As I work through analysis of recent earnings reports for the companies currently in my portfolio, I also like to check back on progress for companies that I had owned previously. In each case, I had a reason to own the stock, based on their product portfolio, growth potential and alignment with broader trends in software infrastructure. However, every thesis does not play out as expected and subsequent missteps in company execution can lead me to downsize or exit a position until the company appears poised for growth again.
Fastly (FSLY) is one such stock. I had owned FSLY during its COVID driven surge in 2020, enjoying price appreciation from the $20 range in May 2020 to break $100 by August. After some disappointing earnings reports later in 2020, I began reducing my position in increments around $80-$90 and closed it in early 2021. The stock continued to underperform in 2021, dropping below $10 in 2022. While software infrastructure stocks in general have lagged in 2022 due to macro effects, FSLY began its slide in 2021. Yet, peers were hitting all-time highs.
Audio Version
View all Podcast Episodes and Subscribe
Strategically, Fastly’s product portfolio is well positioned to capitalize on the confluence of several secular trends in software and security infrastructure, including edge compute, application security and distributed content delivery. In 2020, as enterprises rapidly shifted to digital channels for consumer and employee services, Fastly experienced a surge of interest. Revenue growth shot up to 62% y/y in Q2 2020 from 38% in Q1 and 34% a year prior. Additionally, operating margin quickly passed break-even on a non-GAAP basis and gross margin improved 600 bps year/year to 62%, both reflecting improving operating leverage. Their customer list included many Internet leaders, like TikTok, Shopify, Stripe, Spotify, Pinterest, GitHub, Twitter and more.
In addition to offering products for basic CDN and video streaming, Fastly had build a fairly robust edge compute solution, called Compute@Edge. It offers a serverless runtime that is distributed across Fastly’s PoPs in almost 80 cities and 35 countries. Developers can program in familiar languages like Rust, C, C++, JavaScript and most recently Go (in beta). With a custom-built runtime environment, Compute@Edge supports near instantaneous startup times.
Fastly’s acquisition of Signal Sciences in August 2020 accelerated their application security capabilities and bootstrapped a new product offering. Fastly had already built basic WAF and DDOS services. The Signal Sciences acquisition would be packaged into an adjacent market offering labeled Secure@Edge, which effectively doubled their TAM.
Signal Sciences was already highly rated for their WAF and RASP offerings, and had a stable of marquee customers as well, including Under Armor, Datadog, Duo, One Medical, DoorDash, Onelogin, Airtable, Postmates and Procore. The opportunity to cross-sell between Fastly’s and Signal Sciences’ enterprise customers appeared significant. Signal Sciences also brought $28M of ARR, higher revenue growth than Fastly’s 62% and gross margins in the 80% range.
At their Altitude User Conference in November 2020, Fastly leadership teased more product categories, expanding their planned market penetration even further. These included Perform@Edge, which would enhance their existing content delivery offerings, and Observe@Edge, which promised new capabilities for application tracing, advanced analytics and alerting. Finally, they had been hinting at a broad expansion of their data storage capabilities (referred to as Data@Edge), moving far beyond the rudimentary key-value store in Edge Dictionaries to a full-featured distributed data store with a SQL-like interface.
While the product roadmap appeared very promising in the second half of 2020, little of it materialized within a reasonable timeframe. Compute@Edge remained in beta for an inordinate amount of time, finally going GA in November 2021. Secure@Edge and other @edge products were never launched in the form described during the 2020 user conference. The full integration of Signal Sciences WAF and Compute@Edge was delivered in Q1 2022.
A review of Fastly’s quarterly release notes is pretty underwhelming. Granted, it’s meant to be a summary, but a paragraph for each quarter seems limiting. The descriptions for the past 4 quarters fit on a single web page. This is in contrast to other software infrastructure companies with a rapid product development cadence. Competitor Cloudflare would need multiple pages to cover a year’s worth of releases.
Along with product development, financial metrics worsened in 2021, while software infrastructure peers were putting up record performance. Fastly’s revenue growth decelerated into the 20% range and gross margins regressed. Profitability also started backsliding, from positive operating income in 2020 to consistently negative operating margin. This culminated with the decision to replace CEO Joshua Bixby in May 2022, coinciding with the Q1 2022 earnings report. In August, they announced that Todd Nightingale will be replacing him, effective September 1, 2022. Nightingale joined Fastly from Cisco, where he led business strategy and development efforts for Cisco’s multi-billion dollar networking portfolio as Executive Vice President and General Manager of Enterprise Networking and Cloud. He joined Cisco in 2012, through an acquisition. Prior to that, he held leadership roles at Motorola. This should represent an improvement. It may also be positioning Fastly for an acquisition. In the past, Cisco has been floated as a potential acquirer.
