The Internet Is Global, But The ISP Business Is local

Last night I was having dinner with C.J. Cherng, the very able CEO of Taiwanese ISP Seednet, a Fon partner. During dinner we were talking about how global the Internet is and yet how different local rules are for ISPs around the world. On one side, there´s the US, who invented liberalization and, after being successfully copied by the whole world, retreated into a duopoly situation that is hurting the US consumer. On the other, there´s Europe, which implemented unbundling to such an extent that now European ISPs are the most competitive in the world. ISPs, like another Fon partner Neuf in France, have a menu of choices that would make any American forget about French rudeness and want to move there. For around $40 you get free national and international long distance, up to 20meg download speeds, a TIVO like service, cable TV channels, storage, mail, etc. And if you are an Internet fan and you happen to live in Sweden, then Labs2 offers you 100MB both ways for 89 euros a month.

Asia is somewhere in the middle of liberalized Europe and monopolized USA. Asia has implemented liberalization without unbundling. While ISPs can compete in Japan, Korea and Taiwan for example, not only do they use the copper of the incumbent but they also use the routers and DSLAMS provided by the incumbent. The result is that their offerings are unoriginal in terms of products (they can´t be creative without their own CPE), but their offers are very good. With Seednet you get 2 meg down and 256 kb up for the equivalent of 9 dollars or 7 euros, an incredibly low rates by global standards. On top of that you have to pay 10 dollars to the former monopoly for your regular phone line, a very competitive price as well. But Seednet as opposed to Neuf can´t offer innovative services because whatever is installed with the DSL service comes from the former monopoly. Whatever they want to offer that is original must come as an extra box and the incumbent´s box only has one ethernet port. In general Asians (by that I mean Japanese, Korean and Chinese) managed to create a more innovative rate environment but a less innovative product environment. The Asian market though is good for Fon because as Seednet must offer another box if it wants to offer WiFi by choosing Fon they not only offer WiFi but they offer social WiFi.

Now the paradox of the ISP business is that because of regulatory constraints while the Internet web site business is so global the Internet access business is very local. As a result there´s tremendous replication and fragmentation with each ISP having R&D teams that design practically the same services but with tiny differences that make it very hard for Fon work with. DSL is hardly a standard and there are great variations from country to country, ISP to ISP. DSL unbundling rules are very different. For Fon´s engineers it´s been incredibly tough to build the Fonera that works with all ISPs in the world. We have succeeded by now but only thanks to the ingenuity of our developers.

Given this environment of inefficiency in the ISP business, is there an opportunity for global consolidation of ISPs? I think so. Seednet for example is a small ISP but it´s highly profitable charging only 7 euros per customer and providing such luxuries as customer service over the phone which at that monthly rate with only one phone call a European customer would become unprofitable. But when I mentioned to the Seednet CEO that he must be an incredibly efficient ISP if he is able to make 12% profit margins with a rate of 7 euros providing DSL and that he should consider entering the European ISP markets using his technology I realized that he would not give this idea serious consideration. ISPs around the world are all very local and seem to want to stay that way. They don´t seem to mind replicating their R&D. Reinventing the wheel again and again with yet another new billing system, customer care system, deployment system, similar but not the same as their counterparts in other countries. Personally I do see an opportunity for a top ISP in one country to start buying ISPs around the world, there must be enormous economies of scale in doing so that nobody has exploited. And the business opportunity is gigantic, the market cap of these global ISPs would be in the tens of billions. But nobody has tried this in the extremely fragmented industry of Internet provisioning, large telcos like Telefonica do this with mobile services but rarely with fixed line services. Is this something I would get into personally? Probably not because I am more of an idea/product driven entrepreneur. But if I was running a large telco in a competitive country I would certainly think about this. Why should the Internet be global, the mobile business be global (witness Vodafone) and the ISP business be so local?

• bussiness
• globalization
• internet
• ISP