Prediction: Fastly Will Surpass Level 3 and Limelight In 3 Years

Fastly Will Soon Be In the Tbps Club

Fastly, the inventor of the SuperRack is not only one the fastest performing CDNs on the planet, but also likely the fastest growing CDN in the bandwidth metric. The bandwidth metric (95/5%) is usually one of the most confidential metrics for a CDN. Rarely does a CDN announce the number publicly, because the bandwidth metric shows the world the State of the Union for that CDN. Once the bandwidth metric is known, the monthly recurring revenue can be guesstimated. Fastly, being who they are – have nothing to fear but fear itself, and that’s why they plaster the bandwidth metric on their home page. The screenshot is a bit small, so here are the numbers on the image.

  • Origin RPS: 171k
  • TTFB: .083
  • Hit Rate: 88%
  • Bandwidth: 635Gbps

Over the last last sixteen months, Fastly’s bandwidth metric has grown 3.8x. That’s “Phenomenal” for a CDN start-up.  My prediction is that Fastly will join the Tbps Club by year end, and surpass Limelight and Level 3 CDN in three years. Limelight and Level 3 are pushing somewhere in the 4- 6Tbps+ range. The most important thing to keep in mind is that it’s exceedingly more difficult to grow bandwidth from 0 to 1Tbps, then 1Tbps to 5Tbps, because at 1Tbps – everyone in the world knows who you are. Name recognition is everything to a CDN.

Fastly’s Bandwidth Growth

  • March 2014 – 166Gbps
  • June 2014 – 280Gbps
  • June 15, 2015 – 635Gbps
Digiprove sealCopyright secured by Digiprove © 2015

9 thoughts on “Prediction: Fastly Will Surpass Level 3 and Limelight In 3 Years”

  1. Where did you get your info about Limelight capacity being 4Tbps? Wikipedia put them at 11Tbps while their own analyst presentation back in February 2015 shows 9 Tbps.

  2. Fastly has a lot of work to make it into a position to surpass LL or L3. First off, the CDN world is changing and evolving into a multi-faceted services platform and not remaining a constant around small file caching or large file delivery, which has lousy margins, but is easy to scale and operate. Instant purging can only take you so far without real transformative solutions to surpass where Akamai and others are planning on being in 3 years.

    By your own thought process and math – Fastly would have to accelerate their growth considerably while LLNW and Level 3’s growth slowed by a greater margin in order for your prodiction to come true. Also growth by percentage it one thing when you are talking small numbers like 100Gbps and something verry different when relating them to 1Tbps.

    I’m not saying they can’t do it – I’m saying you are making this into a trivial pursuit and not a huge up hill slog.

    • What’s up brother. Good to hear from you.

      Fastly will become a video streaming powerhouse in due time. They are supporting quite a few live streaming customers today, including 24×7 broadcast channels. Thus, they have figured out streaming. This is where the volume will come from. I think they can reach 1Tbps by year end, then 3Tbps end of next year, and finally 5Tbps+ the following year. So its not the small file delivery that will get them there, but the streaming piece.

      Also, its widely known that many of Level 3’s technical CDN employees have either left or have been let go. These engineers are irreplaceable. L3 can never get them back. Just imagine some of the best CDN engineers in the industry with so much experience are gone to the competition. The worst part is L3 can never attract top CDN technical talent. Why would top talent go there when they can go work for Apple, Netflix, Google, or others more forward thinking CDNs.

      Now throw in the acquisition of TW Telecom, and then you start to get a full picture of L3’s CDN future. Today, the #1 priority for L3 is integrating TWT, not growing the CDN business. I personally think L3 will become a one trick pony that’s is a good fit for one type of customer – the Netflix and Hulu’s of the world that demand the cheapest possible price for a high volume of data transfer. Also, once Verizon Corp completely swallows Edgecast in a year or two, my guess is customers will flee. So if customers flee Level 3, and EdgeCast, there will be a trickle down effect for everyone, and Fastly will be there to pick up some of the good pieces. In regards to LLNW, they are doing a good job, and have steadied the ship. We’ll start seeing some decent growth in the next three years.

      • Again, I am not saying that Fastly can’t get there and I agree with most of your positions. Edgecast is in a spot right now where their growth went from start-up mode to Telco mode, which is a drastic difference. This will slow them down considerably; however, thinking through the growth from 1Tbps to say 5 Tbps is a massive undertaking not so much from an infrastructure perspective, but also from a offer and management perspective. Can the current team take it there in 3 years? I don’t know.

        It will be interesting to see how much money they can raise in the next year, who they can attact from a talent perspective and if they can become more than a self-service boutique with many pops. I like the team there and have nothing but respect for all they have accomplished, but I also honor that this market is not fluid and companies with already 10 times their traffic will not have their growther stunted either. All this means is that they have to grow and a 3-4X measure to the growth paths of the other competitors in the market to over take them on the time frame you are imagining, while investing a ton in diversified services, security, performance, optimization, and internal maturation. Its a tall order – that’s all I am saying – but if anybody has a chance, it’s Artur and team.

        • Well said. They have a strong foundation in place, and a feature rich platform to grow the business many fold. Now its an execution play. Can they execute and make it happen? I think they can. Also, I’m sure Fastly we’ll be announcing another round of funding soon.

Leave a Comment