Fastly (FSLY) Q1 2020 Earnings Results Review

Fastly released their Q1 2020 earnings report on May 6. The results were a blow-out. Q1 actuals were comfortably above estimates for revenue and profitability. Q2 projections and full year raise were even more impressive. At the midpoint, Q2 annualized revenue growth was increased over 20%. Much of the outperformance was provided by the COVID-19 situation, but management anticipates the tailwinds from higher usage across their product set to continue through the year. On top of this, the new Compute@Edge beta release is progressing nicely and should drive incremental revenue streams in 2021. The stock surged almost 46% the day after earnings, as the market digested the stellar results. All analysts raised their price targets. In this post, I will parse out the earnings results, give an update on the CEO transition, check in with comparisons to Cloudflare and then draw conclusions for the investment thesis going forward.

Headline Financial Results (EPS is Non-GAAP)

• Q1 2020 Revenue was $63.0M, up 38% year/year. This compares to consensus estimates for $59.4M, or 30% year/year growth. This represents a beat of $3.6M or 8% of annualized growth outperformance.
• Q1 EPS was ($0.06) vs. ($0.12) expected, representing a beat of $0.06. This compares to ($0.30) in Q1 of 2019.
• Q1 Non-GAAP operating loss was $6.0M, representing an operating margin of -9%. This compares to an operating loss of $7.0M in Q1 2019, representing an operating margin of -15%. Q4 operating margin was also -15%.
• Cash used in operations was $7M in Q1. Capital expenditures were $12M. FCF was ($19)M for Q1, representing a FCF margin of -30%. Q4 FCF margin was -15% and in Q1 2019, it was -32%.
• Q2 2020 Revenue estimate of $70-72M, representing growth of 53.6% over Q2 2019. This compares to consensus revenue estimate of $60.0M or 29.9% growth. Obviously, a substantial raise here of over 23% of annualized growth.
• Q2 EPS estimate of ($0.02) to ($0.00) vs. ($0.11) expected, representing a raise of $0.10 at the midpoint. Note the rapid increase towards break-even.
• Q2 estimated Non-GAAP operating loss of ($2.0)M to ($0.0)M for an operating margin of -2.8%.
• FY 2020 Revenue estimate of $280-290M, representing growth of 42.1% over FY 2019 at the midpoint. This compares to consensus revenue estimate of $259.6M or 29.5% growth and Fastly’s own guidance in Q4 of $255-265M. Updated 2020 revenue estimate represents a raise of $25M over prior guidance or annualized growth of about 12%.
• FY 2020 EPS estimate of ($0.15) to ($0.08) versus ($0.27) expected, representing a raise of $0.16 at the midpoint.
• FY 2020 estimated Non-GAAP operating loss of ($20.0)M to ($10.0)M for an operating margin of -5.3%.
• Ended the quarter with cash, cash equivalents and marketable securities of $187M.

Other Performance Indicators

• Q1 Non-GAAP gross margin of 57.6% up from 57.0% in Q1 2019.
• Breaking down Q1 Non-GAAP expenses by category, we see gradual year/year reductions in relative percentage of revenue:
  • R&D = 20% (versus 21% in Q1 2019)
  • S&M = 28% (32% in Q1 2019)
  • G&A = 18% (18% in Q1 2019)
• FCF is affected by capital expenditures. Fastly is particularly sensitive to FCF margin variances, as they have to invest in network infrastructure to drive expansion. For 2020, they expect to spend 13-14% of revenue on capital expenditures. Long-term, this should drop to 10% of revenue on an annual basis. For Q1, the CEO stated that they were proactive towards COVID-19 surge and pulled some of the CapEx spend into the quarter that was planned for Q2.

