Why I bought FSLY

Investment Analysis Clubs Saul’s Investing Discussions
1

THE BASICS

What they do: SaaS edge cloud platform, designed to help developers extend their core cloud infrastructure to the edge of the network, closer to users. The purpose seems to be speed (as the name suggests), but I believe there are security features and implications as well. I have only begun to look into it, so consider this a not at all thorough description.
Mkt Cap: $4.0 Billion (95.4 million shares outstanding)
TTM Revenue: $218 million
PS ratio: ~18
Revenue Growth in most recent quarter: 38%
NRR most recent quarter: 133%
GM in most recent quarter: 56.7%
TTM Cash Flow: -$31 million in 2019, -$7.1 million in Q1 this year
Cash: $187 million

We can see that this is a very small company with a reasonable PS ratio, but that revenue growth is slower than we’d want (but wait). Gross margin is also low for a SaaS company, and they are losing a lot of money each year, but only a fraction of the cash they have on hand, so they’re not in trouble. But basically, the numbers to date look meh, so we would have to see some extraordinary trends moving positively to have much interest here.

Spoiler alert: that’s exactly what we’re going to see.

IMPORTANT TRENDS AND OTHER SPECIFICS TO HIGHLIGHT:

YoY Revenue Growth
Jun 2019 Q: 34%
Sep 2019 Q: 35%
Dec 2020 Q: 44%
Mar 2020 Q: 38%
Jun 2020 guidance: 52% - 56%

So this is a big time acceleration. They also raised full year guidance from $265 million to $290 million at the high end. They mentioned, As social distancing measures increased over the March period, we continue to see increased traffic across the internet and our platform, which certainly provides an additional boost to our results. They said a couple times that this quarter only included 2 weeks of the benefit (for obvious reasons). In short, this next quarter is going to be huge for them, and that trend should continue thereafter. Their 56% guide is surely conservative, so just imagine.

Gross Margin
Jun 2019 Q: 55%
Sep 2019 Q: 55.2%
Dec 2020 Q: 56.7%
Mar 2020 Q: 56.7%

They don’t provide guidance on gross margin, but it’s about to go up big time. Their language is very formal: Despite the uncertain economic environment, we remain confident in our ability to deliver incremental annual gross margin expansion as we continue to scale and deliver innovative security and edge computing solutions. But the translation is that margins are going to soar. There were a lot of questions about their huge guidance raise and what gross margin it implied. One analyst said when you look at the remainder of the year, the guidance seems to imply…close to 80% operating margins. I don’t know if he’s correct, but they didn’t argue with him.

INTERPRETATIONS AND CONCLUSIONS

This is an unprecedented acceleration for them. Fastly is built for digital transformation, so they’re thriving in this environment.

Their customers are huge for them. The CEO said like 10 times on the call that Fastly is the platform for innovators. Fastly is usage-based, and their customers are growing like mad, so Fastly grows like mad with them. Online shopping in Q4 is huge for them.

Their new CEO is a rock star. The founder and old CEO was on the call, but these guys are just night and day as communicators. They made a great hire. Listen to the call.

I’m psyching myself up about this company. The stock is down a little this morning because most things are, and also maybe because FSLY is wisely offering 6 million shares to raise capital. I’m taking advantage of the dip and buying a lot more.

Bear

93 Likes
2

https://seekingalpha.com/author/richard-durant#regular_artic…

Bear

I’d like to refer you to yesterdays article by Richard Durant which in my opinion is a clear exposition of FSLY and its prospects.

I read it as somewhat negative despite FSLY’s redeeming features. I’d certainly be interested in your perspective.

Also Beth Kindig had a good exposition recently on Edge computing and FSLY which I found helpful

2 Likes
3

Draj: I read it as somewhat negative despite FSLY’s redeeming features. I’d certainly be interested in your perspective.

[From the Durant article] Although these large customers are creating rapid revenue growth and the potential for increased spending over time as bandwidth usage increases, it has also resulted in concentration, with Fastly’s 10 largest customers generating 37% of revenue in 2018. Fastly currently has 297 enterprise customers (>100,000 revenue annually) who account for 88% of revenue.

