Expensive Hotels Really Do Have Faster Wi-Fi

OpenSignal, by means of mobile apps for iOS and Android, has been amassing data on Wi-Fi and cell-network signal strength. They released yesterday a few of their findings on the speed of Wi-Fi available at U.S. chain hotels (download speeds, specifically). Though it shouldn't be surprising that (as their data shows) more expensive hotels generally have faster speeds, I know it hasn't always matched my own experience. (Hotel chains also vary, even within brands, in whether the in-room Wi-Fi is free, cheap, or exorbitant.) If the in-room connection is flaky or expensive, though, from the same report it seems you'll do better by popping into a Google-networked Starbucks location than one fed by AT&T, and McDonalds beats Panera Bread by quite a bit.

How much is due to Congestion

[robstout]

If WIFI is free, everyone will use it, clogging up the pipes. If there's a charge, less people will be on, making more BW available for those who shell out the cash. I also hope that the hotels that charge use the money to miantain the infrastructure, but that's wishful thinking on my part.

Re:How much is due to Congestion

[Kjella]

Maybe it won't be good enough for Netflix streaming since they want to "upsell" you to their PPV service, but providing decent Internet should be no problem at all, not counting tropical islands and such but normal hotels in big cities. Heck, here at home I got 100/100 Mbit all to myself and it would easily let 100 people surf the web and check their mail as long as the hogs gets last priority. Hotels know they got you hooked on the convenience, you can get cheaper drinks at the corner store, cheaper food at a local restaurant and cheaper internet at an internet cafe but grabbing a drink from the minibar, order some room service and hook up the room wifi is so easy you'll overpay for it. Fortunately many hotels now consider it part of the basic service.

Useless data is useless

[mellon]

I get widely differing performance at different hotels in the same chain, for some chains, and consistent performance for others. And of course, different performance in the early evening than in the early morning. So these numbers are basically garbage.