Fastly CEO: One-on-One – CML

Fastly (FSLY) is our number two Spotlight Top Pick and we have just competed our one-on-one conversation with the CEO.

This places Fastly in the inner most circle of this logic diagram:

We most recently wrote about the earnings report for Q1 and the steep sell-off in the stock here:
Fastly Shows Largest Customer Growth Ever – Wall Street Doesn’t Care

Fastly’s stock has been hit hard off of the earnings report in a reaction that did have us surprised.

While Wall Street saw a “miss” to Q2 revenue guidance and a replacement of the CFO, we saw three phenomena that were entirely different.

First, we saw the largest new customer additions ever in a quarter immediately following the addition of its first ever chief revenue officer — an executive with experience running sales for enterprises with multi-billion dollars in sales.

Second, we saw a full year revenue guidance raise, which far surpasses any sort of intra-year analyst estimates in importance.

Third, we saw the company following its pattern of hiring executives from multi billion dollar enterprises like PayPal and VMware. We are pleased to see the CEO act in the best interest of shareholders, understanding that professional relationships are eclipsed in importance by commercial success.

We pressed on these topics and then broadened to exciting new opportunities from talks with Microsoft, a continued cementing of the defacto standard of edge computing, and an imminently monetizable new product that was announced in late April.

Please enjoy our summary, and then the actual transcription.

CEO Joshua Bixby was well aware of the stock drop and the questions surrounding credibility for the full year raise.

He addressed the credibility issue head one when he said

… I understand there’s some skepticism. I didn’t think there’d be this much skepticism, but I understand there’s skepticism.

With regard to Wall Street’s reaction to what was a miss for Q2 revenue guidance yet a raise for full year guidance and the seeming disconnect he first spoke to the Q2 estimates from analysts:

… [W]e don’t guide Q2 until the end of Q1 and I think [analyst] numbers… [where] a little bit off.If we could have sketched out what we thought Q2 was at the beginning of the year, it would have been in line with how we had guided this quarter.

Nothing’s changed from our perspective.

And then he commented on the company’s raise of full year guidance:

What we see for our vantage points four months into the year is that existing customers are way more predictable this year than they’ve ever been…[W]e get to see the pipeline and all the stages of the pipeline, including those that are closed.

… [T]he customers that really matter make buying decisions in Q1 and Q2 and implement in Q3, just shut down everything in Q4 in order to get themselves ready for an end of year push.

So, we’re, what? Four months and a week into the six months that we really reserved for people to make decisions that we know will impact revenue, right?

So, I think people are forgetting that the window to have an impact on the full year is… We’re well through it. And I think people forget that about our business

He points to the fact that the actual material selling season for revenue follows a Q3 to Q3 arc, not a calendar year arc, and in that vein they have seen the substantial portion of the selling season making full year revenue estimates far more visible and far more accurate.

As for Fastly’s ambitions and questions as to whether it will materialize into a large company he notes

…[E]ither we’re a house cat or a lion……[W]e know who we are, we’re a lion.

When addressing the announcement that Fastly would be going in another direction from its CFO of five-years, Joshua points out that this is a repeating pattern that Fastly is engaging in, to scale the business with executives that have worked with multi billion dollar sales.

The company recently hired the former chief people person of PayPal, an executive from VMware as chief revenue officer, and a CTO with scale experience.

He is looking to do the same with the CFO position — “you can rest assured that one of the characteristics that we are looking for .. is having seen around the corner.”

We then discussed Microsoft’s recent joining to the Bytecode Alliance and the impact on the industry he noted:

…[W]hen Microsoft joined and you see Google and you see Intel, it really starts to signal that WebAssembly is the defacto standard.This is the VHS of the VHS versus beta story.

I then asked if Fastly was engaged with Microsoft on this topic to which answered “we are absolutely in talks.”

Finally we discussed Fastly’s new Nearline Cache product and whether this was an immediate opportunity to revenue and growth and he said:

I think this is one area where absolutely, we’re going to see immediate impact in the existing customer base and in new customers.