Fastly’s focus on large, performance-oriented customers creates the risk of their largest customers integrating backwards to reduce costs and allow customized CDN solutions. Many large companies have elected to manage their own CDNs for cost or performance reasons, including Netflix (NASDAQ:NFLX), Google (NASDAQ:GOOG) (NASDAQ:GOOGL), Facebook (NASDAQ:FB), Amazon (NASDAQ:AMZN), and Twitter (NYSE:TWTR). It seems possible, if not likely, that, as customers like Spotify (NYSE:SPOT), Hulu (NYSE:DIS) and Pinterest continue to grow, they will launch their own CDNs to improve control over user experience and reduce costs, which could have a large impact on Fastly’s revenue.

I don’t disagree that this is a risk, but it seems paranoid to worry about it. Durant discloses that he’s long NET (Cloudflare), and their go-to-market is different from Fastly’s. I’m not familiar enough with them to know if both can succeed easily, or if one is better than the other, but if Durant prefers NET that may be why he’s relatively less bullish on Fastly.

Agree to disagree. Fastly’s path to success seems really clear to me, and I don’t see any problem with their strategic placement or go-to-market. Durant’s concerns seem like the concerns we often hear on the board from techies that analyze the competitive threats to death, to their detriment. I’m sure I’m a little biased, but that really is how it seems to me. And we do see it all the time. Perhaps some folks with an ear to the ground can weigh in here.

Bear

8 Likes
4

Hi Bear,

Great write up… thanks for sharing…

I would agree that its gross margins are likely to go up with scale… thats the nature of this business.

However, there are two concerns I have with FSLY - and they ultimately manifest into what valuation I am willing to pay, rather than anything fundamentally wrong with the company.

  1. At the end of the day, FSLY is about renting compute / storage capacity, sure they have advantage in “serverless computing” for now… however, it can change and with relatively low switching cost, nature of business is it tends to commoditization. I worry as an investor, I would not know when the landscape changed and be left holding the bag!

  2. Because its compute / storage capacity driven business, it is required to spend huge money on capex… profitability measure of this business may be tough based on gross margins or EDITDA type measures… unless you are really keeping a tab on progress of gross margins vs depreciation an capex and have good judgement on how it progresses, there is room to get carried away by temporary hold back on capex spend.

I hope this makes sense and also looking for insight from more knowledgeable observers on this company.

10 Likes
5

One other concern is that they got a huge growth pulled in by sudden change due to WFM… however, even if that business is recurring, that growth rate may not sustain in a quarter or two from now.

2 Likes
6

Bear - I’m mostly a lurker since I feel that I may not add much value to what is being discussed and still feel in learning mode. That said, I’ve been invested in Fastly for some time and recently listened to their call. It was clearly very upbeat.

Let me also suggest that people read Peter Ofringa’s take on their quarter including a high level comparison against Cloudfare. I particularly like his tech POV since he is a CTO and former VP of Engineering for several companies https://softwarestackinvesting.com/fastly-fsly-q1-2020-earni…

Ganador

12 Likes
7

One thing to keep in mind is that the same enterprise can be a customer of multiple of these companies.

I know of one Enterprise which was originally on AWS CloudFront. CloudFront had an outage which made it this company’s website crash for over an hour.

Since that time they signed up for Fastly and Cloudflare. They route 90% of the traffic to CloudFlare, 5% to Fastly, and 5% to CloudFront.

If Cloudflare were to go down they would direct the traffic to Fastly or CloudFlare instead.

I wouldn’t be concerned much about companies rolling their own CDNs. That’s hugely expensive and not a good use of time. It may only make sense for FAANG which already has distributed data centers around the world.

7 Likes
8

this is a great analysis on FSLY:
https://softwarestackinvesting.com/fastly-fsly-q1-2020-earni…

2 Likes
9

I am delighted to have read the write up from Bert on Fastly. But since then the run-up in the stock price has been extraordinary even by recent SaaS standards. I am most keen to double my position but I think I will wait for a better entry point. I know that is some degree of market timing but there may be some effect from the recent secondary / dilution

2 Likes
10

I’m one of those nerds some popular contributors to this board like to, shall we say, tease? I think it’s important to under both what a company does and how they do it. During NTNX’s heyday on this board there was much misunderstanding of what Nutanix’s products actually did or didn’t do. There were a few of us nerds who were able to point out that Nutanix wasn’t actually yet making a product that would let you seamlessly move your application between your private cloud and any of the major public clouds. The company’s marketing hype was too great and some investor expectations were too high.

Anyway, if we’re talking Fastly, we need to talk CDNs - Content Delivery Networks The basic way a CDN works is that they set up lots of servers all over the world that contain content (say, articles in the NY Times). The idea is that if you’re a NY Times subscriber in Los Angeles reading the NY Times the content you get comes from a server in SoCal. And as a result, you get that content more quickly.

