Amazon Says It Has 20,000 CDN Customers; But What Does That Really Mean?
On Tuesday, Amazon announced via their AWS blog that the company now has 20,000 active customers for their CloudFront CDN service. In addition, Amazon believes that their 20,000 customers would make Amazon CloudFront “the largest global CDN according to published customer counts“. The problem is that while the 20,000 number sounds impressive, without more granular details, there is no way to know what it really means.
If the majority of Amazon’s customers are start-ups, or developers, then the average revenue per user (ARPU) is probably a few hundred dollars a month, with many paying even less than that. And since Amazon does not require monthly commits, a customer could be here today, but gone next month which means their business would have a very high rate of churn. Amazon did define in their post that the 20,000 “active” customers means that the customer “has used CloudFront in the given month,” but we still don’t know what kind of volume the average customer is doing or even what percentage of their customers come from which verticals. In a follow up with the company, Amazon did verify for me that all of their “active” customers are in fact “paying” customers.
Is the number of customers a vendor has really the best way to distinguish who the “largest” CDN is? I would argue no. Customer count is one of the many data points that the industry should be looking at, but total revenue and profitability matters more. Would a CDN vendor rather have the most customers in the industry and be losing money, or have fewer customers and be profitable? One CDN could have many smaller customers, who don’t do as much volume and another CDN could have fewer customers who do much larger volume. One could be “larger” based on the customer count, but the other could be “larger” based on the volume of traffic going over the network. In the end, there is no one data point alone that can be looked to say who the largest CDN is. Multiple data points on customers, ARPU, churn, rate of growth, volume of traffic, capacity of network and profitability all have to be looked at together.
In Amazon’s case, they do have a big advantage over other CDNs in that they can afford to spend years building their CDN business, at a loss, and will still be around years from now to turn the business into something that is profitable when this market really begins to grow. Many other CDNs didn’t and still don’t have that advantage. So I do expect Amazon to be one of the few big players in the CDN space years from now and the company does have a lot of advantages with CloudFront thanks to their entire portfolio of Amazon Web Services (AWS).
Amazon won’t say how much revenue CloudFront does, which is pretty typical of the CDN vendors as a whole since most of them never discuss revenue or number of customers per service type. In August of this year I detailed that by my estimates, Amazon’s CloudFront service should do around $75M in revenue this year. Of course, that’s not a scientific number and Amazon has never given any guidance, but talking with some of Amazon’s competitors who watch what kind of volume and capacity the company has for their CDN service, some of Amazon’s competitors have expressed to me that they feel pretty confident that Amazon is well under $100M in CloudFront revenue for 2011.
All this aside, Amazon’s CloudFront service is seeing some dramatic growth in traffic volume and getting some nice traction in the market with mostly small and mid sized content owners. But Amazon has bigger ambitions and also plans to really start going after the larger content owners in the New Year which should make things interesting, especially since they bring so much transparency with pricing to the market. In the last couple quarters, I have gotten more questions from content owners about Amazon’s CDN service than any other vendor and Amazon has been pushing their service via marketing, speaking engagements and sponsoring of events. So word is getting out that Amazon’s CDN service is no longer something that is only thought of for developers and the company has rolled out an amazing amount of new features and support for the service in the past 18 months. (see below for a list)
Plus, it’s only a matter of time before Amazon starts offering services that fall under the “value add services” umbrella for CDN companies, especially when it comes to the acceleration of content. And one would expect Amazon to bring the same kind of transparency to those services as they did for CDN, which is going to make things very interesting considering those are the services that currently have high margins that so many CDNs now rely on.
It’s clear that Amazon’s CloudFront team really wants to show off the success they are having in the market, but have their hands tied by the fact that as a company, Amazon really never divulges too much info on their AWS platform. So it makes sense they would put out whatever data they can, even if it is just a customer count number that really does not tell us much. Over time, their customers will act as their mouth piece as we’ll hear from more and more of them using the service who are willing to talk about the success they are having with CloudFront. One thing though is for sure. Amazon is not going away, they are heavily investing in their AWS platform and they will be one of the leading CDNs in the market, for all kinds of content delivery services, years from now.
As a side note, Amazon currently has 341 900 job openings worldwide, just for their AWS group. That alone shows you the kind of long-term investment they are making in AWS.