I think Fastly is selling off so hard because of its run up, because of the not as awesome as hoped for guidance and because of Tik Tok fear.
Tik Tok was ~12% of revenue over the last 6 months and 50% of that was in the US.
So if Tik Tok gets banned that hurts.
I think Fastly is just fine long term and I’m keeping my shares ~15% position
What really jumps out at me reading through Fastly’s results is the deceleration of new wins of Enterprise customers. Enterprise customers make up 88% of Fastly’s revenue, and Fastly’s QoQ growth can therefore be put down largely to the increased spending of these Enterprise customers (DNBER of 137% up from 133% in Q1 20, average Enterprise customer spend of $716,000 versus Q1 $642,000). My understanding is that the increased usage has driven this increased customer spend to date, but that usage levels will be flat going forwards based on their full year guidance.
All good so far. However the number of Enterprise customers is as follows:
Q2 19: 262
Q3 19: 274 (+4.6%)
Q4 19: 288 (+5.1%)
Q1 20: 297 (+3.1%)
Q2 20: 304 (+2.3%)
Clearly the number of customers is not the be all and end all, depending on each customer’s relative size (for example signing a significant customer Shopify or TikTok is more meaningful than a SME) however the broader question remains why they only managed to sign 7 new customers in the last quarter despite there being an ongoing “demand for our mission critical services” and what is the significance of that. Is there a lag from signing a new customer to getting it up and running and recognising revenue perhaps? My concern would be a slowdown in signing up new Enterprise customers, Fastly’s core revenue stream, coupled with usage flat to current levels (as implied with their FY guidance) might not lead to the kind of growth rates that many of us are hoping for.
Working in Finance selling IOT to Enterprise customers I can attest that new business has not been easy to come by the last quarter, and perhaps Fastly is feeling that too despite their Covid usage tailwinds to date and their product which is ‘mission critical’ for many businesses.
Therefore while Tiktok at ~12% of Enterprise revenue would be certainly a risk, I am more concerned about the continued growth of new Enterprise customers as a whole if we can assume the usage tailwinds are now running at run-rate. Appreciate any thoughts on what is going to continue to accelerate Fastly’s growth from here, perhaps before ‘Compute Edge’ becomes realisable.
Enterprise customer growth something I’m watching, but not a red flag at this point. I’d expect that to slow a bit in the current environment and it’s pretty impressive that revenue is growing as fast as it is without bringing on lots of new large customers.
Also very strong for increased Enterprise customer spend jumping up to $716k
Q2 20: 304 (+2.3%)
They mentioned this number includes a number of customers who dropped out of the category because they were in covid affected verticals. I would have also liked to see this growth be higher though.
Great questions and a really good catch on you part. For me, I’d just be careful not to read too much into it given we have such a small base of both customers (+12, 14, 9, 7) and quarters. And only one YoY comparison (+42 or 16% YoY). Of course, you can only read what you have so the progression is certainly worth continued attention. I’d agree “demand for our mission critical services” is playing it up a bit. Just not sure there’s a definitive answer or trend at this point.
“it’s pretty impressive that revenue is growing as fast as it is without bringing on lots of new large customers. Also very strong for increased Enterprise customer spend jumping up to $716k”
I guess my point is that a spike in usage (due to Covid) has driven this QoQ acceleration and increase in customer spending. But now that usage is at ‘peak’ levels presumably, what is going to continue to accelerate this growth going into Q3. They beat Q2 estimates due to a spike in usage, but how are they going to beat Q3 estimates.
Perhaps their forecasting flat in Q3 & spike in Q4 due to seasonality (thanks Rafe’s modelling post no. 70339) is realistic
Personally I was actually hoping for more of a usage driven spike in customer spend than we’ve seen, given the spikes in traffic from PINS, Shopify, Amazon, TikTok & Etsy that we’ve seen recently. All things being equal increased usage = increased revenue, albeit not at a 1-1 correlation. Perhaps I (and others) had overestimated this correlation. These customers had their traffic explode this quarter, I was expecting a similar explosion of Fastly, and in my opinion an acceleration to 62% YoY growth (while in isolation is healthy) was at the lower range of my expectations - and is probably why the market has reacted the way it has. It was priced in.
