maslow’s hierarchy of needs and with fast being like self actualization)
  • true serverless application, includes compliance, data localization, 100 countries, can write to one platform and still be compliant
  • continue to move up market
  • companies developing IP on platform which they can resell like IBM
  • channel parterns can sell the full suite
  • china revenue low single digits
  • selling to US companies which sell in the Chinese market
  • JD cloud center promising, 2 categories, Chinese companies selling outside China has more visibility and strength, companies selling into China is more uncertain
  • JD turning on more in 1st half of 2021
  • Baidu is impressive partner as well
  • Migrating customers to use more and more of platform
  • Analysis of bot traffic, then offering solution to the customer
  • compelling ROI for customers on Cloudflare investment
  • typically customer land with one product and expand in following Qs
  • Cloudflare Teams on per seats basis, 2,000 companies signed up and free till September, many will convert
  • Magic Transit sleeper product, directly connects your product into Cloudflare network
  • Big fan of kubernetes but it’s just one step to tech nirvana of getting rid of techops team
  • Goal is to be able to write code and have it scale up or down automatically
  • Workers will enable the first company to become a billion dollar company without ever needing tech ops support
  • Workers is winning workloads from AWS/Azure, because cheaper, faster, easier to use, compliant
  • Scaling up containers on AWS Lambda is poor experience, Workers more effecient, nextgen, zero nano-second start time, pricing incredibly competitive and less overhead
  • largest enterprise companies adopting and backlog is growing
  • 9 out of 10 of the largest expansion deals were with existing customers
  • customers saying they can no longer delay on migrating to the cloud

Overall impressions and comparisons
This was easily the most impressive conference call I have heard this quarter from a company yet. Seems covid impacts were minimal and big customer upticks. I cannot get over the contrast with the Fastly call where they gave scant details about any of the products and spent half the call discussing TikTok. While usage may be helping Fastly revenue in the short term, it’s not beneficial long term to have a pricing model which can shock customers. I have some Fastly shares as well, but after this call I am much more optimistic about Cloudflare going forward.

Also comparing to the DataDog call, there was little mention of cloud slowing down. It was spoken more of reaching a plateau but still seeing growth.

The mention of the 2 categories of nice to had versus must have made me think of AYX, and I believe AYX is a nice to have solution.

61 Likes
2

I’m confused about how the market is reacting to NET after posting such a great ER - Today NET is down 8%. Any ideas what it is that the market is not liking?
(I cannot see anything to dislike in the results or in the conference call)

1 Like
3

Looks like sector rotation to me, happens once in a while. Almost everything SaaS/cloud-related is down while the Dow is up significantly. Wouldn’t worry too much about it (fyi: my portfolio had a 20% haircut in a week, but I was up a lot for the year).

Good luck!

Paul

6 Likes
4

Thanks Paul,
I wasn’t worried at all - I was only curious to know; in fact on the contrary I will probably take advantage of this to buy more NET.

5

I took advantage of Mr market and loaded significant position in NET with the proceeds from LVGO a few days ago.

Zoro

2 Likes
6

I’m confused about how the market is reacting to NET after posting such a great ER - Today NET is down 8%. Any ideas what it is that the market is not liking? (I cannot see anything to dislike in the results or in the conference call)

Hi moneyspin, I agree with you. I thought the earnings report AND the conference call were excellent. I think it’s just down because lastly sold off a bunch last week and Cloudflare didn’t. I sold some more Alteryx today, and also sold Fastly at down 1.5%, to buy NET at down 9.5%. This equalized my Fastly and NET positions where before I had much more Fastly. Now they are both about 9% positions.

Best,

Saul

54 Likes
7

Cloudflare looks very attractive on a variety of metrics. Some areas that give me pause, however, are their operating cash flow/sales is strongly negative, while most of the companies we follow here are slightly positive (Fastly, Twilio, Roku, Alteryx, Sea Ltd) to very positive (Zoom, Crowdstrike, Datadog, MercadoLibre). I also am shocked by their capex/sales, ~20%. This is insanely high capex relative to other companies we follow - most spend less than 10% of sales on capex, and many spend 5% or less. Why is Cloudflare spending so much on property, plant, and equipment?? They don’t seem like a truly “asset light” company to me.