I do note that for some questions I took a rather deeper dive into technology with Joshua than I normally do. I hope it was instructive.

Please enjoy our conversation.

One-on-One with Fastly CEO
Ophir Gottlieb:
Thanks for speaking as always. Before I start, how are you what are you hearing about the earnings report and Wall Street’s reaction?

Joshua Bixby (CEO Fastly):
A little bit of a gut punch over the last day or so as we talk to our investors.
We’re hearing exactly what you would expect, which is people see the long-term vision. They’ve got some questions about Q2.

Thank you for that. So, Joshua, I have a lot of questions. The first three are about the quarter and then I don’t want to talk about it after we talk about it, I want to go into bigger things.


Okay. So, as I see it, Fastly raised full year revenue guidance from… I’m talking to midpoints now, from 380 to 385 [million].


Left EPS guidance unchanged full year, reported the largest number of new customers in its history, even excluding SigSci [Signal Sciences acquisition], as far as I can see in public disclosures.

I know the company hasn’t been public that long. So, before I get to my questions, is everything I just said in fact an accurate representation of 2021? Did I misstate something?

I’d need someone on the call to double-check the customer number.

Maria Lukens (Fastly Head of IR):
Yeah, Ophir, let me double-check that and get back.

I mean, it’s as strong as we’ve seen. I just don’t want to say it’s the number one.

I think it’s in the top few, I just don’t know… Five here, 10 there, I don’t have my sheet in front of me where I can check that, but I would say it was as strong as we have seen in our history.

And we can double-check whether it’s actually number one, but it’s in the top, for sure.

I’ve done the checking, but yeah. No problem. Okay. So, now the real questions.

So, the stock is obviously down significantly on the report and since what I just said was true, it leads me to believe one of two things is happening.

Let’s just start with the first one, which is the easiest one.

Analysts don’t think that Fastly will reach that new raised full year guidance. And in fact, actually the Q2 number, which was a little bit below consensus, is more representative of what they see happening for 2021. Right?

So, in that regard, can you discuss as much as permissible how the 2021 full year guidance raise played out?

So, what happened either broadly or specifically such that you felt comfortable adding to full year expectations?

Yeah, so, I think first, as you know, we don’t guide Q2 until the end of Q1 and I think everyone’s numbers… we’re looking at some linearity that hasn’t been represented in the past history of the business and that I think was a little bit off.

So, I would say, if we could have sketched out what we thought Q2 was at the beginning of the year, it would have been in line with how we had guided this quarter.

So, nothing’s changed from our perspective, but I understand there’s arbitrary lines drawing a graph.

So, from our perspective, it’s not like Q2 was $5 million short. We knew what Q2 was going to be, give or take. We still have confidence in where it’s going to land.

What we see for our vantage points four months into the year is that existing customers are way more predictable this year than they’ve ever been; than they certainly were last year, right?

So, last year they’d be like, “Hey, we’ve got a Prime-like selling day. Oops, it’s off. Oops, on. Hey, I’m going to tell it to you two days before.”

And in the old world, it was like, “Here’s our schedule for the year, folks, and we’re going to stick to it.”

So, we are seeing much more of the latter than the former, which gives us way more visibility and predictability for the year.

I know what events are happening. I know when they’re happening. I know when e-commerce sites are going to ramp, when they’re not going to ramp. So, that gives us more confidence in terms of what we see.

The other piece that’s happening is we get to see the pipeline and all the stages of the pipeline, including those that are closed.

And what you and I know, and I think the rest of the world doesn’t always know, is that the customers that really matter make buying decisions in Q1 and Q2 and implement in Q3, just shut down everything in Q4 in order to get themselves ready for an end of year push.

And that’s not just… That’s e-commerce, even high tech thinks this way a little bit.

So, we’re, what? Four months and a week into the six months that we really reserved for people to make decisions that we know will impact revenue, right?

Because somebody signing in Q4, they just don’t have any big revenue impact given the business work.

So, I think people are forgetting that the window to have an impact on the full year is… We’re well through it. And I think people forget that about our business. So, I would say that.