• DBNER was 133%, compared to 130% in Q1 2019 and 136% in Q4. The sequential drop is not significant, as Fastly’s business is seasonally skewed towards heavier growth in Q3-Q4. This DBNER rate compares very favorably to peer software companies.
• Fastly leadership introduced a new metric of Net Retention Rate this quarter. It is defined like DBNER, but includes the impact of customer churn. For Q1 2020, this value was 130%.
• At end of Q1, total customer count was 1,837, up from 1,743 in the previous quarter and 1,621 in Q1 2019. This represents annual growth of 13.3% and sequential growth of 5.4%.
• At end of Q1, enterprise customer count was 297, for an increase of 22% year/year. Enterprise customer count was 288 at end of Q4, representing a sequential gain of 3.1%. Fastly is usually most seasonally strong in Q3-Q4. Enterprise customers spend greater than $100k in the prior 12 months.
• Enterprise customers drove the majority of revenue growth and accounted for 88% of trailing 12-month total revenue. This compares to 85% a year ago.
• In Q1, average enterprise customer spend increased to $642k from $607k in the previous quarter. This represents sequential growth of 5.8%.
• Regarding exposure to industry sectors being particularly impacted by the COVID-19 situation (oil/gas, travel, hospitality), the CEO stated that it is a very small percentage of total revenue, in the “small single digits”. The CFO mentioned a few customers have asked for slight changes to payment terms, but that it doesn’t rise to meaningful magnitude.
• In the shareholder letter and on the earnings call, leadership talked about progress with the new Compute@Edge offering. It is in beta release now, and they hope to open it up to limited availability in 2H2020. Management provided several examples of real-world customer edge implementations for digital encryption, authorization/authentication, data loss/protection and machine learning models. Additionally, Artur Bergman (Founder, Chief Architect) talked about personalization as a major use case for several beta customers. This makes sense for two reasons. Personalization generally drives incremental revenue for any web content or service provider. Second, once authentication is performed at the edge, it is straightforward to throw in some personalization logic, using cached data about user preferences.
• Leadership provided some examples of customer wins in Q1:
  • Affirm – a fast growing FinTech company. Chose WAF solution to protect API endpoints and Shielding for origin offloading to speed up the payment processing flow.
  • Neiman Marcus – Chose Fastly solutions for edge delivery for all five of their online properties.
  • One of the world’s largest e-commerce brands (unnamed) – Expanded usage to include edge application delivery and instant purge capabilities to increase performance on several properties. Chose Fastly due to unique capabilities in edge compute. There is some evidence this customer is Amazon, using Fastly to serve content on IMDB and the main Amazon.com web site.
• Leadership also talked about several platform companies that build their SaaS and PaaS services on Fastly and expanded usage in Q1. These included Mux, Pantheon and GitHub. This usage by other software platforms is significant, as these companies will be the most technically demanding given that they are providing software-based services for their customers.
• Discussed increased usage from some customers who are considered technical innovators. These included Shopify, Pinterest and Khan Academy. The implication is that discerning buyers are selecting Fastly due to their superior technology and developer-friendly motion.
• Attributed some of the revenue outperformance to marketing efforts started last year to educate customers on the opportunity and uses cases for edge compute. This go-to-market motion is expected to continue.

Analyst Reactions

Following Fastly’s Q1 earnings, 6 sell-side analysts provided updated coverage ratings. Interestingly, in reviewing Q4 2019 results, I could find only one analyst that posted an update. This earnings report must have been significant enough to garner more active interest from sell-side analysts. This might drive new institutional buying.

Of the 6 analysts providing updated coverage ratings, all raised their price targets. All but one rated the stock at a Buy equivalent. The average price target for these updates is $32.66, representing a 3% reduction from the closing price of $33.58 on May 7th.

Date	Analyst	Rating	Price Target
5/8	Citigroup	Neutral	Raised from $25 to $31
5/7	Bank of America	Buy	Raised from $26 to $31
5/7	Robert Baird	Outperform	Raised from $28 to $35
5/7	Craig Hallum	Buy	Raised from $29 to $36
5/7	Oppenheimer	Outperform	Raised from $27 to $32
5/7	Piper Sandler	Overweight	Raised from $27 to $31

Craig Hallum set the highest price target after earnings. Analyst Jeff Van Rhee provided the following commentary.

Craig-Hallum analyst Jeff Van Rhee raised the firm’s price target on Fastly to $36 from $29 following a “wow” quarter and guide. The analyst keeps a Buy rating on the shares.

TheFly, May 7, 2020

CEO Transition

As investors know, as part of the Q4 earnings report in February, Fastly announced that founder and CEO Artur Bergman would be transitioning into the Chief Architect role. This would allow him to focus on his areas of strength in technology and customer development. In Artur’s words, he would “focus more of my attention on our customers and to work more with the product and technology part of our business to build out our edge cloud platform.”

Replacing him was Joshua Bixby. Joshua had been the President of Fastly since May 2017. He joined Fastly as the VP of Marketing in 2015, after being a part-time advisor since 2013. Between 2015 and 2017, Joshua held Senior VP roles over marketing, product and sales. Prior to joining Fastly, Joshua co-founded a web application acceleration company called Strangeloop Networks in 2006, which was eventually acquired by publicly traded Radware (RDWR) in 2013. From 2002 to 2006, Joshua was founder and CEO of IronPoint Technology, a content management software company. He has a degree in Management from University of Toronto.