The new CEO has been on the job for a few months and has already laid out opportunities to increase revenue growth and improve operating margins going into 2023. As a first step, let’s take a look at the Q3 earnings results.
Sponsored by Cestrian Capital Research
Cestrian Capital Research provides extensive investor education content, including a free stocks board focused on helping people become better investors, webinars covering market direction and deep dives on individual stocks in order to teach financial and technical analysis.
The Cestrian Tech Select newsletter delivers professional investment research on the technology sector, presented in an easy-to-use, down-to-earth style. Sign-up for the basic newsletter is free, with an option to subscribe for deeper coverage.
Software Stack Investing members can subscribe to the premium version of the newsletter with a 33% discount.
Cestrian Capital Research’s services are a great complement to Software Stack Investing, as they offer investor education and financial analysis that go beyond the scope of this blog. The Tech Select newsletter covers a broad range of technology companies with a deep focus on financial and chart analysis.
Q3 Results
Fastly issued their Q3 earnings report on November 2nd. The results were better than expected on revenue and EPS. The stock popped following the report, increasing about 3% after hours and then closed up 11% the next day. To be fair, FSLY stock had been pushed below $10 a share by this point, allowing for a larger swing. As of the close this week (Dec 16, 2022), FSLY stock is at about the same price as the post earnings close.
Revenue in Q3 was a record $108.5M, exceeding the high end of the company’s guidance for $102M-$105M. Analysts were looking for $103.7M, delivering a beat of about 4.6%. This represented 25.1% annual growth and 5.8% growth sequentially. Q2’s growth was 20.6% annually and 0% sequentially, so Q3 showed a bit of acceleration. Looking forward, leadership projects a revenue range of $112M-$116M for Q4, representing 16.7% annual growth. This beat the analyst estimate for $112.1M and would equate to 5.1% sequential growth at the midpoint. Coming out of Q2, Fastly projected almost no sequential growth in their initial estimate for Q3 revenue.
Fastly leadership also raised the full year revenue estimate from the range from Q2 of $415M-$425M to $425M-$429M with the Q3 report. At the midpoint of both ranges, this represented a $7M raise, which exceeded the Q3 beat by about $2M. In Q1, they had set the full year estimate at a range of $405M-$415M. Given that some peers have been keeping their full year revenue estimates the same each quarter in 2022, it’s nice to see Fastly gradually increasing this. The latest range would generate an annual growth rate of 20.5%. Leadership discussed the intention to increase this growth rate in 2023.
Shifting to profitability, Fastly is focusing on margin improvements after some backsliding. Fastly’s Non-GAAP gross margin decreased from 57.5% a year ago to 53.6% in Q3. This did show a nice improvement from Q2’s 50.4%, however. Fastly leadership sees more opportunity to improve gross margin going forward and expects another 200 bps increase in Q4 (implying close to 60% Non-GAAP gross margin). To reach this goal, they have several ongoing efforts that are addressing gross margin performance. These include reduction of some PoP duplication expense as they upgrade their network topology, better alignment of capacity with customer usage patterns and a reduction in bandwidth costs through better use of network peering relationships. Fastly leadership has set a medium-term target for Non-GAAP gross margin in the low 60% range.
Non-GAAP operating loss was ($19.8M) in Q3, representing an operating margin of -18.2%. This is worse than the year ago period with ($12.9M) operating loss for a margin of -14.9%. Compared to the prior quarter, though, Fastly also showed sequential improvement in operating margin from -26.2% in Q2. The operating loss translated into Non-GAAP EPS of $(0.14) for Q3, which beat the analyst estimate for ($0.17) and the company’s own guidance from Q2 for $(0.18) – $(0.15).
Looking forward, Fastly leadership estimated a range of $(18.0M) to $(14.0M) of operating loss in Q4. This would be an improvement over Q3’s performance by about $3.8M at the midpoint. They lowered the full year target range from $(78.0M) – $(72.0M) to $(82.0M) – $(78.0M), however. On an EPS basis, analysts had modeled a loss of $(0.10) for Q4, versus Fastly’s actual estimate for $(0.15) – $(0.11) Non-GAAP EPS.