So, I know what’s going to land. I know who’s going to land. And I think that $5 million raise is a combination of all of those things, which is we’re sitting there and saying… We’re coming out of COVID, we’re making the assumption the world goes completely back to normal.

And when we look at our pipeline, which is very much influenced by the security deals, we’re feeling really optimistic about it.

I think people are struggling, because there’s nothing I can do to show you numbers about that optimism. They have to rely on us for it.

And I understand there’s some skepticism. I didn’t think there’d be this much skepticism, but I understand there’s skepticism.

Yeah. A really good answer, a lot of color. I appreciate that. Okay.

So, one thing is there’s incredulity surrounding a raise. Okay. That’s one.

The other possible phenomenon to impact the stock, given that the report, in fact, it could actually be qualitative rather than quantitative, and that would be surround Adriel [CFO] exiting.

Can you talk about what Fastly is looking for in its new CFO?

So, I know when Brett was brought on… I think it was one of the first times, maybe it was the first time.

Fastly brought in a C-suite exec that actually had experience guiding an enterprise with multi-billion dollars in revenue. Right?

Is this new CFO spot reserved for someone with a similar experience, this scale experience?

Yes, sir. I mean, there’s no question.

I think if you look at… I was telling this sort of metaphor inside the business is everyone knows we’re a small baby cat, right?

And so, either we’re a house cat or a lion, and you can’t always tell.

There are little baby house cats that when they’re babies maybe look lion-like, right? And so I think we know who we are, we’re a lion, and we know that we know the kinds of people and the kinds of experience that it takes to tame the lion and grow the lion and give it… So, there’s no question this is about this hire, like Brett.

And I would point to Doniel {Sutton] as well, right?

Because Doniel was chief people officer at PayPal.

I mean, a thousand people is chump change in terms of numbers for how she’s dealt with.

Same with Brett.

And I’d argue the same with Nick. I mean, if you look at the last three hires, Nick Rockwell [Executive vice president and chief technology officer], he was the CTO of the New York Times, and before that ran very, very large businesses.

So, yes, I think you can rest assured that one of the characteristics that we are looking for, which Adriel doesn’t have, but he has many others, is having seen around the corner.

Every day, this is the largest business that I run and I’m not shy to say that, and I’m not embarrassed by that.

And so, I need to surround myself with people where that’s not the case in order for me to be successful.

well said. Okay. I don’t have any more questions about the quarter reported, and I want to talk about broader based themes, but before I move on, since we’re here, is there anything I should have asked, but didn’t ask?

Or maybe a smarter way for me to say it, is there anything else you want to discuss about the quarter with respect to revenue, customer additions or any other topic before I move on?

Nope, I’m good. I mean, I think you saw the SigSci enterprise customer count numbers, which was eight, which was incredible.

I think that you heard the 15% quarter over quarter SigSci growth, which is a testament to the integrated sales effort that we have.

And I think as much as when I have the stock price sitting behind me on broadcast news, it doesn’t feel this way. We’re optimistic.

And we think the quarter was optimistic, and I don’t know how else I can communicate that, but no, let’s keep going.

Yeah. I’m with you. If I would have not been looking at a screen and I just read the press release like, “Oh wow, raised full year” and I did the math, most adds ever, like, oh, this is going to be huge. And I was like nope.

A billion dollars in the bank, growing at 30-35% plus, gross margins improving year over year, more customer adds.

I’m kind of in the same boat.

No, sometimes you can’t always take the label of the player that you’re looking at off when you look at the field. Right?

And I think, listen, we’ve had some volatility. We had a miss. I think all of that is impacting us, at least in the short term.

Yeah. That’s a good perspective you have.

Okay. I don’t want to talk about the quarter, I want to talk about other stuff.

On April 29th, Fastly announced essentially Nearline Cache, right?

So, the storage solution for customers with these gigantic content libraries, and then automatically populates the content in cloud storage, so it doesn’t incur the egress costs near the Fastly POP.