When this transition was first announced, I had mixed feelings. The transition of Artur made sense, as he clearly prefers working with the technology and wasn’t leveraging his strengths in the CEO role. I also like software stack companies where a technical founder/CEO runs the company for a long period (DDOG, AMZN, TWLO, ESTC, OKTA). While Joshua has experience founding two prior companies in the content delivery space, these weren’t brand names in enterprise software. I felt like Joshua was unproven in a larger opportunity and might represent a risk, versus selecting a replacement CEO with a little more experience, similar to the CEO transition at MongoDB with Dev Ittycheria. However, there is also an argument for selecting a fresh face, who can bring a lot of vision and passion to the role and is eager to make their mark.

Thus far, Joshua has performed superbly. Given Fastly’s Q1 results, the transition is obviously going smoothly and Joshua is clearly setting a high bar. I have been happy with his communications, both on the earnings call and in the shareholder letter. My concerns are mitigated and I think the combination of Artur focusing on technology development and Joshua focusing on business growth will be strong.

Coinciding with the Q1 earnings report, Fastly also announced the transition of its EVP Sales, Wolfgang Maasberg, out of the company. Wolfgang will stay on through November. He joined Fastly in 2016, after having served in sales leadership roles at Adobe, Omniture and Coremetrics. Usually a departure in sales leadership would be disruptive, but Joshua will be directly overseeing this function and had previously maintained oversight to sales. On the earnings call, Joshua mentioned that he will be evaluating the executive leadership structure in this area going forward and will make appropriate strategic changes. This might result in a better executive alignment for the long term.

Comparisons to Cloudflare

I often see comparisons made by investors between FSLY and NET in various public forums. I have also been asked by readers if I prefer FSLY over NET. With my coverage of FSLY’s Q4 earnings results, I tried to provide some apples-to-apples financial metrics from both companies. With Q1 results, I will try to explain my personal investing preferences and then give an update on relative performance of the two names.

At a high level, based on my observations, Fastly’s target audience is mainly developers, certainly evidenced as we look forward to Compute@Edge.  They want to help engineering teams build and deliver faster and more efficient software applications.  As an example from the Q1 earnings call, the CEO stated “the core of what we have always brought is to allow developers to innovate, and Compute@Edge is the next step in maintaining our significant advantage in that market.”

On the other hand, Cloudflare’s target audience appears to be primarily CIOs/CISOs who want to make their corporate network and enterprise apps more secure.  For example, on the Q1 earnings call, Cloudflare leadership talked about displacing hardware vendors of on-prem firewalls and VPN. “There are a lot of Cisco boxes out there, that have to be replaced.” Additionally, significant sales emphasis is being put on their new Cloudflare for Teams offering, described as a “real standout for this time”. Teams consists of Cloudflare Access, which replaces traditional VPN solutions by wrapping access to enterprise apps with Cloudflare network tunnels, and Cloudflare Gateway, which replaces on-prem firewall solutions with Cloudflare’s global network to filter and secure outbound traffic requests from enterprise users to the public Internet.

There is currently some overlap between the two companies’ product suites (Fastly has security offerings and Cloudflare has developer offerings), but reading content on each web site and the earnings call transcripts, the difference in target market going forward seems evident to me. The distinction is subtle and doesn’t stand up to all cases, but I think it may help investors appreciate the future product direction for both companies. I believe they diverge.

This difference is important to me because I prefer investing in companies that have a developer focus.  I have deeper personal experience in that space, as I used to run software engineering teams as a CTO/VP Eng.  I have some understanding of security matters, but could not be a CISO.  That’s really the crux of it.  I like to invest in what I understand.  For example, I couldn’t explain to you in depth whether Cloudflare for Teams will compete well with ZScaler’s offerings (assuming that is a thing).  But, I can certainly parse out how well Fastly Compute@Edge will compete with similar offerings from Akamai or cloud vendors (or Cloudflare Workers).

That is the main reason that I feature Fastly on this blog and haven’t covered or personally invested in Cloudflare. I think Cloudflare is a fine company and will likely continue to perform well. It’s just not an investment area where I feel comfortable offering insight that readers can’t find elsewhere. I don’t want to waste readers’ time or worse, misrepresent a technology underpinning in evaluating Cloudflare.

