When Google made its offer to San Francisco to do Wi-Fi for free, the first thought through many minds was that this Internet upstart will be upsetting the apple cart of many an incumbent service giant. First, the city by the Bay, then the world on Google Wi-Fi!

I've obviously been around telecom way too long because my first thought — or at least my second — was, “And how are they going to do customer service?”

In a world where there seem to be no barriers to entry in becoming a communications service provider, there remain serious barriers to success, and customer service is at the top of the list. That is true for municipalities and utility companies that want to provide broadband services, it's true for voice-over-IP providers and it remains true for incumbent service providers such as telcos and cable companies.

Spending $20 million to set up Wi-Fi hot spots in significant density to cover San Francisco is not a small task, but it pales in comparison to offering 24/7 assistance to the customer who can't get his Internet connection to work — even if that customer's problem has nothing to do with the Internet service.

The quick answer, of course, is to give the customers what they paid for — and since the ad-supported service is free, that would be no customer service at all. Many free and low-cost software and hardware providers address the issue by putting all help online and charging customers if they want to get real human assistance.

That may work in the tech world, but municipal networks such as San Francisco's are part of government-provided service that quickly becomes an entitlement.

Given the speed — 300 kb/s — of Google's planned offering, it clearly isn't aimed at businesses and high-end consumers, most of whom already have cable modems or DSL and are accustomed to broadband speeds. The market at which this service is aimed includes current dial-up users, Internet newbies, those who can't get a wired service because of geography and casual users, who want Internet everywhere. This is a crowd that will need some hand-holding — is that going to be the city's job or that of the service provider it chooses?

This is the same dilemma VoIP players face in trying to crack the mass market, which is one reason service providers such as EarthLink and AOL are viewed more seriously, even though they are a bit late to the party. If the network infrastructure required is no longer a billion-dollar, decade-to-build proposition, the internal operations infrastructure remains costly and time-consuming to construct.

That is what any long-term player will have to do, as past history shows. The cable industry learned the hard way in the 1990s, and the telcos suffered their own customer service debacle in rolling out DSL. Those who don't learn from that history are doomed.