So, if from a very sophomoric point of view, because I’m not great with infrastructure, this feels like a virtual holy grail.

So, better, or if you prefer, faster and cheaper together, right?

That’s the mantra of business, “make it better and cheaper.”

So, can you discuss that and how we should be looking at its impact on 2021 and then beyond 2021?

And does this essentially give Brett [CRO] a new product to sell with immediately addressable opportunities?

Yes, it gives Brett… I mean, I think it does two things.

One is for our existing customers; it provides them with immediate cost savings and immediate performance gains.

So that allows us in… I guess stepping back.

In many of the customers we’re in, there’s still a lot of market share, wallet share to get. And one of the ways that you get that is by bringing innovation to the table and saving the money. So I think this is one area where absolutely, we’re going to see immediate impact in the existing customer base and in new customers.

So that’s number one.

Two, I think… The story of the media business is one that, as we’ve talked about, this is a business that has to grow in line with the rest of our business, because it provides unquestionably the volume and, in some cases, the margin.

But in some cases, the margin is sort of margin neutral, not accretive to the business. So, we have to be really careful about how we grow that business.

This helps, but I do not want to have a business which is growing where the media side of the business overtakes the other side of the business. That’s not interesting to me.

So, I think the answer is yes, it does, but where we really need to see even more growth in our business is on the enterprise side.

And that’s really what Brett’s been brought in to continue to optimize. And SigSci obviously is a huge part of that.

Okay. Yeah.

So, interestingly, on that same day, April 29th, Microsoft announced that it had joined Bytecode, the Bytecode Alliance.

So, Web Assembly, at least in my opinion, is one of the core underpinnings of getting the value out of Compute, right? Compute@edge.

Yes, sir.

Can you talk about the Bytecode Alliance and the significance of Microsoft joining?

I don’t know if you can answer this, but I’ll just ask, has Microsoft been in conversations with Fastly about Web Assembly?

What’s your view of that announcement broadly? What should investors take away from that? Because I think we also saw Google is a part of it, too.

Yeah. I mean, I think there are a few messages. I think the first is that there are competing views on what this is going to look… what sort of dominant force is going to own the edge compute worldview, right?

And I think when Microsoft joined and you see Google and you see Intel, it really starts to signal that Web Assembly is the defacto standard.

This is the VHS of the VHS versus beta story.

When you get Microsoft… You think about Microsoft and what we all forget is how fundamental they are to the tooling environments of the world.

The development tooling environment revolves around the axis of Microsoft. And had they not gone out and got GitHub, it’s possible that it would have been a different evolution, but that business is amazing and it’s growing.

So I guess I would say the one strong message is what this says is Web Assembly is the defacto step.
That’s one.

From a Fastly perspective, what it says is that, to your point, we are absolutely in talks.

And what’s really important for us is to continue to expand the language base. So, .NET or C++ or C… These are all native, C and others, to Web Assembly. And that’s awesome.

I was having a conversation with CTO.

He’s like, “Yeah, I don’t really want to use Rust. I don’t really want to use Ruby or whatever. Hey, can you just do C++?”

We’re like, yeah, it works. And that was like, “Awesome. Now we can use it.”

So, I think it’s really significant. I think we’ll see that significance as the Microsoft tooling environment starts to bring Web Assembly in.

When Web Assembly is really injected into that tooling environment, then deployment, which is actually one of the most complicated aspects of this, can be solved in your developer environment. Right?

So, “Where do you want this to run?” is not a question we should be asking developers, but we have to ask them today. In the future, the programs themselves should make those decisions.

Yeah. If .NET… I mean, this is just sort of off-topic. If .NET allows for easy launching onto the edge, or… It’s just like crazy, crazy change.

I mean, I think it impacts Fastly the most, but it impacts a lot of companies.

Well, I think everyone. I think, yeah, exactly. Not just the Fastly story.

If you actually look at GitHub and our work there, they’ve launched a whole suite of new products that integrate with us. So, I don’t know, I’m very bullish.

The war of edge compute is going to be won at the developer library level, in part.

