Tuesday, February 12, 2008

End to the Wi-Fi mirage

Earthlink is now trying to sell or close its municipal Wi-Fi operations — perhaps because it lost $80 million on municipal Wi-Fi last year (versus $20 million in 2006). Apparently some of the failed services are getting turned off. Earthlink also pulled back from the ill-fated Helio MVNO venture with SK Telecom.

This is probably the beginning of the end of attempts to build self-supporting municipal Wi-Fi systems (as opposed to those subsidized like parks and libraries as a “public good”).

The cause is not that different than Ricochet’s failure in the 1990s. Yes, Ricochet had slow nonstandard modems while Wi-Fi is ubiquitous and cheap. However, the economics of building infrastructure coverage are the same, and today the desirability and adoption of substitutes (other access methods) are probably worse.

Most seriously of all, it seems like the era of paid Wi-Fi is heading towards extinction — because free Wi-Fi at restaurants and coffee shops is becoming more the norm. I had lunch in Mountain View on Sunday and it seemed every bar on Murphy Street had free Wi-Fi. Starbucks announced Monday that it’s offering (limited) free Wi-Fito match all the other free sites — further fueling the commoditization of Wi-Fi hotspots. About the only place that people will pay for Wi-Fi is in an airport, because you can’t easily go down the street to find a better alternatives.

Without crunching the numbers, my intuition is that WiMax is going to face the same problem. Yes, the radios have longer range, but you still have to build more cells (and negotiate access and install backhaul) for lots of cells. If WiMax fails, then Sprint's 4G strategy will fail with it.

1 comment:

Doug k said...

The problem with wifi hotspots as a standalone business is that in reality there's too little value add. Even if you exclude the more forward thinking municipalities it's just too cheap and easy for Joe Bob's diner to stick an AP on the end of their DSL line and let customers have at it. You don't really need any sophisticated services or management since the use cases tend to be pretty simple, including the fact that the users aren't really 'mobile' like we think of in the cellular case, but rather are 'portable'. (Think of the original Osborne - it was useful since you could move it from place to place. You wouldn't consider using it while you were moving :).

I'm pretty pessimistic about wimax becoming the replacement for cellular. I am, however, quite bullish on its ability to really open up the 'public utility' model. Lighting up downtown Sunnyvale (or any similar area) just becomes free, which can only be a good thing for all sorts of interesting new applications.