7 Likes
8

Cloudflare (and Fastly) have to build out their network, so they do have hardware costs. You can’t compare them to pure SaaS like ZM, DDOG, etc when it comes to GM. In theory Fastly should spend less in the long run since they plan to build less PoPs.

Just to share some changes I made today to my portfolio as well, now that things are starting to settle a little (I think):
I also sold out of AYX (only 2.5% of port by today after selling a bunch before the drop, like many did here, and then the drop took me down to this final size) and shaved some off SHOP (dropped from 9% to 7% of port). Added to DDOG, FSLY, and NET. The new purchase proportions were like 60%, 25%, 15% of the cash spent (not % of port). I chose these proportions mostly based on degree of correction and rough value, BUT I may bring NET up like Saul did. I just want to understand a bit more about Workers and how it compares with, or lives along side, Fastly’s Compute@Edge. I don‘t like how NET advertised this zero-second cold-start when it doesn’t apply to most sites (I’m not even clear if this is CDN or edge computing related), for example, so I want to know some more about the tech that will compete for edge processing. Please don’t reply to this last point here though. Off topic for this thread.

4 Likes
9

“Please don’t reply to this last point here though. Off topic for this thread.”

Funny enough this is wrong. This IS a NET thread after all. Feel free to comment :wink:

The “zero cold start” thing is indeed about edge computing. I posted about it, and what I now perceive as the difference between Fastly and Cloudflare, here instead:
https://discussion.fool.com/some-initial-answers-and-responses-l…

1 Like
10

I also am shocked by their capex/sales, ~20%. This is insanely high capex relative to other companies we follow - most spend less than 10% of sales on capex, and many spend 5% or less. Why is Cloudflare spending so much on property, plant, and equipment??

They are building data centers (or renting space) to be able to deliver content faster. In this call and previous calls they’ve mentioned that hardware costs are falling while speeds are getting better.

I don‘t like how NET advertised this zero-second cold-start when it doesn’t apply to most sites (I’m not even clear if this is CDN or edge computing related), for example, so I want to know some more about the tech that will compete for edge processing.

The zero second start time is for Cloudflare Workers which is their server-less product. This is the one that competes with AWS Lambda, and it costs 75% less and is basically a superior product, which is why I think they are so excited about it. What I’m less clear about is if server-less and edge computer are used interchangeably. I thought edge computing was programmable logic within the CDN, and server-less was separate, but sometimes they are used as synonyms.

After getting more familiar with Cloudflare and Fastly I believe they serve somewhat different markets. Cloudflare has 3 million total customers, 90,000 of which are paid customers, 637 with 100k+ ARR. Fastly has 1951 total customers and 303 with 100k+ ARR and they mention that the average spend of these 303 customers is $716,000. Unfortunately Cloudflare does not break out the average spend statistic, so it is hard to compare on the number alone.

The biggest standout to me is that Cloudflare added 80 100k+ ARR customers, while Fastly netted 6 100k+ ARR customers. What I’m trying to reconcile is how Fastly has DBNRR go up to 130%+, and Cloudflare had DBNRR go down two points to 115%. I’m guessing it is just existing customers doubling down on Fastly, and Cloudflare small customers have a ton of churn. Looks like Fastly revenue growth mostly comes on expanding usage while Cloudflare’s comes on expanded customer count.

From what I have heard the free tier of Cloudflare provides a lot of value and out of the box. Fastly seems more geared to enterprises that need speed and is too expensive to just test out without massive scale. Cloudflare has a wide range of products, and Fastly has a more targeted suite of products. Fastly has ~600 employees and Cloudflare ~1500.

When Cloudflare is down, it seems like the half the internet is down. That seems consistent with having 3 million customers and contributes to have an endlessly deep pipeline to convert free members to paid, and convert existing paid to 100k+ ARR. Keep in mind supported 2.9 million free customers is a big cost, but the company is still able to perform well despite giving away so much.

I am also wondering longer term about the companies that have a usage model versus fixed costs. While it is boosting share prices massively for companies like Twilio and Fastly, it leaves the door open for others to compete on pricing.

34 Likes