In terms of comparisons of financial performance coming out of the Q1 earnings reports, a table is below. I am including annualized growth for forward looking revenue estimates, as there does seem to be some acceleration associated with FSLY’s Q1 earnings report.

Metric	FSLY	NET
2019 Revenue Growth	39%	49%
Q1 Revenue Growth	38%	48%
Q2 Revenue Growth (Est)	54%	40%
2020 Revenue Growth (Est)	42%	36%
Q1 Non-GAAP Gross Margin	58%	78%
Q1 Non-GAAP Operating Margin	-9%	-16%
Q1 FCF Margin	-30%	-34%
Q1 DBNER	133%	117%
Trailing EV/Revenue	17.0	26.4

With the forward revenue and profitability guidance for the rest of 2020, it appears to me that Fastly’s growth is accelerating.  If I look at updated 2020 revenue guidance, Fastly went from 30% year/year projected growth to 42% year/year.  Cloudflare went from 34% year/year projected growth to 36% year/year.  At this point, Fastly is projecting higher revenue growth for 2020 than Cloudflare.   Also, profitability measures for 2020 stepped up significantly for Fastly. Non-GAAP EPS estimate raised from ($0.40) estimate to ($0.11) projected for Fastly. Cloudflare kept EPS inline at ($0.20) for the full year.

I am not saying that Fastly is a better investment than Cloudflare, as I am sure investors have differing views. Fastly needs to work on gross margins, for one. I just wanted to provide my perspective on the two companies, explain my personal philosophy and update some of the metrics I used to compare the two in the past, as I have been asked about this by readers.

My Take-Aways

• DBNER of 133% continues to be best of breed, as compared to other software solution providers. Fastly’s focus on technically discerning enterprise companies is driving this. With enterprise customer count increasing by 22% annually and representing 88% of total revenue, we should continue to see large customer expansion driving meaningful revenue growth going forward.
• While gross margins are still relatively low at almost 58%, Fastly leadership expects this to improve as more high-margin products like edge compute and security features gain revenue traction. On the earnings call, the CEO referred to Compute@Edge as “the highest margin side of our business.” Also, they expect to drive better utilization of hardware resources through continued optimization through software improvements.
• The new Compute@Edge offering is still in beta, yet could become a large revenue driver going into 2021. For investors not familiar with the story, I dug into this offering and competitive comparisons in my original Fastly review. While it is deployed in a serverless mode, Fastly’s technology solution can start up in 35 microseconds, which they claim is 100x faster than competitive solutions. Since an ideal request/response server-side loop needs to be under 20 milliseconds, Fastly’s server start up time leaves plenty of room for other processing work. The big advantage here is that this serverless model can be applied to synchronous use cases. Most other serverless models are relegated to asynchronous processing, where a human isn’t waiting on a response. By applying a serverless model to synchronous processing, a whole new set of use cases is available to Fastly’s edge compute solution. Also, the serverless model requires less hardware support, because a server isn’t idling waiting for new requests. This drives efficiencies in Fastly’s capital allocation.
• This performance bias also comes through in their thoughtfulness about selecting a second language for Compute@Edge support. It is clear that the choices are being constrained to runtime performance, versus other cloud serverless providers that include the standard laundry list of popular interpreted frameworks to support for broad developer appeal.
• Fastly’s depth of integration with some customers is significant and impressive. This will be a key driver of the sustained growth story going forward and is important for investors to understand. As an example, subsequent to the earnings report, Khan Academy published a post on their Engineering blog describing how they were able to rapidly scale to 2.5x usage over a two week period in March. The author describes how Fastly, along with GCP services, provide the “backbone of their serverless and caching strategy.” Further, the author states that all client requests are passed through Fastly first, in order to improve performance and reduce traffic sent to App Engine (GCP compute offering). Besides caching static content with Fastly, they also extensively cache common queries, user preferences and session data. This is all used to speed up data retrieval performance, by eliminating the round trip back to centralized servers. This goes far beyond legacy CDN usage patterns. Imagine what will happen once Fastly adds code interpretation to the mix with the full roll-out of Compute@Edge. Increasingly large slices of data processing will relocate from centralized server tiers to Fastly’s edge compute platform. Given that their data is already being cached on Fastly’s edge, customers would logically use the same edge solution to process it. It wouldn’t be efficient to cache data with Fastly and then use a cloud vendor’s solution to compute it.