Akamai is the grand-daddy of CDNs. It’s got over 288,000 servers (called “POP” for Point of Presence) around the world (https://www.akamai.com/us/en/about/facts-figures.jsp ). (https://www.fastly.com/network-map ). Akamai claims that 85% of the world’s Internet users are within a single network hop of an Akamai POP. Fastly has just 59 POPs.

According to Fastly, that’s by intention: we’ve focused our efforts on placing fewer, more powerful POPs at strategic markets around the world. With Tier 1 transit, solid-state drive (SSD) powered servers, and an engineering team that lives to optimize for speed, we’ve built a blazing-fast network that requires less hardware to deliver comprehensive global reach.

Fastly gives an analogy of legacy CDN vendors having lots of convenience stores, while Fastly has a few Supermarkets. Having lots of stores made sense 15-20 years ago, before mobile and when lots of people were still on dial-up. But today getting to one of Fastly’s supermarkets takes only a few additional milliseconds. But, the advantage of having larger supermarkets is the hit rate is so much higher: smaller POPs can’t hold as much content, forcing many more requests to go back to origin and resulting in hundreds of milliseconds of additional latency (which translates to slow load times for end users and additional hardware costs for customers). So, while super popular-content may be a few milliseconds slower, most content is hundreds of milliseconds faster.

One problem with a large supermarket is that it can take longer to find what you want versus in a smaller convenience store. Fastly addressed that by using SSDs instead of spinning disks.

Fastly’s different approach means that the old comparison metrics (eg, how many POPs and how close to you are they) don’t work. Here’s such an article: https://www.codeinwp.com/blog/maxcdn-vs-cloudflare-vs-cloudf… , which is useful for other things, like features and pricing.

Competitors in that article are:
• Akamai Edge (800 pound gorilla with over half the existing market, also offers performance tools)
• Amazon CloudFront (second place in market; but works only if you’re using AWS of course)
• StackPath (formerly known as MaxCDN)
• CloudFlare (not really a CDN; it’s a reverse proxy that takes over all a company’s traffic and then serves cached versions where possible)

Of these, Fastly and Cloudflare each have the smallest market share.

Fastly is known as a “real-time CDN,” meaning that you can update content on it really quickly (they claim within 150ms).

In terms of performance, I found this comparison on the web: https://www.cdnperf.com/ Here is the state when I looked (it changes by the minute):

  1. Google Cloud CDN
  2. Akamai
  3. jsDelivr CDN
  4. AWS CloudFront
  5. Azure CDN
  6. Verizon (Edgecast) CDN
  7. CDNetworks
  8. CloudFlare
  9. Fastly
    …
  10. StackPath

(Ranked by median HTTP request time)

In terms of business model, Akamai seems only interested in big players. You can’t do anything without talking to a sales rep. Fastly has the most transparent pricing.

The nerd in me is drawn to Fastly’s new approach, based on the current internet architecture, but while their points about fewer larger POPs make sense, it also strikes me then that if distance doesn’t matter, it would seem that any major site should just set itself up with fast servers and SSDs and cut out the Fastly middleman.

Of course, I’m probably missing something. Fastly is very developer focused (eg, they have company blogs written by engineers discussing which new language they should support), but that may not be the best for business. Perhaps the new CEO will help change that focus.

I bought a tiny position to force me to track the company more.

75 Likes
11

I started a position in Jan/Feb and added in March and it’s grown to about 8% after the last report

I’m not technically inclined so will get out of my league if I try to explain really what they do. The softwarestackinvesting posts from Jan and Feb do an excellent job explaining the business. I can’t imagine a better explanation.

Here are a couple of thoughts, although I would welcome any feedback.

I initially took the position based on the combination of their high NER (136%), valuation (6x forward sales on 30+% growth), and small size (2 bil market cap in Jan). In March, once it seemed like they would benefit from COVID that helped me feel more confident as the stock tanked to $11 from $22.

I looked at Cloudflare too and tried to compare them and what I came away with is that Fastly is more specialized. Compared to NET, FSLY has better operating and net margins DESPITE having a much lower gross margin. Even though they had slower revenue growth that seemed to say they were getting leverage.