Ultimately I am happy with 62% growth this quarter despite these increased expectations, the key for me is understanding its growth acceleration going forwards.
“They mentioned this number includes a number of customers who dropped out of the category because they were in covid affected verticals”
Agree that churn will offset some of this new business, and that if they were losing some ‘small fry’ to focus on larger higher quality customers that might be no bad thing. Is there any information available over how much revenue they lost through churn vs how much brought in through new business? And what is their strategy on new business, is there a focus on signing up larger Corporate clients on bigger contracts?
Another factor to consider is whether they have been offering incentives/discounts to stimulate new business growth. Perhaps I am misunderstanding their business model, but I was expecting the spike in traffic from its customers & usage to indicate that there would be an inherent demand for their services, perhaps even a backlog. Other factors include what is the lead time from signing a new customer to fully on-boarding them on their platform, and are customers signed with any minimum revenue commitment or is it entirely usage based (which could be a risk, of which the upside seems capped based on the above explosions of traffic/revenue correlation). For example, signing up Shopify might not mean that the benefits of Shopify’s usage have been fully realised yet, or is Fastly already responsible for 100% of Shopify’s traffic.
Put simply, I see Fastly’s core revenue as follows:
No. of Enterprise customers x DNBER + new Enterprise customers
So if we are saying that the growth of DNBER is due to increased usage, but usage is to remain flat to current levels going forward, and that the number of new Enterprise customers is slowing - this represents a risk to me.
Perhaps I am underestimating how much these new Enterprise customers are worth, or misunderstanding something about their business model. It is entirely possible that they have churned low quality customers and are focusing on a few high potential customers, but I haven’t heard anything about this?
I’d just be careful not to read too much into it given we have such a small base of both customers (+12, 14, 9, 7) and quarters
I agree that in itself the number of customers is not the be all and end all, because there are gaps to be filled from my points above. It would be more meaningful if we were to understand the absolute revenue contribution from new business in each quarter, but I don’t suppose that detail is available? I feel like I could be missing something fundamental here.
Really I am trying to get an insight into whether 1. That DNBER value is likely to accelerate further (eg from selling its other services, or even increased traffic from current run-rate and 2. if there are significant new customers in the pipeline. 1&2 together should lead to the growth we want. However without 1&2 there could be a slowdown. The number of Enterprise customers, while not a determining factor in itself, could be an indicator of this.
From Fastly’s Q2 report on Enterprise customers:
Continued Momentum with Strong Enterprise Customer Growth
Enterprise customers (defined as spending $100,000 or more in a twelve-month period) continue to be our primary revenue driver, generating 88% of our trailing twelve-month revenue for the calendar year. We focus on partnering with large companies that have the capability and available budget to increase engagement on our platform. We highly value these customers because they tend to have the quickest ramp in usage and the highest demand for ancillary products and features, as well as the ability to provide us with a certain degree of visibility and stability relative to other market segments in uncertain times.
During Q2 2020, we further expanded our enterprise customer base. This was driven by further adoption of our platform by both new and existing large customers. We increased our enterprise customer count to 304 from 297 in the previous quarter, increased our Dollar-Based Net Expansion Rate (DBNER) to 137%, as well as increased our average enterprise customer spend to $716,000 from $642,0002 in the previous quarter. Even though enterprise customer count can be impacted by a number of factors that impact spending, including customer consolidation and usage variation, we were pleased to see a net increase in our number of enterprise customers. The increase in spending is due primarily to their usage on our platform and further adoption of optimization and security products such as Media Shield, Cloud Optimizer, WAF, and our TLS services.
In addition to our growing enterprise count, our strong DBNER results reflect the success of our land and expand strategy. Once a customer chooses to utilize our platform, our historical experience is that they realize our value and migrate more traffic or purchase more software solutions. To illustrate, in the second quarter, we won significant additional business with a leading global payments company that onboarded onto our platform at the beginning of the year. This customer added Security and Dynamic Site Acceleration services to their core offerings and has continued to move more application logic to our edge enabling it to move faster and securely on its digital transformation journey.