And so, the fact that Microsoft’s on board and has really sort of rubber stamped and put their name behind this is deeply important.

Yeah. I so agree. I saw that news, I thought it was market moving. Not that anybody cares what I think, but… Okay.

I want to go deeper on Web Assembly, then.

I’m going to say some things and I think they’re right, but if they’re not, then you can correct me.

Web Assembly does not support GPU access, so it won’t be able to leverage the computational power of Nvidia products.

So, I bring this up because Cloudflare made an announcement about partnering with Nvidia, and I’ll be honest, I don’t know a lot about it, so I don’t want to overstate or understate it.

Basically, I saw a presser and I moved on.
But WebGPU builds on top of the lower-level graphics API movements, right?

So, it enables fuller feature GPU accelerated compute capabilities compared to the predecessor, WebGL.

And Web Assembly is a similar effort for unlocking CPU based compute capabilities.

So, does Fastly plan to incorporate WebGPU at some point?

Since I think the open-source community has already made quite a bit of progress in terms of compiling machine learning languages with Web Assembly and WebGPU. For example, like Apache TVM.

So, one, I’m no expert on this. Let me tell you what I know. I think the first is I also I don’t know what the Cloudflare announcement meant, but it certainly gave them a nice short-term boost.

I think that a couple of things are necessary to have this work.

One thing that’s actually been in traditional CPU usage for years, because it’s been driven by the large clouds, is multi-tenant architectures.

And more importantly, multitenant security.

So, if you’re using part of a CPU, the advent of the shared global compute environments forced everyone to spend time thinking about the security paradigm.

GPUs are way farther behind on that trajectory, because not a lot of people are sharing GPUs, and they’re certainly not shared in public environments in a meaningful way at scale.

So, I think the way we’re looking at it is we are working with the vendors, both through the Alliance and just one-on-one to really try to understand what the timeline is for multi-tenanted GPU security, so that we can then build sort of a model going backwards of how that could be used in an environment like ours.

So, I think there’s a big question mark that has to be addressed.

It’s not even a Web Assembly question.

It’s like, have the developers of these GPUs thought about the fact that they’re going to be used by 500 different customers all at the same time? All of whom want to attack each other, on the same box.

So, that’s the step in my understanding.

Again, I’m happy to get the people who actually are much deeper in this question to answer, but that’s the biggest barrier we see right now which impacts the timeline.

So, if I were to give a timeline, I think we’re two to three years from seeing multi-tenanted GPU at the edge, but I believe we’ll see it.

And I think when we see it, it’s going to be really interesting. We’ll see experiments well before that of actually production running the most secure workloads.

Okay. Interesting. That’s a good perspective.

So, now I want to talk about something Fastly can do, but in a different industry, because these are some of the companies that I cover.

So, as incredible as the technological landscape has expanded, just in general, technology, internet infrastructure…

Still today, from what I can see from my perch, the most prolific monetization in technology still sits in advertising.

No matter how cool of a technology you and I could talk about, or Artur [Artur Bergman, founder and Chief Architect] is conceiving, it’s Fastly’s customers’ Twitter and Pinterest, et cetera.

They’re making all the money.

So, in 2019, Fastly wrote about edge side ad insertion, that it was now possible.


Is there any update to the capability and usage?

I ask because it’s sort of a bit of a hidden opportunity if advertisers start to realize that there are core underpinnings to engagement with ads, right?

And one of those underpinnings is the speed of delivery.

So, if you tell people who do ads for business, “Hey, your ads are going to be faster and they’re going to get delivered right,” that’s just like a $300 billion market. So, I’m just curious…

Right now, the answer’s maybe, and let me tell you why it’s maybe.

And to the question of progress whatever, yes.

Where we keep seeing the challenge is in the auction nature of the ad business.

So, if I allocate a budget of 50 milliseconds to go out and query the universe, because so much of this is done in real time, right?

Then your ad only has to get there within that budget to be valuable. And so we’re sort of right now… I think where we will see this become way more valuable is when we start seeing the ad latency budgets for auctions come down, and we’re already seeing energy in that direction.