Management acknowledged the poor performance on operating margin in spite of the revenue beat. They attribute this to heavy investment in Sales and Marketing in the second half of the year to drive revenue growth for 2023. The CFO anticipates that sales and marketing expense will increase sequentially from Q3 to Q4, while R&D and G&A remain flat. Additionally, leadership sees opportunities to drive further efficiencies in business operations, especially across G&A. The new CEO expects to meaningfully reduce operating loss in 2023.
Other key metrics showed some progress. Fastly added 31 customers in Q3, bringing the total to 2,925. This compares to 2,748 customers in Q3 2021 for 6.4% annual growth. The sequential growth in total customers was just 1.1% in Q3, and has slowed substantially in the last two quarters. Q3 additions were at least higher than Q2’s 14 new customers, which was up from Q1 by 0.5% sequentially. Prior to 2022, total customer additions each quarter were averaging around 100. The heavier investment in S&M is intended to increase new customer lands in 2023.
Fastly defines enterprise customers as those spending more than $100k in the prior 12 month period. Fastly ended Q3 with 482 of these larger customers, representing 16.5% of total customers. As compared to other software companies, this concentration is relatively high. This highlights Fastly’s go-to-market strategy, which currently focuses on targeting the largest Internet properties. Fastly added 11 enterprise customers in Q3, up 2.3% sequentially and increased by 152 in the last year for 12.1% annual growth. Growth in enterprise customer counts is higher than in total customers. The contribution to total revenue from enterprise customers is now 89%, up from 88% last quarter.
Because of the slowdown in total customer additions and concentration on larger customers, most of Fastly’s new revenue growth is coming from existing customers. This is captured by their net retention rates. Fastly provides a couple of different metrics to measure the change in spend from existing customers over time. The first is the traditional DBNER (Dollar-Based Net Expansion Rate), which measures the increase in spend from existing customers when comparing the last 12 months to the 12 months before that. For Fastly, DBNER excludes the impact of customer churn, though, which inflates the numbers somewhat. In Q3, DBNER was 122%, up 2% over Q2 and 4% year/year. This represents Fastly’s highest DBNER in the last year.
Accounting for customer churn, Fastly provides the Net Retention Rate (NRR) LTM. This performs the same comparison of the prior 12 month periods as DBNER, but includes the impact of churn and spending reductions. As such, it is more comparable to the Net Retention Rates published by peer software infrastructure companies. NRR LTM was 118% in Q3, up 1% from Q2 and 4% better than a year ago. Fastly also publishes the average spend for their enterprise customers. As expected, this is growing nicely, up $29k in Q3, increasing from $730k to 759k for a 4.0% sequential increase. On an annual basis, average enterprise customer spend is up $61k or 8.7% annually.
This also supports the observation that most of Fastly’s revenue growth in the recent quarter was generated by increasing spend from large, existing customers. This represents a sound strategy for now, but Fastly will need to keep filling the sales funnel to capture the next set of customers that can drive expansion in the future. On the earnings call, Fastly leadership discussed the opportunity to pursue the high end of the mid-market customer segment to land new customers. This will be a big part of the focus of the increased S&M investment for 2023.
Fastly has been able to cross-sell multiple solutions into existing customer deals. For example, from content delivery, they are upselling their Compute@Edge product. Further, customers layer on application security capabilities, including their next generation WAF, gained through the Signal Sciences acquisition. For Q3, the contribution from Signal Sciences was 13% of revenue and is growing at 44% per year.
In the third quarter, we saw continued momentum in our portfolio expansion strategy with strong cross-selling activity in both Compute@Edge and security. We saw Compute@Edge cross-selling wins in marquee accounts like New Relic and with Canada’s leading content creation companies. With security, we saw cross-selling motion with one of the largest drugstore chains in Europe, which is now using our Next-Gen WAF on top of content delivery from Fastly, and from a major Japanese video game company, who is now using our Next-Gen WAF in addition to our content delivery and Compute@Edge capabilities.
Fastly Q3 Earnings Call, November 2022
A year ago, as part of the Q3 2021 earnings report, Fastly set a goal to reach $1B in revenue by 2025. If they finish this year with $430M in revenue (slight beat of current guidance), then they would need to generate a 32.5% annual growth rate from 2023 to 2025. Fastly just finished Q3 with 25% annual revenue growth, but estimated the next quarter at a lower growth rate. They are projecting to finish 2022 with a growth rate of about 20%. The $1B goal looks suspect at this point, but if accomplished, would represent an acceleration of growth going into the next 3 years. At this point, Fastly leadership has not changed this target, but the new CEO would presumably have the latitude for a one-time reset if needed.
