JacobRosenberg.net

 

Just the other day, I stumbled on Yet Another Startup CDN Provider, SimpleCDN. The CDN business has become very crowded as facilities-based carriers (AT&T and Level), start-ups (Limelight, Peer1, Panther Express), and the established leader (Akamai) engaged in a free-for-all over price. SimpleCDN’s value proposition is straightforward: pay a one-time up-front fee based on object size, instead of monthly based on usage. The big question I’m inclined to want answered is: how could this possibly work as a business?

This is a pretty unusual concept, so I decided to check it out myself. They let anyone sign up on their portal and start delivering without so much as a sales call. I wanted to ask some questions, so I went looking for real-world contact information. 

Here’s where the first problems became obvious: 

- The only address they provide (the Highland Park, IL one) appears to be a Mailboxes Etc location. 

- They also list some “NOC” mailing addresses on their contact pages. If they seem familiar, it’s because they’re major carrier neutral data centers. Perhaps they host something at these locations, but they sure don’t have people there. 

- There was, however, a 24/7 support address. So, I sent them some email.

I received an almost-immediate response (on a Sunday night) from Frank Wilson, who described himself as “Chief Engineer of SimpleCDN, and also a co-founder.” He offered to answer any questions, so I asked:

I’m assuming your data center bills you for power and space every month, and I’m sure you’re paying someone for transit monthly, so how exactly does this work over the long run? On the surface, it seems like you’re banking on the demand for content dropping dramatically over time, and the marginal cost to store it approaching zero. 

Assuming your cost for delivery is the 7 cents per GB transferred (which implies Cogent-level transit costs). At the 1MB = 1 credit = ~14GB xfer’d price point you’ve set, you’re collecting that one-time fee sufficiently to cover 14000 times while at the 1GB = 35 credits = 500GB which would only cover around 500 downloads. 

Days later, with no answer to my question, I decided to do a bit of poking around. Here are a few things I discovered:

1) The entire SimpleCDN operation seems to be hosted in a single Chicago location. Looking at looking glass servers from Europe, Asia, South America, and Australia reveals the exact same set of servers in Chicago serving everything. They own AS46177, but don’t seem to be using it.

2) All of their routes come via a self-declared ISP called Resisoft (AS32205). Don’t bother trying to click that link — they don’t seem to be a working web site. It isn’t clear if they’re troubled, or just host troublesome companies. It seems like most of their transit is Hurricane Electric, which isn’t exactly premium bandwidth. 

3) Frank Wilson seems to be the only visible employee of SimpleCDN–he answers their support email, manages their network, and markets their services. Yes, you heard that right: from sites from the technical web-hosting to stock boards to commenting on industry sites he pops up, promoting SimpleCDN. Sometimes he identifies as an employee, other times he talks about SimpleCDN in the third person, as if he weren’t affiliated. I find this very shifty.

Here’s what I think is happening: SimpleCDN is using current revenue from those initial “one-time” payments to cover their current bandwidth and hosting bills. As long as usage remains reasonable, and they keep growing their customer base, they can probably keep up with their costs for a little while. 

Except, this just isn’t a very good business model. Most of these customers will pay a few dollars a few times a year. The transaction charges eat alive a small company — even Amazon’s EC2 products aren’t really designed around someone using 5 instance-hours a month. And, all it takes is one really big download (Firefox 4, maybe?) to call their bluff.

So, worst case, they’re some sort of scam that will collect a bunch of pre-paid hosting charges and disappear. Best case, they have a naive business model which will cause their collapse if they get even moderately popular. I suppose I can understand why someone currently using S3 as a CDN (seriously, guys, why?) might consider this an appealing alternative, but it’s pretty hard to imagine that Frank’s business will be here in 2 years. 

Oh, and to Frank: I’d love to see a business plan that proves me wrong. I’m always interested in seeing legitimate competitors in the space, so show the world some evidence that SimpleCDN is more than just a great big promise that can’t go the distance!

Tags: akamai, cdn, content distribution networks, internet business, limelight, s3, simplecdn, web 2.0

This entry was posted on Saturday, August 9th, 2008 at 11:18 pm and is filed under Infrastructure, Internet. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.