Risks and Items to Watch

• While Fastly benefited from surges in usage by some of their customers, like the example with Khan Academy, we have to consider what might happen after the COVID-19 situation clears. Are these surges in usage temporary, returning Fastly to more normalized growth afterwards? I think we will naturally see some of that. At the same time, this question was raised on the earnings call. The CEO feels that COVID-19 accelerated various digital transformation initiatives and upgrades to a more modern architectural stack, off of legacy CDN solutions. He feels that this represents a more lasting and fundamental shift in how customers view the role of edge computing and how to leverage robust data caching and processing capabilities that exist at the edge to speed up performance and reduce costs. Nonetheless, a slowdown in future activity after COVID-19 is a risk to the investment thesis.
• Growth in enterprise customer counts slowed in Q1, from 14 adds in Q4 to just 9 in Q1. This is worth watching, as 88% of Fastly’s revenue comes from enterprise customers, defined as those spending more than $100k a year. This could be seasonality, or a large concentration of customers just under the $100k threshold. We will need to watch this metric for Q2 and beyond, along with overall customer growth. However, if this metric were a concern for future revenue, I wouldn’t have expected such a large increase in Q2 and 2020 estimates. Perhaps DBNER will go through the roof.
• While Fastly seems to have a head start in edge computing, the cloud vendors are not remaining idle. They see the risks of encroachment on centralized server compute and are investing in extending their dominance to the edge as well. Fastly appears to have a technology edge here. More importantly, since Fastly is already caching customer data at the edge, processing it in the same place is a natural extension. Because discerning customers are choosing Fastly’s solution for data caching now, they have a proximity advantage for providing the compute to process it.
• The CEO leadership transition seems to be progressing well. However, there could be other disruptions as Joshua asserts his leadership over the executive team. Artur seems on-board with the new arrangement as well, but if he were to leave, it might result in some engineering talent leakage.

Investment Plan

I first covered Fastly in January when the stock was trading at $24 and then gave an update with their Q4 results in February. While my disposition towards the stock was favorable, I didn’t set a price target then because the company was within one year of its IPO. As a general rule, I like to allow four quarters of public performance to pass before making a long term investment recommendation. In the case of FSLY, they IPO’ed on May 17, 2019, closing at about $24 that day as well. At this point, I am ready to formally recommend FSLY for investor consideration.

The main driver of my investment thesis rests on the potential for Compute@Edge. Fastly is already experiencing accelerating growth in 2020 and Compute@Edge is still in beta. I think this has the potential to generate incremental opportunity going into 2021 and offset any potential downside post COVID-19. As I discussed in the Khan Academy example above, the addition of code interpretation at the edge, where customers are already caching common user session and preference data, will drive a significant shift of compute processing to edge servers. This processing will enable enhanced personalization capabilities, as well as potential for IoT and gaming use cases. As a leader in edge compute, Fastly should grab a large share of this shift to the edge. Because cached data is already being stored locally by Fastly’s edge network, it wouldn’t make sense to transport the data somewhere else to process. The market potential for edge compute is huge, as significant spend is already directed at centralized compute resources.

In terms of setting a price target, FSLY is currently trading at about $38. This is up 65% from the pre-earnings price of $23.05. While there was a substantial run-up after earnings, the stock has been range bound for the past year. FSLY is targeting $285M in revenue for 2020. I think they can deliver $300M and approach break-even on EPS, based on the substantial raise of $25M in revenue in the Q1 report. This would represent 50% annualized growth. Today’s EV/Rev ratio is about 16. With $300M in revenue, the end of year EV target would be $4.8B. Current EV is about $3.5B. This would imply growth of 37% or a share price of about $52 by end of 2020.

Pushing out four years to end of 2024, I think FSLY can maintain high annualized revenue growth of 40% in 2021, dropping to 38%, 35% and 32% in subsequent years. This high revenue can be justified by the large addressable market, the addition of Compute@Edge and Fastly’s relatively small annual revenue amounts currently (law of small numbers). This would put 2024 revenue at approximately $1.0B. I think we can assume positive operating margins and FCF margin at this point. For 30+% growth and positive operating margin, a reasonable EV/Rev ratio range is 12-16. At the midpoint, this puts EV at $14.5B in 2024, or 4.1x today’s value. Therefore, I think a 5 year price target of $155 is reasonable.

Personally, I initiated a small position in FSLY earlier this year and substantially increased my investment following earnings. I currently have a 20% allocation at a cost basis of $32. I am comfortable leaving it there, unless the investment thesis changes. I will continue to monitor FSLY’s progress and provide investors with updates as the story progresses.