Me: I think this provides a good overview of what Fastly’s strategy is for its focus on Enterprise customers (> $100k customer spend) and it touches on a quicker ramp up in usage for these customers being a driver of this strategy, but it would be great to hear if anyone has any more detailed insights into this strategy in relation to continued growth of new Enterprise customers, or shed some light on what the other factors are that can impact Enterprise customer as is quoted above.
I believe FSLY is working on finding a right salesman(EVP)for their poor enterprise customers win. They have advantage to choice right person at this COVID period and they also not rush to need great EVP for now. So JB himself do this job in the same time.
About TikTok impact, I think they already wiped out this part(at least a huge discount on that) on their guidance. As JB said on ER, TikTok have 12% their sales and US probably 50% of it. So it’s 6% impact. As a 60’s growther. I believe even the worst situation (TikTok got banned), FSLY will still move on. The recently price move just a dog went too far.
I believe FSLY is working on finding a right salesman(EVP)for their poor enterprise customers win
I believe FSLY is working on finding a right salesman(EVP)for their poor enterprise customers win
This would be interesting if so… do we have a source?
Perhaps I am understating the DNBER growth that Fastly can gain through additional business of existing customers, and then new customer growth becomes less important in parallel. The big question is what is the wallet share opportunity of its existing base, and will its existing customers’ traffic growth be enough to continue to accelerate Fastly’s own usage growth. On Shopify’s website they have Fastly down as their only CDN (“Shopify provides merchants a world class CDN run by Fastly”) perhaps suggesting 100% of its traffic through Fastly - but that’s just one customer, it would be good to understand the wallet share opportunity to go after for the rest of their customer base.
(apologies I accidentally pressed enter & posted this before - I am not sure if it’s possible to delete or edit posts, but please delete my previous one)
Muji referred to our pick and axe approach the other week.
If you believe in the Fastly engineering culture and fascination of solving compute problems at the edge, then the Fastly platform has a fair chance to continue to deliver the best axe for the TikTok use case (regardless if delivered by Bytedance or otherwise).
Short term potential pain, medium long term certain opportunity with this proven use case.
Nik
Long fastly, my last add at $113 
This reminds me a lot of the Twilio - Uber situation a few years ago. Twilio ended up just fine.
Also… the fact that Tik Tok uses Fastly says a lot imo. And Tik Tok is just one app… apps come and go, but why wouldn’t the next one or the one after choose Fastly too if it was the best option for Tik Tok?
Everything about this screams short term headwinds for a fantastic business that the market got too excited about over the last couple months.
Holding my shares.
Austin I think it’s more accurate to say that Fastly is ONE of the CDN providers that Tik Tok uses. Because of Tik Tok’s size, it makes up a large part of Fastly’s revenue. Other apps might choose to use Fastly but so far the customer growth numbers have not shown that story. (Of all of the large enterprises out there, they only have 300 spending 100k+ a year? contrast that to customer growth numbers from CRWD/DDOG/ZM, which deserve the valuation they get)
The parallel you share about Twilio is a good one, but to provide a counter example that I can think of from Twilio is right after the Sendgrid acquisition where the company was reporting (short term) blowout growth (I believe it was 80%, mid 50s organic), and there was a lot of exuberance over “Flex” being the future innovation/techology that was going to continue to accelerate growth rate. Time will tell which scenario Fastly is in right now.
Also… the fact that Tik Tok uses Fastly says a lot imo. And Tik Tok is just one app… apps come and go, but why wouldn’t the next one or the one after choose Fastly too if it was the best option for Tik Tok?
Agreed, TikTok is just one (albeit a significant) customer. I’m more concerned about the growth of Enterprise customers as a whole, than Fastly as a proposition.
Everything about this screams short term headwinds for a fantastic business that the market got too excited about over the last couple months.