But right now, if you actually just looked at why does that ad take time, it’s because of the auction process, not because of the delivery process.

The delivery process can absolutely be made better and every vendor wants to, but they all are at the mercy of the auction process.

So, I think the answer is the auction process needs to change, which we’re seeing a change, and only then will the multi millisecond advantage you’re going to get be seen at a broader scale.

Now, if you start talking about first party ads, that’s totally different because they don’t have an auction process. And in that case, the advantage is all about personalization.

So, where we have seen, where we continue to see the value, and where I think we’ll see the first value in this is actually more so on that first party personalization.

So, imagine you were watching a game and the ad understands something about your behavior and then targets it to you.

So, I think it’s going to be first party non-auction and possibly video that is actually the killer use case to start.

That’s a good point too. Yeah. And there’s just the idea of engagement.

I mean, engagement is important for everything, even if you’re just in a grocery store, like the owner of the grocery store wants you to stay in the cereal aisle as long as possible.

But engagement with ads, that’s their whole business.

And once we get there with the edge… Part of engagement is speed, right? Content is content.

Fine, they have to create good content.

But a part of engagement is how easy it is to work with as a consumer of that ad and that’s speed.


Ophir Gottlieb (22:13):
It’s just so big and the advertising world is just one market… They actually move quickly if there’s something that’s better.

You can see what they did with Facebook many years ago.

They realized, “Oh my God, the appification of the internet is coming and this thing called Facebook sells better.”

I mean, it was just en masse, right? From Facebook being not really a thing to a hundred billion dollars.

I mean, that’s one of the most unbelievable trajectories.

I mean, I think the other element here, which is an underpinning, is all of this work that Apple is doing to create a private internet is confounding the entire industry.

Everyone is trying to assess this.

And I think we are hearing just a lot of… It’s stalled innovation, because no one knows what they’re going to be able to target, who they’re going to be able to target, and in what form.

I think people are figuring this out, but I was talking to a few folks, a couple of public CEO folks, because I was trying to get their feel for this.

And they’re just like, “None of us know what the impact is going to be.”

So, I think that’s also having an impact in the short term. That’s going to be a six-month impact, but Apple’s not done.

I mean, Apple is going to continue to drive privacy deep into that experience, I believe. Certainly, from their public statements.

Yeah, for sure. Yeah.

We cover Pinterest and we cover Alphabet.

We cover a lot of companies and there’s a lot of confusion.

I think some of them are wondering maybe if this is kind of like, I don’t if you remember, in 2000, the 2000 bug. The idea that because dates had to roll over, that that would be a problem.

And Y2K was actually not a big deal.


All right. That’s all I have. Joshua, Maria, thank you for your time.

It’s always a pleasure. We appreciate the support. We love that you try to understand and actually deeply understand the story. So, if there’s anything we can do to help you, we’re here.

Awesome. All right. I will talk to you in three months.

We maintain our number two Spotlight Top Pick status on Fastly.

We said last quarter “if the stock dips, so be it, it likely will.” Well, we got the dip part.

We don’t know a lot about the short-term, but the long-term future, that is clearer to us — albeit quite different than Wall Street sees it.

And in our opinion, even as stocks go down, even as panic shakes the investing world, it doesn’t have to shake us.

We just have to accept that stocks go up and down, and it will be scary when the ‘down’ part comes. Look forward.

We see the future of technology, by simply listening to the companies and executives themselves and applying our own expertise.

And all of that research is bullish — maybe not for next week, or even next year, but in 3-10 years, yes, we see a huge shift higher in technology.

Thanks for reading, friends.

The author is long shares in Fastly at the time of this writing.

Please read the legal disclaimers below and as always, remember, we are not making a recommendation or soliciting a sale or purchase of any security ever. We are not licensed to do so, and we wouldn’t do it even if we were. We’re sharing my opinions, and provide you the power to be knowledgeable to make your own decisions.

Get a nice roundup of new retro gaming content once or twice a month.