Dear Jeff,
Good question, difficult to answer. But, I would like to take a brief crack at it. Please understand that the following are my personal thoughts, and should not be taken as any indication of PSINet's corporate position.
It is impossible to look ten years ahead in this business. Five is extremely difficult. Two is tough. But, in the space in which we are operating, I can see the following:
As you may understand, the Internet is one of the most global phenomena that exists. Its appeal to businesses is obvious, for almost all businesses today need to move information, people, goods and money. Businesses rely (the "mission critical" euphemism applies) on these things daily.
Expansion of the ISP business globally is something that will be undertaken only by a handful of providers, due to the constraints of time, competition, and effective personnel. Competition will continue to be extremely heavy. But outside the U.S., it takes only a handful of POPs and some good connectivity to cover the major business concentrations in a given country.
For example, companies like PSINet and UUNet have something like 250 POPs in the U.S. In a country such as Japan, with a huge economy, coverage of about 75% of the businesses can be achieved with only about a half dozen POPs. Likewise for countries such as France. Germany requires only a few more. Countries such as Italy or Spain only require a few POPs. This means that everyone in the global business will need to be in the same places; and one can see a case made that the businesses in a given area will not require more than a fairly small number of ISPs.
So, while there are numerous local/regional ISPs in any given country, you will find that few, if any, are global in that country. This will change, and the change will occur in the next two to five years. The lag time is a function of (i) telco deregulation in a given country, (ii) proliferation of computers in the business environment, (iii) availability of bandwidth to connect elsewhere, and (iv) (to a lesser extent) the business model in the country.
In our experience, the UK and Canada are the most like the U.S. in the business model and proliferation of computers. Telco deregulation is a fact. So, it is not surprising to see companies like PSINet and UUNet going to those countries with major early efforts to gain international presence. As you can also see from our press releases and other information, Europe figures strongly in our growth plans, consistent with the factors outlined above.
A couple years ago, one read a lot about the "consolidation" of the Internet. While top-level ISPs have been bought by Telcos (PSINet being the main exception), local and regional ISPs have proliferated. I do not see that proliferation lessening. But the backbone, facilities-based global network providers will, I think, remain few.
I hope this wordy response doesn't violate some basic principles of the Thread. Let me know if this helps. |