The CDN business is a commoditized business so it takes more than just offering a CDN service. Fastly’s future is in their “Compute@Edge” network. It’s only just getting started and it’s higher margin according to the CEO. NET is also specializing of course too, but I didn’t quite like that they were going up agains ZS and ETSC.

What does “compute at edge” mean? I had to ask myself. First off, they aren’t actually “computing” in the sense that they don’t make chips. They offer a network that connects devices (IoT, wireless cameras, etc). They offer other CDN services but that is their speciality and future (hopefully).

They for sure have short-term tailwinds from WFH/Covid but the CEO on the last call seemed to think it was an acceleration into the future more than a one-time event. In other words, once companies get hooked on Fastly’s service they won’t go back. Here was a comment, but he had a few that said the same thing

We are very confident in our underlying business model. The structural changes we anticipate coming out of this pandemic help our business in the short and long term. We remain confident in the long term efficiency of our business model and we anticipate we will continue to see leverage as we grow.

After the last ER I didn’t understand the pop but I think it can be attributed to it being a small company, was undervalued, and of course the accelerating revenue guidance. But because they have such high recurring revenue and a high NER it means they have a good chance of retaining that revenue.

One other note, after the last ER the stock sold off despite accelerating revenue PROBABLY because it was announced the CEO/Founder was stepping down to CTO. But can you imagine a better time for him to make that move … right before the pandemic? Now you have the founder/leader overseeing the technology and a CEO who knows how to run a conference call

Last thing, I imagine the stock was down today (although now it looks like it’s up) because they priced the equity offering just under where the stock was at the open.

https://seekingalpha.com/news/3576628-fastly-prices-equity-o…

Regardless, this equity offering seems like a good balance of dilution that will help provide funds for the surge in their business.

Perhaps their customer growth is a concern, but given their high NER, high recurring revenue, improving margins, specialization, relatively small size, still trading at 7-8x forward sales (based on 30+% growth), and obvious near-term tailwinds I’m willing to watch it play out.

6 Likes
12

Great writeup. So, I’m skeptical about the TAM and growth potential out of partial ignorance

  • because in both AWS and Azure, deploying an edge configuration locally “closer to users” is a few mouse clicks and an edge configuration naming. AWS’ Cloudfront “deliver(s) content to end users with lower latency, Amazon CloudFront uses a global network of 216 Points of Presence (205 Edge Locations and 11 Regional Edge Caches) in 84 cities across 42 countries”.
  • multi-layered security, firewalls, configurability

Azure provides similar edge computing resources.

I haven’t researched Fastly at all; just pointing this out as an item to investigate with regard to competitive differentiation vs the behemoth and the mammoth already in the space.

13

Thanks Bear,

great write up as usual. I too, have bought into Fastly and a smaller position in Cloudflare. Elsewhere on this thread there are references made to both Bert’s and softwarestackinvesting write ups. I would urge anyone who wishes to understand why this company might be worth investing in to read them. There are also Tim Beyer’s pieces on MF for those who subscribe to various services.

I sold Akamai after having held it for many years despite to buy thee two companies.

The reasoning was as follows :

CDN became a commoditised many moons ago and whilst Akamai made serious money most of it came from their move into associated security products and not from CDN itself.

Since the delivery of content is now moving to the edge of networks( I know this is a fact rather than the technical nuts and bolts of how this works) then companies that sell a solution which move the content there are more relevant to today’s environment. (sorry to state the bloody obvious).

So hence Fastly and Cloudflare. I also like to laws of small and large numbers. As an aside , I sold some of my datadog earlier this week @70 due to it being a 22bn company. Can it go on to become a 100bn company ? Absolutely. Is it harder for them to go 100bn than fastly to go to 15bn ? yep.

This, of course, is dependent on TAM etc but generally it is harder given the laws of small and large numbers.

Why Fastly more than Cloudflare ? Just personal preference perhaps. Cloudflare essentially uses a freemium model and the type of land and expand strategy which is familiar to many on this board through the companies talked about. Fastly , a more top down approach where they essentially go after more established businesses who need a more specialist service.

The move of Artur Bergman to CTO makes sense. He is akin to Wozniak at Apple. Just brilliant but not particularly interested in running the company. The new CEO gave me kittens when I listened to the first CC he did. He sounded young, eager and naive. I thought how old are you ? 8.

I then thought oh dear what have I done putting so much money into this company . However, he sounded much better on the CC of last week and I am willing to cut him some slack though I don’t like that he has fired the head of sales.

