Bear On Business

Sometimes when you add 2 + 2, it doesn’t equal 4 even if it looked like it did at first glance.  On Ike Elliott’s Telecosm blog, he probed into a chart that I posted on bearonbusiness.  We referenced the report “Estimating the Exaflood”, published in January, 2008 by Brett Swanson and George Gilder.  One of the authors, Swanson, commented on Ike’s blog:

“The chart you posted, which you found on [bearonbusines], is not our chart. George and I never claimed such 500-600% growth in 2007. Not even close. The posted chart appears to be a mix of two distinct data sets and does not even appear to be Cisco’s chart, which is the cite.”

Click here to see the bearonbusiness chart that caused the debate.  It is sharp looking, isn’t it?

Swanson and Gilder’s report contained two charts.  One spanned from 1990 through 2006.  The other spanned from 2005 to 2015.  One was titled U.S. Internet Traffic; the other titled U.S. IP Traffic Projection.  Note the similarity in names. 

Here is the first graph, as it appears in page 8 of the Exaflood report:

This chart shows growth from almost negligable terabytes in 1996 to 700,000 terabytes/month in 2006.  

On page 1 of the Swanson/Gilder report, Cisco estimates were used for 2006 through 2011 while Swanson’s was used for 2015.  In this chart, the 2007 appears to be 4.3 exabytes or 4,300,000 terabytes.  See below.

If you compare 2006 data on both reports, clearly the numbers are materially different (though 2006 is two years ago).  I understand that measuring Internet usage is an art, not a science, and we should have realized it was inappropriate to combine the two data sets.  Note, though, that Swanson combines Cisco’s estimate with his own future projection of 2015 in the chart above.  This is what were doing–but we also tried to tie it to Swanson/Gilder’s historical actuals. 

Our goal was to show data that spanned the bubble, the meltdown, and the resurrgence.  Clearly the text of the report is clear in their belief of growth between 2006 and 2007–and it is not the large step function implied by our co-mingled graph.  We made an error and will be more careful in the future.

Nonetheless, bandwidth is growing fast–and this will persist.

Today, Ike has a post titled Neutral Tandem: A Smart VoIP Business on his Telecosm blog. It is worth a quick peak, as I won’t repeat the ground Ike convered in his post. I will add a few thoughts.

Neutral Tandem is a public company. Their enterprise value was a staggering $510M in March. Their annualized revenue is probably in the $120M vicinity, which means the Neutral Tandem is trading at over 4X revenue. Nice!

I am not close enough to the company to know whether this is a sustainable valuation, though as a University of Chicago guy, I guess I should trust the stock exchange. In any case, Neutral Tandem is an example of how clever entrepreneurs can innovate as the telecom technology and ecosystem evolves.

[Ike Elliott, a friend and colleague, has a blog called Telecosm. It is worth reading. Ike had a blog entry on December 18th which I found insightful and fun to read. I thought the bearonbusiness.com readers would enjoy it.]

I am hoping to find a Kindle under the tree at Christmas. I think that Amazon has hit a home run with their new book-reading device.

What is so cool about it?  It has built-in 3G EV-DO hardware, so you can download newspapers and books wirelessly, wherever you are.  It has a built-in wireless data subscription on Sprint Nextel’s network, meaning you never get a bill from Sprint Nextel or any other wireless provider. The cost of the network access is built in to the price of the books and newspapers that you buy.

It is easy on the eyes, using E-Ink, which looks like regular ink on paper.  It automatically saves any books or papers that you buy up in the network, in case your Kindle dies, or in case you lose your Kindle, so you never have to worry about backing it up.

It is environmentally correct. Think of all the paper you avoid using!

I admire Amazon for daring to innovate on the business model for Kindle, and hiding the wireless data subscription costs from their subscribers. Think of it: what brick and mortar book store charges you extra for the truck that brought the books to the store?  What you really want to buy is a book, so why should you have to worry about renting space on the truck to get it to you? This is similar to the way that Amazon has free super-saver shipping for orders that are large enough, to help remove shipping costs as a barrier to sale. Amazon has done the same with Kindle, and just removed the distribution costs from the offer price for a book. And the book price is often lower than the paper book would be, because they saved the cost of printing the book and putting it on a truck.

Good product. I hope Amazon sells a ton of them. Now, if they would only integrate an RSS reader with it, for access to free web content, I wouldn’t even wait for Christmas to get one.

Please visit Ike Elliott’s Telecosm blog.

[Ike Elliott, a friend and colleague, has a blog called Telecosm.  It is worth reading.  Ike had a blog entry last week that reminded me of when I first met Ike and some of the initial VoIP experiences we shared.  Ike--remember SIOS?   With Ike's permission, this post is a re-print of Ike's blog entry.]

There was a good post by Daniel Berninger over on GigaOm today, recounting the 1995 paranoia about VoIP at AT&T Bell Labs.  Daniel laments that a dozen years later, VoIP still hasn’t eliminated the phone company business model. 

It took me down memory lane, because I was part of very similar conversations at MCI at the same time.  As an engineer that was leading some of the VoIP research at MCI back then, I was pulled in to meetings with executives every now and then to talk about VoIP, and I always felt like the execs couldn’t make up their minds about whether VoIP was a threat or an opportunity.  Sometimes they would say disparaging things about VoIP, such as “it is like ham radio, a toy for hobbyists.”   I saw it as a big opportunity for MCI, though, arguing that we finally had the tool we had been looking for to end our slavery to the circuit switch manufacturers.  Sure, our top line revenue might suffer as prices declined, but if we played our cards right, we could rapidly compress our costs as well by moving off of old circuit switched technology and onto an all-packet network, while focusing our attention on growing our Internet backbone.
Wow, I was naive. 

MCI ended up getting bought by Worldcom and never really bought into the risky strategy of replacing circuit switches with new networking technology based on VoIP.  In fact, while all of the major long-distance carriers did a lot of research on VoIP and applied for patents on the technology, none of them really wanted the technology to succeed and all of them tried to milk the circuit-switched cow for the next decade.  They watched their residential phone businesses shrink and were eventually acquired by their arch-enemies, the Regional Bell Operating Companies. 

But was it VoIP that caused the collapse of the long-distance carriers?  No, in the end it was a combination of wireless encroachment, lost regulatory battles, and corporate malfeasance (in the case of WorldCom) that did a lot more damage to the long-distance carriers than VoIP.  Sure, VoIP did help crater the high-margin international long-distance business, but that by itself wasn’t enough to kill the inter-exchange carriers. 

Like Mr. Berninger, I expected VoIP to wreak more havoc by now.  I wouldn’t have guessed back in 1995 that twelve years later, I would still be thinking VoIP’s best days are ahead of it.  My views are more nuanced now; I no longer believe VoIP wins in a frontal assault on the gates of the telco fortress.  VoIP is already winning in more subtle ways, as the technology is incorporated into mainstream product offerings (Comcast Digital Voice, for example), and into emerging standards (the IP Multimedia Subsystem, for example).  VoIP is sneaking its way into fixed-mobile convergence standards, too.   VoIP still wins in the end, but folks might not notice, or care, that it is VoIP under the hood.  

Please visit Ike Elliott’s Telecosm blog.