As we continue to invest and execute against these opportunities, we are focused on achieving $1 billion in revenue by 2025. We have set clear milestones to drive success, including winning developers, enhancing our security offerings, and driving enterprise customer adoption. Combined with a strong roadmap of new products that we will discuss further in upcoming quarters, we are optimistic about the journey ahead of us.
Fastly Q3 2021 Earnings Call, November 2021
Investment Plan
Fastly’s current market cap hovers around $1.1B and has a trailing P/S ratio of 2.6. Given its poor progress on gross margins and operating margins, this low multiple is justified for 25% revenue growth. However, if Fastly does start to show improvement on profitability and push revenue growth closer to 30%, then the multiple could re-rate upwards. For comparison, Elastic (ESTC) just logged 28% revenue growth with break-even Non-GAAP EPS in their most recent quarter. ESTC has a trailing P/S ratio of 5.3. Their gross margins are higher at nearly 75%.
While Fastly’s near term target for Non-GAAP gross margin is 60%, they have product lines like application security (Signal Sciences acquisition) that enjoy much higher gross margins. This part of the business is growing faster than the rest (over 40% in Q3), but only makes up 13% of total revenue. At some point, application security could generate a larger share of overall revenue, pulling up gross margins further. Over the next 3 years, improvement in gross margins, operating income and the $1B revenue target could make FSLY an interesting investment. If we assume they increase the revenue growth rate to 30% annually for the next 3 years, they would end 2025 with $945M in revenue. Additionally, if they transition to positive operation margin and improved gross margins, the valuation multiple could re-rate up to a P/S ratio of 4-5. This implies a market cap of over $4B by 2025 for an increase of about 4x from its current $1.1B.
While I held FSLY stock previously, I am not planning to open a new position at this point. However, if revenue growth and profitability measures start to show improvement under the new CEO’s leadership, then FSLY stock could offer favorable upside. I might consider initiating a smaller position then. The other variable for investors to consider is acquisition potential. With a low market cap and a relevant product portfolio, Fastly might represent a nice tuck-in for a larger technology player (perhaps the CEO’s former company Cisco). I don’t purchase stocks in anticipation of acquisition, but that might provide another reason to own FSLY if execution improves, as the acquirer would pay a premium over an increasing valuation multiple.
NOTE: This article does not represent investment advice and is solely the author’s opinion for managing his own investment portfolio. Readers are expected to perform their own due diligence before making investment decisions. Please see the Disclaimer for more detail.
Thanks for the info and insight. Is the product development slower than Cloudflare’s because Fastly doesn’t have composable software?
Hi – Likely the architecture is a limitation in so far as they haven’t developed as many building blocks as Cloudflare beyond compute (like multiple data storage options). Also, culturally Fastly seems more inclined to hold products in beta for a longer amount of time. Perhaps this is to allow sufficient time to test, tune performance and fix bugs. This is where Cloudflare’s free tier is useful, as it provides a faster feedback loop on new product testing.
Given a sustained DBNER of >120%, Fastly can growth in the 30% range and re-accelerate. and assuming growth comes back into favor, price/sales of 10x will become more commonplace in the years ahead.
Your 4x assumption is something I deem to be conservative. Slight difference of opinion. I know people who work at both Fastly and Cloudflare and from the Bay Area. The pressure to succeed, as you well know, is felt palpably in the Bay Area. Long NET and FSLY as security/edge play out to our favor.
Thanks for the feedback. I would certainly welcome a re-rating of that size. I agree that if Fastly really turns around and re-accelerates growth with favorable operating leverage, then my estimates could be conservative.
Talk about a different cadence… Fastly hasn’t updated their Compute@Edge docs since May-22. (7mo ago)
Since then, competitor Cloudflare has had two dedicated Innovation Weeks advancing their developer ecosystem and their edge-native primitives since then (Platform Week, Developer Week) along with the slew of Workers ecosystem announcements in the broader GA Week & Birthday Week in Sep-22. (All of which Peter has covered well…)
Sources have informed me of a significant brain drain there. They’re moving slower for a reason, and they’ve installed some rather underqualified dummies, who are better at talking than innovating, in some important technical leadership positions from what I am hearing. A pop is definitely possible in the short term, but I would be careful long term.
Thanks, Paul. Not surprising to hear. As companies lose momentum, they generally lose talent. This is likely exacerbated by the CEO transitions. My interest in FSLY would be as a short term play, given where the valuation is now. If they can show some improvement in financials and the macro turns, then a rapid re-rating over a relatively short period might happen. However, in terms of innovation speed and product roadmap, Cloudflare is the leader. My allocation to NET is large and an entry into FSLY would be a small portion of my portfolio.