The “Compute@Edge” network is what they have bet on to make their future. I will leave it to others more knowledgeable than I to assess whether this makes sense. It seems to but …

12 Likes
14

Annavalluc, I think that sales performance was definitely lackluster recently - they just added like 8 or 9 enterprise customers in recent quarter (OK, Covid, but still). So, they’re firing head of sales and CEO will be supervising sales for the next months. This whole thing is obviously not positive.

Another thing to watch - Fastly’s core business - content delivery on the edge - how big cloud players will be reacting to this recent “edge” trend - AWS, GCP, Azure - all have their edge products to my knowledge. So, we have traditional David vs Goliath topic. NET is less a competitor due to the different focus - although they want to move to enterprise, but to my understanding - inherently their platform is more focused on small and mid-sized companies.

Long FSLY, but will be closely watching if the thesis won’t be broken in next quarters.

1 Like
15

Yesterday: Fastly, Inc. (“Fastly”), provider of an edge cloud platform, today announced the pricing of its follow-on public offering of 6,000,000 shares of Class A common stock at a price to the public of $41.50 per share. In addition, Fastly has granted the underwriters a 30-day option to purchase up to an additional 900,000 shares of Class A common stock at the public offering price less underwriting discounts and commissions

I added my second third yesterday morning. It looks like that was an inopportune time to add. However, after reading the notes I’ve taken from this board and elsewhere, I’m hoping to add more soon. I really like the guidance for acceleration prior to the Compute@edge is made G.A.

16

Sorry… I meant that post to go out yesterday… so two days ago they priced their secondary.

J

17

Great post Bear, I’ve owned Fastly for a while now and believe the company has tremendous runway for growth ahead.

One thing, Joshua Bixby wasn’t a recent hire. He’s been with the company since… 2013 I think (off the top of my head) and I believe they were planning for him to take over as CEO for a while.

He’s a natural fit as their CEO and Bergman, the previous CEO and Founder will remain highly involved, but in a more technical role which is exactly what he wants/where he should be.

8 Likes
18

Just a follow-on to my post, since I responded to a similar thread on the paid FSLY board.

It’s actually ironic that Fastly touts itself as an “Edge Cloud Platform,” because Fastly is less at the edge than traditional CDNs. It’s kind of like Porsche putting “Turbo” in the name of their first all-electric vehicle. I guess people associate both “turbo” and “edge” with Fast, so that’s why marketing did those things.

What Fastly is trying to do with Edge Cloud Platform is to go beyond what CDNs do today. Fastly is trying to be more than just delivery of cached content with more speed and security (traditional CDN stuff). It’s about streaming video. It’s about load balancing. It’s about customization - you can even run Fastly’s CDN software on your own hardware if you’re big enough (like Spotify does). Heck, Fastly has even started a Beta program to compete with AWS, Azure, Google Cloud Platform and Oracle Cloud Platform in the realm of serverless computing (AWS’s Lambda was the first public serverless environment, btw). Here, Fastly is again touting speed, saying they can startup services 100X faster (measured in microseconds instead of milliseconds).

This is smart, at least technically. Fastly wrote a bunch of software and put serious high-performance hardware assets in place around the world for their CDN service. They are now leveraging that to provide new services at a low (to them) development cost. Put their software on your hardware if you want to customize where the POPs are located for performance or if you want the CDN servers inside your own network to use your existing security mechanisms. Put your own applications on Fastly’s super fast hardware, where it’ll outperform AWS handily. Use Fastly to deliver not just cached content, but to provide streaming services.

Of course, it still remains to be seen whether these new initiatives will be successful in terms of revenue and profit. But since the development cost is likely low since they leverage what Fastly has already done, the cost of failure should also be low.

24 Likes
19

Bear,

I just listened to the call. I won’t repeat what you have already written. Some things that made my ears perk up:

  • school closures have had the largest impact on the business. So, I take this to mean that Zoom and other such tools are using their live streaming service. I also think that schools will remain closed at least through January or such time as there is a viable vaccine, so this wind should be in their sails for some time.

  • While this might have had the largest impact on their business, this platform, the CEO said, is not mean for live streaming - it is not an event driven business. I took that to mean that streaming is just icing on the cake.

  • Regarding expenses, I heard him say early in the call that they have had to acquire extra capacity, thus increasing expenses.

The tone of the meeting was go-go. Hiring presents some challenges, which if not fixed obviously could impact growth. Growth impact is my assumption.

Gordon

6 Likes