"Perhaps the best-known network of hotspots is that operated by T-Mobile, a wireless operator, in over 2,000 Starbucks coffee shops in America. Around 25,000 people access the hotspots each week, which works out at an average of less than two users per day per hotspot. But the cost of connecting each hotspot to the internet is several hundred dollars a month. The Wi-Fi hotspot at Amsterdam's Schiphol airport is used by only a dozen people each day. (...)Excuse me ? 130 euro per month ?? That's about three times what I pay for my DSL-line at home... No wonder that they are desperate for customers. Looks a little bit like the problem with GPRS I wrote about a couple of weeks ago.
Users may be deterred by high prices. Even after a recent round of price cuts, using T-Mobile's network of hotspots costs $6 per hour, $40 per month, or $360 per year. Other operators in America charge $40-70 per month. Prices in Europe are as high as euro 130 ($150) per month. (...)"
So why does it cost "several hundred dollars a month" to connect a small hotspot, like a Starbucks, to the Internet ? After the initial investment for the router, the running costs shouldn't be much higher than the price for a normal DSL-Line.
Here in Hamburg, for example, we have a service called Hotspot-Hamburg, which is free to use for everyone. Every owner of a cafe or restaurant can get there a hotspot installed for just 50 Euro per month - which means 1,6 Euro per day. So he breaks even, if he sells 1 additional cappuccino every other day.
At these rates, the Internet access becomes part of the normal infrastructure of a cafe, which you use, while you are there - just like a chair or a table (Ever got asked if you want a chair with your latte ?). And providing the service for free, also saves the costs for things like billing or customer service.
So if there is a business model for Wi-Fi services in public places, like cafes, that's probably it: keep the running costs as low as possible, provide the service for free, tell everyone about it and recoup the investment with additional sales. Jupiter's Joseph Laszlo talked about a very good example for this on his weblog yesterday.
But there is still one problem left: who will pay for the large overhead of something like T-Mobile ? Well, I definitely won't...