Playing devil’s advocate slightly, but shouldn’t Fastly be benefitting from short term tailwinds rather than headwinds. In their report they say “we believe the ongoing pandemic has permanently accelerated the need for businesses to focus on digital transformation”, and some of their significant customers’ traffic (Spotify/PINS/TikTok/Etsy etc) has exploded. Shouldn’t this only accelerate Fastly short term?
Fastly also reported “our largest quarterly increase since our IPO” on total customers, which makes me think why would they have any trouble converting new Enterprise customers with this general demand behind them, particularly as its stated target customers are Enterprise. Notably Enterprise customers contribution has stayed fixed at 88%, indicating their greater impact on DBNER than non-Enterprise customers. Perhaps one possible reason is that it takes more leg work & lead time to sign up a new Enterprise customer than a non Enterprise customer - and (who knows) there could be a backlog of Enterprise customers to be caught up in Q3 and beyond.
There is a further angle over what % of non-Enterprise customers have potential to grow into an Enterprise customer (trailing twelve month revenue > $100k). While Enterprise customers have been fixed at 86-88% over the last year, might this proportion shift >90% as many new customers signed in 2020 have 12 months revenue behind them. Or do all customers start out as ‘non-Enterprise’ by default until they have billed $100k, even the Shopify’s or TikTok. In which case total customers might well be a viable indicator of future Enterprise customer growth (good news if so!)
Compute Edge seems to me something of an unknown at present, so I am keen to understand that what has been driving Fastly’s growth so far will continue.
For what it’s worth, I am also holding. I would just like some more clarity over these points
This would be interesting if so… do we have a source?
This was stated in the earnings call in response to a (partially inaudible) question about sales. They are “looking for a leader” .They also said that currently their sales result were good…(??) Meaning during the current 3Q period.
In response to a question about traffic they said they saw growth in Q2 continuing into Q3. Also that Q2 was unique in regard to growth and Q3 would be different.
Perhaps their forecasting flat in Q3 & spike in Q4 due to seasonality
Hi Athinkingfool,
From the conference call I got the following thoughts if I understand what they were saying:
During the second quarter usage was hugely accelerated because of work from home, school from home, more time at home, more streaming videos, gaming, etc etc.
Third quarter, while usage will be up it will be more like a normal quarter as it is summertime. Everyone is spending more time outside, just the way they do in a normal, un-covid year. Kids are home, but they are home in the summer normally. It’s just less contrast with a normal year. So they are “only” guiding to 50% or so growth.
And I think they are also being conservative too because they don’t know what will happen with TikTok.
Saul
The big slowdown in new customers came with Covid. We might consider whether number of new customers and revenues are driven by the same things. I suspect the economic contraction driven by Covid has led enterprises to concentrate IT spending on existing software , cyber security, and telecommunications.
It is late in the game to be investing in purely Covid driven software.
https://siliconangle.com/2020/08/06/aws-rolls-wavelength-edg…
Published on August 6th,
“Wavelength, an edge computing service that Amazon Web Services Inc. is delivering with help from wireless carriers, today became generally available in Boston and San Jose, California. Use cases for the service range from video streaming to medical image analysis and factory monitoring. Over the next few weeks, AWS plans to extend availability from San Jose to the rest of the San Francisco Bay Area. It’s also working with Verizon to add additional U.S. locations later this year.”
“LG Electronics Inc. is employing Wavelength to develop low-latency road safety services for vehicles, while Tata Consultancy Services Ltd. used it in an application that provides automated quality inspection of assembly lines. Sony Corp. has also adopted Wavelength to support real-time video streaming use cases. AWS’ long-term goal is to take Wavelength global. The cloud giant detailed today that it’s also working with multiple international telecommunications providers to make the service available in Europe, South Korea and Japan.”
Just figured I’d post this bit of news and a few interesting use cases for ‘edge computing’. It reminds me of when Amazon announced a product that competed with MongoDB. I obviously have zero idea of the impact this may have on the competitive landscape or FSLY and NET’s short term stock price, but it at least demonstrates that the TAM is indeed very large if AMZN thinks it worthwhile to get involved. The wording makes it seem like some companies have already begun using the product.
This is a shy first post - I hope it’s net beneficial. Thanks for all you guys do.